ERM

List the trade-offs of each approach of risk aggregation methods

- Accuracy (model and numerical) - Consistency of methodology - Calibration: data availability - Intuitiveness and ease of communication - Flexibility - Resources required

What does the notation ci mean for the default risk notation in Jorion Ch. 18?

ci is the cumulative default rate

What does the notation di mean for the default risk notation in Jorion Ch. 18?

di is the (conditional) marginal or annual default between year i - 1 and i

What does the notation ki mean for the default risk notation in Jorion Ch. 18?

ki is the (unconditional) marginal or annual default between year i - 1 and i

List the possible modeling approaches with respect to catastrophic underwriting (Cat UW) risk and reinsurance (RI) credit risk

(Possible) modeling approaches (increasing complexity): 1. Variance-covariance matrix - Determine RI Credit and Cat UW risk capital separately - Aggregate with correlations 2. Like #1, but capture RI Credit risk in the Cat UW's MRD - Will result in higher Cat UW risk capital - May need a copula for residual dependency 3. More complexity: - Model reinsurance defaults and LGD as a function of insurance losses and/or asset values - Use stochastic interest rates to PV reinsurance losses

State the drawbacks of implied standard deviations (ISDs)

- A systematic bias could be introduced between the risk-neutral volatility and the actual volatility forecast - Implied volatility can change dramatically even over short time horizons

List at least 5 characteristics of strong economic capital models according to AM Best

- Address correlations with conservative assumptions on positive correlations - Contemplate increased correlations with larger events - Show the volatility in results - Reflect diversification, concentrations, macro economy, underwriting, reinsurance - Accepts deterministic scenarios for testing - Sufficient data on extreme events - Parameters fit well and reviewed regularly - Dedicated staff - Data quality reviewed/audited/tested - Output easy to read/understand - Results tied to objectives and default probability - Cash flow projections by scenario - Tested against historical adverese events

List the effects of transformations and outliers (with respect to correlation measures)

- All correlation measures are invariant under linear transformations - Only rank measures (Spearman and Kendall's tau) are invariant under non-linear monotonic transformations - Rank correlation is less affected by outliers

Describe walkaway / tear-up features - what is the exposure formula if a contract has this feature?

- Allow an institution to cancel transactions in the event that their counterparty defaults - An institution would walk away only if there was a negative MtM value, walking away from what they owed when the counterparty defaults - A walkaway agreement means an institution can cease payments and will not be obliged to settle money owed to a counterparty on a mark-to-market basis - Exposure with a walkaway feature = MtM

Describe target populations. State the three-step process that can be used to narrow the mortality gap between target groups and the general population

- An example of a target group is "healthier groups in the general population that act as targets to evaluate and focus medical intervention" - A public policy goal should be to narrow the gap between the morbidity/mortality experience of target and wider populations. This could be done with a three step process: 1. Identify an ideal/target population, either in terms of absence of disease or favorable risk factors 2. Specify interventions, either behavioral or treatment, that might benefit groups with worse experience than the target population 3. Evaluate through control studies how effective and efficient the intervention might be, and therefore how widely it should be adopted

Describe how actuaries may be involved with investment strategy with respect to climate risk

- Analysis of securities to determine climate risk exposure - Quantitative measures for securities (carbon footprint, etc.) - Risk scoring investment portfolios for climate risk - Engagement with companies on climate-related risk approach - Reporting on climate-related risk exposure in the investment portfolios - Development of products that: * Allow for climate-related risks * Seek a particular climate-related outcome (reduced carbon footprint) * Pursue a climate-related theme (renewable energy)

What are the complex duties of insurers?

- Answering to policyholders, shareholders, rating agencies, and tax agencies - Managing its group's overall balance sheet and capital adequacy - Reporting financials based on accounting requirements (if publicly traded)

Describe considerations for setting risk appetite for operational risk

- Appetite is usually qualitative instead of quantitative - Examples include brand management, anti-money laundering, compliance policies

Describe Step 5 of GSAM's ALM and SAA process: Risk/Return Trade-Offs

- Apply optimization techniques - Evaluate and compare various efficient frontiers

Describe de-biasing methods that may facilitate more rational behavior

- Ask individuals to think more critically (e.g. "consider the opposite") * Counteracts overconfidence and anchoring - Hold individuals accountable and make them justify their choices * Avoids "conjunction errors" caused by compound probabilities * Sometimes decision justification has the opposite intended effect - Recall relevance method - describe how plans unfolded in some scenario based on personal experience * Counteracts planning fallacy - Provide more financial education (doesn't work well with the planning fallacy) * More information causes "cue competition" (distracts from pertinent info) * Recommendation: make people aware of specific situations relevant to them - Frame risk over a different time horizon * Counteracts overconfidence - Ask individuals to estimate a range around their estimate of an event * Wide ranges are more likely to capture the correct answer

Describe the impact of behavioral biases and externalities on insurer resiliency efforts

- Behavioral biases: insurance decision-makers may underweight resiliency measures due to biases (e.g. tendencies to ignore information about extreme risk, too much emphasis on contingency plans) - Externalities: firms with positive externalities that optimize based on only their firm value may under-invest in resiliency (because they do not model the societal value of the positive externalities) - Summary: both biases and externalities may lead to underinvestment in resiliency efforts

List reasons why the theoretically optimal CML portfolio may not be feasible

- Borrowing cost could be too high - Optimal portfolio may require short-selling, which amy be prohibited - Investment income may be more volatile than the company desires - May be impractical/impossible to solve for the optimal portfolio for a given set of risk combinations Critical to define risk appetite to reflect risk aversion, risk limits, etc. - Ensures that TAA will not put capital/earnings at high risk

Describe contingent credit default swaps (CCDS)

- CCDSs are similar to CDSs but have a variable notional amount - The notional is indexed to the exposure on a contractually-specified derivative - They allow the synthetic transfer (to a third party) of counterparty risk related to a specific trade and counterparty - Can buy protection on an underlying company through a third party

State limitations of stochastic methods (according to ERM-120)

- Calculating a weighted average over multiple scenarios can be dangerous - This does not focus attention on the effect of actual scenarios - There may be actual scenarios with extreme conditions and potentially disastrous results

List general insurance claims considerations with respect to climate risk

- Can be difficult to identify trends in frequency/severity of large catastrophes - Past experience may not be a guide to the future - Agricultural insurance products can mitigate risks of adverse natural events - Carbon-based industries may receive lower investment in the future - Renewable energy producers may have increased needs for insurance - Coverage for climate-related liability risks (litigation) will need updating - Current catastrophe models may not adequately capture climate risks - Models may not be designed to quantify changes in climate risks

Describe the role of external auditors and regulators with respect to the Three Lines of Defense Model

- Can provide additional defense beyond the 3 lines - NOT substitutes for any of the 3 lines - Usually have more focused/narrow objectives * Regulator requirements strengthen governance and control - actively report on the organizations they regulate - External auditors assess the controls over financial reporting and related risks

How can risk appetite be used in capital allocation?

- Capital is not unlimited - should earn a sustainable return for shareholders - Must allocate total cpaital to BU level, then potentially classes within BUs * Statutory or rating agency - prescribed methods that don't reflect company's risk appetite and objectives * Economic capital - more granular but still constrained by statutory and rating agency capital - Using risk appetite to allocate capital is best * Considers risk tolerance and all constraints * Provides explicit probability of loss and maximum loss

Compare the 3 fatality rate measures relevant to epidemic modeling

- Casees should ideally be only those that have concluded in recovery or death * Difficult to determine early in pandemic * May be affected by sample bias: skewed toward severe cases - Not the same as qx's: CFR/HFR/IFR are independent of time * qx's are annual; CFR/HFR/IFR could be the same for a 10-day virus or year-long virus - IFR < CFR < HFR all else equal (i.e. within same region/group) * HFR focuses only on hospitalized cases, so it will be the highest * IFR has the largest denominator: includes milder (subclinical) cases

Explain managing forward counterparty risk with collateral

- Collateralization for a forward has the same role as a future's margin - If spot prices fall, long party must post more collateral - If spot prices rise, short party must post more collateral - Collateralization does not eliminate all default concerns * A new derivative could be harder to find and/more expensive

Regulatory action levels

- Company Action: 150% - 200% - submit corrective action plan - Regulatory Action: 100% - 150% - submit corrective action plan and Commissioner may issue an order specifying corrective actions - Authorized Control: 70% - 100% - Commissioner may place company under regulatory control - Mandatory Control: 70% or less - Commissioner must take regulatory control of the company

State ways that risk managers mitigate the challenges of stress testing

- Comparing ad hoc stress scenarios against the results of historical stress scenarios, reinforcing the recognition that large events do actually happen - Defining a robust risk appetite "envelope" (i.e. range) by risk category

Summarize the key idea of martingale tests in one sentence. What type of models do martingale tests apply to?

- Martingale Test - the average discounted cash flow should (approximately, up to simulation error) equal the time-zero value - Only applies to a risk-neutral framework

List the criteria that determine the appropriateness of a benchmark

- Contains a high proportion of the securities held in the portfolio - Low turnover of the benchmark's constituents - Benchmark allocations should be investable position sizes - Investor's active position should be given relative to the benchmark - Variability of the portfolio relative to the benchmark should be lower than its volatility relative to the market portfolio - Correlation between Rx - Ru and Rb - Ru should be strongly positive - Correlation between Rx - Rb and Rb - Ru should be ≈ 0 - Similar exposure style between benchmark and portfolio

Describe credit exposure

- Credit exposure is defined as the cost of replacing the transaction if the counterparty defaults - Also referred to as simply exposure - Exposure = max(MtM, 0) - Assumes no recovery - The loss in the event of a counterparty default - Conditional on counterparty default

Which is broader - credit risk or counterparty risk?

- Credit risk is broader - Counterparty risk is a subset of credit risk - A risky bond is an example of an instrument that only has credit risk

Describe data profiling

- Data profiling uses different kinds of descriptive techniques and statistics (e.g. min/max/average) as well as aggregate measures, such as count and sum, to analyze data according to known patterns - Using these, experts can flag unexpected values that are potentially incorrect - Profiling can help insurers identify missing values, which can then be replaced by more logical values

State the factors for loss given default (LGD)

- Debt priority - Industry - Asset type - Legal environment

State the high-level model development steps

- Define and develop specifications for the model - Develop or modify code/program the model - Test/validate the model - Approve the model for implementation - Validate that controls are in place and documentation is complete

Describe examples of reviews and testing procedures used for model validation

- Design use/fit - Design methods/processing - Data - Assumptions - Actual results - Sound/stable results - Results communication - Governance

Why do life insurers struggle under a prolonged low interest rate environment?

- Difficult to earn returns on the asset portfolio that were assumed during pricing of the life insurance products - A rapid decline in interest rates can hurt insurance companies due to their convexity exposure * Possible for the duration of liabilities to increase by a greater amount than the duration of assets - Difficult to immunize traditional guaranteed products with long durations * Companies may not want to invest in long-term, low-yield bonds and lock in losses

Describe the concentration of risk diversification benefits and how to quantify them

- Diversification reduces overall risk * Aggregate risk < sum of individual risks that are not perfectly correlated - Critical in both bottom-up and top-down approaches - Diversification assumption should be reasonable and protected from manipulation - Quantifying diversification benefits from different categories or regions: 1. Correlation matrix approach - aggregates risk factors using correlation factors (ρ's) 2. Copula approach - allows for nonlinear correlations (useful for tail events) - Structural scenario approach - scenarios incorporate correlations * Used within each risk category and for market risk volatility

List the problems with Redington's immunization

- Doesn't work well for non-parallel curve shifts - Doesn't work well for large interest rate shifts - Requires regular rebalancing of assets - Often not possible to find assets with: * Long enough durations * High enough convexity

What are the key acronyms used for credit exposure metrics?

- EE: Expected Exposure - PFE: Potential Future Exposure - EPE: Expected Positive Exposure - EEE: Effective Expected Exposure - EEPE: Effective Expected Positive Exposure

Compare insurer vs. bank liquidity risks

- ERM-136 emphasizes that banks are typically more exposed to liquidity risk than insurers - Insurers have the benefit of typically receiving premiums upfront, and then paying out claims later. This inverted cycle of getting a cash inflow first and then cash outflow after is a benefit that makes insurers liquidity-rich - Insurers are long-term investors, and one of their key activities is matching their assets and liabilities - Banks are typically more exposed to immediate and irrevocable liquidity shortfalls (e.g. run on banks) - Banks rely primarily on the wholesale funding market, and are more likely to face systemic liquidity risk - By pooling a large number of risks and by retaining the bulk of the risks underwritten on their balance sheet, potential liquidity issues for insurers are likely to be idiosyncratic without industry-wide impact

Describe the impact of climate risk on actuarial work

- Ensuring model assumptions account for short- and long-term climate risks - Creating/pricing insurance products to adapt to climate change - Designing products that balance all stakeholders' needs - Encouraging climate risk management in investment strategy - Sharing expertise in modeling extreme climate events (cat modeling) - Developing investment strategies/products that address climate risks - Advising governments and policymakers on climate risk - Contributing to public debate on climate-related policies - Disclosing the impact of climate change on an institution's risk profile

Describe Step 2 of GSAM's ALM and SAA process: Asset Universe and Assumptions

- Establish a broad set of asset classes to achieve diversification and portfolio efficiency - Correlations are key * Correlations between assets * Correlations between assets and liabilities

Describe expected exposure (EE)

- Expected exposure is the amount expected to be lost if the counterparty defaults - The expected exposure will be greater than the expected MtM since it concerns only the positive MtM values

Describe the "A Inverse" factor (Also called A')

- Factor A asks "How vulnerable are you now to a certain risk?" - Factor A' asks "how effective is your mitigation?" - Vulnerability and risk mitigation are inversely related (if one is high, the other is low) - Risk assessment questionnaire should ask the four questions for factors A-D but not ask Factor A' since it is already understood through the inverse relationship

State three reasons why direct losses for insurers from pandemics are typically lower than the indirect financial market impact

- Favorable demographics (e.g. if areas with low insurance penetration are most impacted by the pandemic) - Health claims having minimal impact because of government coverage of testing and also because of people delaying elective or non-essential procedures - Offsetting protection risks from business mix (e.g. selling both annuities and life insurance)

Define settlement risk and how to reduce it

- Few transactions are settled on a same-day basis (usually takes 5 days) - Counterparties exposed to the risk of adverse price movements - Settlement risk can be reduced by: 1. Payment netting provisions 2. Exchanging derivatives before they mature (common) - Worsened by foreign exchange risk ("Herstatt risk") * Derivatives are a relatively small percentage of all forex transactions

State possible impacts of cyber risk

- Financial loss - Business disruption - Reputational damage

State the components of the "Three Lines of Defense" model

- First Line of Defense (FLOD) - operating/business units - Second Line of Defense (SLOD) - Individuals or areas responsible for providing assurances of the effectiveness of controls through monitoring the design and performance of those controls - Third Line of Defense (TLOD) - could involve individuals or areas responsible for independent assurance and could include internal audit, external auditors, and regulators

Compare and contrast having an ALM framework's financial objective be based on economic results vs. accounting results

- Focusing on economic reality (based on actual cash flows) ensures the organization can realize higher earnings over the long-term - However, focusing on accounting results can have short-term benefits because these are the results that currently get reported to shareholders, rating agencies, etc.

Describe mortality and morbidity assumption considerations with respect to climate risk

- Food and water insecurity caused by weather changes * May increase morbidity and reduce life expectancy - Temperature changes and volatility * May result in milder winters but more extreme summer weather events * Impact will vary by age - Pandemics and vector-borne infectious diseases * Climate change may increase the prevalence of pandemics * Wider spread of disease-carrying insects may increase malaria and other diseases - Social impacts * Unrest caused by migration and severe measures * Parts of the planet may become less habitable (or even uninhabitable)

Describe potential future exposure (PFE)

- Potential future exposure answers the question: what is the worst exposure we could have at a certain time in the future? - References a certain confidence level α - PFE defines a possible exposure to a given confidence level

State the relationship between the expected value and the variance of the Poisson and Negative Binomial distributions - which is preferred for overdispersed data?

- For a Poisson distribution, mean = variance - For a negative binomial distribution, mean < variance - Negative binomial distribution is used when the data is believed to be overdispersed, meaning that the variance of predictions should not be restricted to be equal to the mean but should be greater than the mean instead

Provide some arguments against focusing on economic results for an ALM's financial objective

- Future long-term economic earnings may not be realized due to forced actions caused by regulatory constraints - Economic results depend on future projections of cashflows that may not be reliable * For example, many economic assumptions are needed to project these long-term cashflows - Long-term interest rates used to discount long-term liability cash flows may not be observable

Identify the key differences between futures and forwards

- Futures are traded on exchanges, while forwards are OTC - Forwards have more counterparty risk * Futures contract is between the parties and the exchange ** Exchange has counterparty risk, but the parties do not ** Exchange reduces risk by pooling contracts and requiring margin accounts *Forward is contracted directly between parties ** Parties post collateral to reduce counterparty risk - Forward contracts are more flexible but more expensive to arrange - Futures are more standardized: cheaper, more liquid

Describe Step 1 of GSAM's ALM and SAA process: Investment Objectives and Constraints

- Gives insight into a portfolio's risk and return attributes under different scenarios - Most important step - Examples of constraints: * Achieve a 5.5% target yield * Asset-liability duration mismatch tolerance: (-0.2, 0.2) * Define exposure limits to specific asset/credit classes

List some best practices of ALM governance

- Have the board of directors and senior management promote ALM - The ALM committee has experienced individuals with the necessary expertise - Roles and responsibilities are clearly defined - Reports are created that clearly communicate the risk profile of the insurance company and supports decision-making - Risk is consolidated at the company level and understood by senior management

Describe Step 4 of GSAM's ALM and SAA process: Risk Measures

- Helps assess effectiveness of different investment strategies - Risk metrics: * Asset-only volatility * Surplus volatility * Economic or required capital, VaR * CVaR (TVaR, CTE) * Surplus drawdown risk

List considerations when determining the percentile used for a risk metric like VaR

- Higher target credit rating requires higher percentile (higher VaR) - EaR and CaR may use the same percentages during normal conditions but could diverage under stress conditions - Returns may change after tail events - Deterministic stress events are critical * Sometimes more meaningful than probability measures

Describe characteristics of companies with strong identification and management capabilities

- Identify and manage the 5 risk categories - Implement ongoing process for identifying/managing significant operational risks - Produce exception reports if metrics are outside maximum tolerances - Business decisions capture hedges/correlations for entire company - Reinsurance based on overall risk tolerance and risk aggregation across lines - Adjust risk profile and risk-management process based on past experience, pro forma model results, future stakeholder expectations, and current market conditions

Describe the following types of dependencies between risk factors: - Immediate Dependencies - Time-Lagged Dependencies - Feedback - Phase-Shift

- Immediate Dependencies - two or more risk factors are causally directly linked, in which case a change in one risk factor causes an immediate change in the other risk factor - Time-Lagged Dependencies - a risk factor is affected by the change in another risk factor after a certain period of time - Feedback - a change in a risk factor causes a change in a second risk factor, which in turn causes a change in the market price of a financial instrument a firm holds - Phase-Shift - a risk factor is affected by the change of another risk factor if and only if the change exceeds a certain threshold

State the benefits of the LVaR approach

- Incorporates the total execution-cost and price-risk components in a consistent fashion - Draws attention to market-impact effects in portfolio liquidation - Illustrates that execution strategies should pay close attention to execution costs and price volatility

State the global benefits of reinsurance

- Increases economic prosperity and growth - Spreads risk and reduces financial uncertainty - Aids in recovery from bad events - Gobal reinsurance diversifies local insurance markets

Describe inherent risk

- Inherent risk is a function of exposure and impact (not taking into account the aspect of vulnerability) - Looks at the potential impact of a risk event with no mitigating factors in place

Describe ways to validate updates to model inputs

- Inspecting model results for reasonableness - Control checks, back-testing, visual inspection, attribution - Automated verification with established controls - Peer review

State the reasons why insurers have less liquidity risk than banks

- Insurers' asset duration < insurers' liability duration - Insurers rely less on short-term funding - Insurers' assets have higher liquidity

List considerations when structuring and coordinating the 3 Lines of Defense

- Intended to be flexible, but it is best when all 3 lines exist in some form - Small companies: harder to separate 1st and 2nd lines - If functions overlap, Board should consider potential impacts - Independence/objectivity of the 3rd line must be maintained - All 3 lines should share information and coordinate activities * Avoid duplication of effort * Use a common risk terminology - Use caution when designing internal audits for the benefit of external parties

What are the two types of nondisclosure?

- Intentional/misrepresentation - Unintentional/incidental

State problems caused by off balance sheet structures, with respect to SPVs

- Investors did not fully understand the risk and illiquidity - Bankruptcy remoteness creates moral hazard for banks - Investors don't scrutinize SPVs because they don't believe the sponsor will abandon them

Describe alternative reinsurance from capital markets

- Key benefit to investors: low correlation between investment and underwriting risk - General structure: 1. Investors provide capital (usually buy purchasing bonds) 2. Investors earn a return on their capital while it is at risk 3. If a covered loss occurs, the insurer can use the investor's capital to pay losses - Sources (investors) of capital supporting alternative reinsurance: 1. Largest source (over half): pension funds, endowments, and sovereign wealth funds 2. Hedge funds and private equity funds (about 25%) 3. Other: reinsurers looking for alternative avenues to obtain returns - Cat bonds have been the most popular, especially for P&C - Other types: sidecards, ILWs, and collateralized reinsurance

List the key SPV risks to sponsoring firms

- Lack of transparency: complexity can make it nearly impossible to manage risk - Risks resulting from a failed or poorly-performing SPV: * Reputational risk (sponsor's credit quality) * Signaling effect (unwanted attention) * Franchise risk (damage to investor relationships) * Liquidity and funding risk (access to capital markets) * Equity risk (if the sponsor holds a large equity tranche in the SPV) * Mark-to-market risk (sale of assets in SPV could depress related assets on the sponsor's balance sheet) - Lax regulation may allow more risk creation in SPVs

Describe how ALM is different between life insurers and P&C insurers

- Life insurers have long-duration liabilities that are primarily focused on interest rate risk - P&C insurers have shorter-term liabilities and are more exposed to catastrophes and mispricing * More focused on maintaining liquidity, given the uncertainty of cash outflows

List general modeling considerations with respect to climate risk

- Limited availability of climate data and expertise - Relationships/correlations among climate and other variables - Short- vs. long-term impacts - Differences in loss experience at the societal level vs. population segment - Varying location, demographic, and socioeconomic profiles

List types of liquidity risk and related risks

- Liquidity risk - unexpected timing/amount of cash needs * Funding liquidity risk (can't meet short-term obligations) * Market liquidity risk (market disruption interferes with selling assets) - Asset/liability matching (cash flow mismatches, especially interest-sensitive business) - Asset-affiliates (affiliate defaults or declines in value) - Capital markets (can't raise capital when needed for any reason) - Catastrophes (short-term liquidity strain from unexpected event)

How can risk appetite help with liquidity management

- Losses during a liquidity event include transaction costs, borrowing costs, impact of fire sale, bankruptcy - Disadvantages of having liquidity risk management without a clear and stated guideline: * Overly-conservative strategy resulting in a low yield on assets * Overly-simplified rules * Underlying risk may not be identified - A liquidity policy without risk appetite might state: * "Cash balance must be > max weekly cash payments in the past 3 months" * "Liquid assets cannot be less than 50% of the total asset balance" - Company should have a quantitative liquidity statement based on a bottom-up analysis of liquidity risks and market impact - Liquidity appetite constrains NB sales, capital management, and SAA

Describe how a more economic, prospective view of capital can supplement the rating process

- Lower BCAR (Best's Capital Adequacy Ratio) could be appropriate based on RM capability - Best is exploring ways to: * Incorporate stochastic modeling into BCAR risk factor development * Tie default probability to required capital for a given rating level - Best will consider using * Internal EC models in the rating process * Output of strong internal EC models for analytical purposes

Identify the recommendations for liquidity management

- Management should establish liquidity risk limits and corrective contingency plans - Manage liquidity at the group level but also consider at legal entity level - Diversify across different asset types and geographies - Large insurance group considerations: * Intra-group transactions can reduce liquidity risk but should be done "at arm's length" * Consider any ring-fencing of legal entities and limits on barriers to asset transfers

Describe cost of capital rate considerations in the context of the IFRS 17 risk adjustment

- May depend on uncertainty in valuation assumptions - Should reflect entity's risk preferences - Should be reviewed/updated regularly - May need to be adjusted for inflation - Should not be "risk-loaded" - don't double count risk already reflected in capital amounts

List the AXXX securitization risks from pricing assumptions and contractual features

- Mortality (A/E results, older ages) - Lapse (dynamic assumptions, secondary market effects) - Policyholder behavior (premium pattern) - Interest rate risk (credited rates) - Non-guaranteed elements - DB changes

Describe the criteria that the IFRS risk adjustment must meet

- Must be an explicit adjustment for uncertainty in timing/amount of future cash flows - Must disclose the associated confidence level (e.g. 90%) - No specific method required/prescribed - Does not reflect financial risk - Does not reflect risks that don't affect insurance contract cash flows * Capital/risk related to growth/expansion plans * Future new business or new products * Investment/portfolio risks * Geographic/legal risks

State ways to limit/mitigate/reduce counterparty risk

- Netting - Collateralization / Margining - Hedging - Utilizing central counterparties - Diversification - Contractual provisions such as close-out features or walkaway/tear-up features - Focusing on trading with high quality counterparties - Monitoring credit lines - Analyzing key credit risk metrics

Describe how social distancing options can affect the projected spread of a virus

- No distancing is riskiest decision: high infection peak will strain hospitals - Early distancing is only slightly better than no distancing * Distancing ends too early while proportion of population that can be infected (Sn )is still high - Delayed distancing may be best: results in shorter peaks * Allows more infections early on * When distancing ends, Sn is less than other options (less "fuel" left) * Better scenario for hospitals - If R0 > 1, the outbreak will continue regardless - Distancing measures are like a "controlled burn"

Describe considerations used to choose between statistical and mechanistic and epidemic projection models

- No simple rule-of-thumb: data availability may drive decision - Statistical models focus more on effects (deaths) than causes * Don't require estimates of R0, CFR, etc. or social structure models * Uncertainty arises because past data may not apply to future or different regions - Mechanistic models focus more on causation (underlying dynamics) * Better at analyzing changing mitigation efforts (e.g. phasing out social distancing) * Can show peaks and valleys in deaths resulting from changing underlying dynamics * Statistical models would require considerable adjustments to do the same thing

List the drawbacks of one-factor models

- Only good for short-term values (up to 1 year) - Not good for multiple points on the yield curve - Does not allow long-term rates to change because of possible future short-term rate changes - Does not work well for long-term liabilities because they are typically influenced by changes in the yield curve shape

How can the operational risk component of other risks be modeled?

- Operational risk is embedded in the other major risk categories - Credit risk with OR >> credit risk alone (think AIG) - Need a way to isolate operational risk * In short-term, combine hard data on other risks with soft OR data * In long-term, identify each risk's contribution to total loss * OR impact = total impact - impact if no operational failure occurred

Describe how different epidemic outbreak scenarios can influence the economic cost of mitigation

- Outbreak models focus mainly on hospitalizations/deaths, not economic impacts - Policymakers may need to marry outbreak models with economic models - Several outbreak scenarios can frame the thinking on economic costs: 1. A fast-spreading, high IFR virus that could wipe out humanity - "Nightmare scenario": any economic cost would be justified 2. A fast-spreading virus with a low known IFR - Suppose R0 = 2, mean generation time = 10, and IFR = 0.1% (similar to flu) - Would be difficult to justify costly interventions 3. A fast-spreading virus with uncertain IFR (between 0.1% and 1.0%) - Much trickier due to uncertainty: could be similar to flu (0.1%) or 10x worse - Must weigh costs/benefits of waiting for vaccines, distancing, and other measures

Describe Step 6 of GSAM's ALM and SAA process: ALM and SAA

- Overall SAA decision reflects an insurer's risk and return objectives, ALM and other constraints - SAA is used to develop customized benchmarks - To recommend a portfolio, must also consider: * Appropriate benchmarks to be constructed in line with SAA * Relevant tolerance for active management to achieve target alpha * How will current market views be incorporated into the portfolio?

State the main types of operational risk controls concerning derivatives

- Oversight of informed and involved senior management - Documentation of policies and procedures - Independent risk management function that reports to senior management - Independent internal audits that verify adherence to the policies and procedures - Appropriate technology and systems for handling confirmations, documentation, payments, and accounting - System of independent checks and balances throughout the transaction process

Describe a close-out provision

- Permits the immediate termination of all contracts between an institution and a defaulted counterparty with the netting of MtM values - Allows an institution to offset the amount it owes a counterparty against the amount it is owed to arrive at a net payment - Close-out netting allows the surviving institution to immediately recognize gains on transactions against losses on other transactions and effectively jump the bankruptcy queue for all but its net exposure - This offers strong protection to the institution at the expense of the defaulted counterparty and its other creditors

Describe considerations for communicating results and updating epidemic projection models

- Present results clearly - Describe data, methods, key assumptions, and any caveats - Don't provide a "most likely" scenario: conveys false precision * Better to provide a range of outcomes (sensitivities) * IHME uses a "cone of uncertainty" similar to hurricane forecasts - Cross-model forecast summaries are helpful (e.g. CDC's webpage) * People will ask "which one is the best?" * Answer: "there is no single correct or best modeling approach" - Models can quickly become stale, and assumptions should be revised * Shifting forecasts may undermine public faith, but it's a consequence of having newer/better data

Compare pro rata reinsurance vs. excess of loss reinsurance

- Pro rata is generally simpler - Excess of loss is generally much cheaper - Pro rata provides surplus relief and help entering or exiting a line of business * Reinsurer pays an upfront ceding commission * Helps offset first-year acquisition costs - Many insurers will enter both in combination - for example: * Automatically cede smaller policies under treaty pro rata agreement * Use facultative to cede larger policies with an excess of loss agreement

What is the purpose of back-testing? List its advantages/disadvantages

- Problem: quantitative measures are subject to model risk, manipulation, and sensitive to assumptions - Possible solution: could backtest against an actual event (earthquakes, 9/11, etc.) - Pros of backtesting: * History repeats (sometimes) * Aids corporate strategic decisions (capital allocation, exit plans, etc.) - Cons of backtesting * The future does not equal the past, especially with so many emerging risks

Describe Step 3 of GSAM's ALM and SAA process: Liability Cash Flow and Replicating Portfolio

- Project liability cash flows to understand the liability profile - Develop the duration profile of the liabilities - Design a risk-minimizing asset portfolio that matches the key economic characteristics of the liabilities

How can hedging be used to reduce counterparty risk?

- Purchasing CDs or CCDS protection - These are instruments that pay off in a default event

List how captive usage affects credit ratings

- RBC ratio less relevant - Less transparent - Reserve funding risk increases - Lower overall capital - Tax efficiency - Increased economic focus

Describe how risk management influences ratings according to AM Best

- RM links balance sheet strength, operating performance, and business profile - Sound RM results in prudent risk-adjusted capital and long-term success - RM should be viewed relative to an insurer's operation scope/complexity - ERM framework must be constantly refined to remain competitive - Insurers will maintain favorable ratings if they: 1. Demonstrate strong RM in core operating processes 2. Effectively execute their business plan - RM should be embedded in an insurer's "Corporate DNA"

Describe how to calculate default probability under real-world vs. risk-neutral approaches

- Real-world / historical approach - determine the actual default probability of the counterparty from historical data - Risk-neutral / market-consistent approach - calculate the market-implied probability

List investment assumption considerations with respect to climate risk

- Reflect physical, transition, and legal/reputation risks - Reflect future shift in investment from carbon-intensive companies to ESG * ESG = environmental, social, and governance

Describe AM Best's traditional approach to assessing exposure to earnings and capital volatility

- Relative risk inherent within the insurer's business profile * Political and regulatory environment * Strategic and operating risk - Earnings and capital trends - Current vs. prior projections relative to actual results * Includes a review of assumptions used in projections

Provide the formula and describe the inputs for quantifying unauthorized affiliate (captive) reserve relief

- Reserve Credit = reserve credit from unauthorized affiliate reinsurers - ModCo Reserve = reserves ceded through unauthorized afilliate modified coinsurance - Total Reserve = Retained Reserve (includes ModCo) + Reserve Credit * Reserve Credit = reserve credit from unauthorized affiliate reinsurers * ModCo Reserve = reserves ceded through unauthorized affiliate modified coinsurance * Total Reserve = Retained Reserve (Includes ModCo) + Reserve Credit

State key considerations for understanding risk appetite for cyber risk

- Risk appetite for cyber risk can be set according to the company's willingness and ability to take cyber risk - Both qualitative and quantitative statements can be formed that describe the cyber risk limits for af irm - Setting the risk appetite for cyber risk requires: * Experience data * Expert opinions * Sophisticated modeling to handle the changing environment

Describe risk appetite, risk tolerance, and risk limit

- Risk appetite: a primarily qualitative document that states an insurer's overall principles with respect to risk taking given its business strategy, financial soundness objectives, and capital resources - Risk tolerance: qualitative and quantitative boundaries that describe an insurer's preference for, or aversion to, particular types of risk in accordance with its risk appetite - Risk limit: quantitative boundaries that express the amount of risk an insurer is willing to take on, beyond which management review may be required * Risk limits are typically more granular than risk tolerances

Describe considerations for setting risk appetite for terrorism risk

- Risks include high mortality, injuries, business interruption, and adverse market movement - Examples: * Don't lose more than 20% of GAAP equity in a terrorism event * Company has contingency plan in place for continuing business operations - Can be mitigated with reinsurance: focus on policyholder concentrations

State the modeling ASOP exposure comments related to the calculations and programming code in the model development process. Summarize key recommendations

- Section 3.2 of ASOP #41 states: "In the actuarial report, the actuary should state the actuarial findings, and identify the methods, procedures, assumptions, and data used by the actaury with sufficient clarity that another actuary qualified in the same practice area could make an objective appraisal of the reasonableness of the actuary's work as presented in the actuarial report" - Consider externally-purchased software from vendors vs. an in-house / homegrown system * External software: request documentation from vendors to support reliance on the purchased software - Recommendation - document calculations and code

State the modeling ASOP exposure comments related to the approval governance and controls of the model development process. Summarize key recommendations

- Section 3.5.2 of the Modeling ASOP exposure draft states: "The actuary should use, or if appropriate, rely on others to use appropriate governance and controls to minimize model risk, to maintain the integrity of the model, and to avoid the introduction or use of unintentional or untested changes" - Recommendation - document governance and controls, such as the following: * Roles and responsiblities - define and document responsibilities for various roles in the governance of the model * Document access controls - processess for limiting access authorization and periodic access review * State change control procedures - processes for authorizing changes to the model, reviewing and testing those changes, and where appropriate, implementing system controls to segregate the testing and production environments * Document model review process - process for reviewing and documenting various validation activities * Model review sign-off - maintain sign-off documentation for model review

State the modeling ASOP exposure comments related to the conceptual framework of the model development process. Summarize key recommendations

- Section 3.6 of the Modeling ASOP exposure draft states: "When the actuary presents the results of the model, the actuary should explain methodology, key assumptions, and parameters" - Also, Section 3.8 of the exposure states: "For model results used in actuarial communications, the actuary should document the nature of the data used, and material assumptions and parameters used in the model" - Recommendation - document the design, theory, and logic underlying the model

State the modeling ASOP exposure comments related to the potential limitations of the model development process. Summarize key recommendations

- Section 3.6.1(a) and (b) of the Modeling ASOP exposure draft states: "In actuarial reports that include information derived from models, the actuary should include explanations, if applicable, of the extent to which a model fails to fulfill its intended purpose, due to limited information, time constraints, or other practical consideration; and... any other known material limitations of the models that have been used and the implications of those limitations" - Recommendation - disclose limitations, and also describe the implications of those limitations

State the modeling ASOP exposure comments related to the intended purpose of the model development process. Summarize key recommendations

- Section 3.6.2 of the Modeling ASOP exposure states: "In actuarial reports that include information derived from models, the actuary should consider including explanations of... the intended purpose of the models and how the user's needs are addressed by those models" - Recommendation - include a statement about any purposes for which the model is not appropriate

Describe considerations for setting risk appetite for concentration risk

- Set appetite for concentrations in products, regions, distribution, asset classes - Example: sell both life and annuity products that ofset mortality/longevity risk

Describe considerations for setting risk appetite for credit risk

- Set appetite for downgrade and default risk (investments and reinsurance) - Map to: obligor limit or counterparty credit rating limit - Example: don't invest in bonds rated below BBB

Describe considerations for setting risk appetite for market risk

- Set appetite for interest rate risk, equity risk, forex risk, alternatives, volatility - Map to: asset allocation limits, forex limits, duration limits, ALM mismatch limits - Example: company will not make bets on forex movement or volatility (no butterfly or straddle investments) - Dynamic modeling is important to understand interaction between policyholder behavior and market risk

Describe considerations for setting risk appetite for liquidity risk

- Set appetite for mass surrenders due to disintermediation, credit downgrade, etc. - Example: maintain enough liquidity to sustain a 1-in-200 year event for 3 months - Less liquidity risk = less investment choice/yield

Describe considerations for setting risk appetite for insurance (underwriting) risk

- Set appetite for mortality, morbidity, lapse, and expense risks - Map to allowed A/E ratio limits - Can be difficult to set limits in practice * May be more useful to monitor trends against A/E ratios * Difficult to define expense limits for new products - Examples: * Set max life mortality A/E ratio = 115% * Don't lose more than 50% of VNB in 1-in-200 year event due to mispricing

Describe the responsibilities of the Board with respect to the Three Lines of Defense Model

- Sets organization's objectives - Implements/oversees the Framework - Oversees implementation of the Model * Ultimately responsible for 1st and 2nd lines * Accountable for selection, development, and evaluation of internal control system - Oversees senior management

Provide some arguments against focusing on accounting results for an ALM's financial objective

- Short-term focus - It can mask interest rate and other financial risks - Changes to accounting rules may give a very different financial picture and risk exposure - It can run counter to the capital objectives in a company

Describe challenges faced by insurers in the financial crisis

- Significant realized losses - Unrealized losses via Accumulated Other Comprehensive Income (AOCI) * Many assets are classified as AFS - must reflect MV on balance sheet * AOCI = MV - Amortized Cost * Flows through equity on the balance sheet * AOCI losses increased sharply between June 2008 and March 2009, then went back to normal levels by September 2009 - Severe drops in market capitalization versus the broader market - Falling RBC ratios created the need to raise capital

State how to compute the solvency ratio. Which component of the solvency ratio formula has tended to be most impacted during a pandemic/crisis? How does this relate to the importance of resiliency measures for insurers?

- Solvency Ratio = Own Funds / Solvency Capital Requirement - Adverse movement of solvency ratio decreases was primarily due to reduction in own funds as opposed to increase in SCR - Management actions during a crisis can only materially reduce SCR but cannot replace lost funds * Own funds take time and effort to generate and retain * Reinforces the message that initial resilience for insurers is more powerful than contingency plans during crises

State common validation report components

- Statement of overall assessment - Identification of the model's intended use, and whether it is fit for that purpose - Documentation of the validation process - Key assumptions, limitations, and opportunities for improvement

Describe SST and RST

- Stress and Scenario Testing (SST) evaluates the strength of the business by subjecting it to varying combinations of economic, liquidity, operational, reputational, regulatory, etc. conditions with differing degrees of severity - Reverse Stress Testing (RST) is intended to identify what event or series of events would cause the insurer to fail (or achieve some predefined adverse outcome)

Describe the tradeoff when choosing the appropriate threshold value u under the GPD model

- Strike a balance between choosing u large enough so that the PBH theorem is applicable from a practical standpoint - But small enough so that a sufficient number of data points can be used in estimating the parameters of the GPD

List reasons why surrendering life insurance is NOT an attractive option

- Surrender charges reduce the amount paid to the policyholder * Often applies in the first years of the policy * Allows the insurer to recover acquisition costs - Inability to obtain new coverage for the same price (or find it at all!) * More significant for policies with mainly biometric coverage * Policyholder's risk profile may have changed - Loss of guaranteed interest rate - more significant when interest rates are down - Loss of future savings benefits (persistency bonuses, GWMBs, etc.) - Tax penalties - loss of tax-sheltered savings, IRS penalties, etc. - Alternatives to surrender - postpone premium, reducing insured amount, etc.

List the reasons why insurance companies use captives

- Tax benefits - Lower regulatory volatility - Improve stat income - Operation in jurisdictions with weaker regulation

List ways that ESGs satisfy regulatory requirements

- Test solvency and liquidity position - Support pricing and reserving - Internationally: ComFrame and Solvency II allow ESGs - US: ORSA recognizes use of ESGs, and NAIC provides ESGs

Describe the Pickands-Balkema-de Haan (PBH) Theorem

- The PBH Theorem states that for a very large family of distributions, for a sufficiently large threshold value u, the distribution of excesses over u can be well-approximated by a generalized Pareto distribution - One of the main pillars/theorems in EVT - Describes the distribution of observations above a high threshold as a generalized Pareto distribution

Historical negligence over counterparty risk can be broken down into which flawed notions that have been widespread amongst institutions trading OTC derivatives?

- The counterparty will never default - The counterparty will never be allowed to default - By the time our counterparty has defaulted, much worse things will already have happened

List different ways risk limits for interest rate risk can be expressed

- The difference between the dollar duration of assets and liabilities must be less than X% of asset value - The net partial duration sensitivity must be less than Y% of the asset value at all yield curve points - The worst case scenario must be less than Z% of asset value

Describe roles and responsibilities in model governance

- The initial development and validation of new models - Documentation of models and processes - Risk assessment of models for classification for determining frequency and level of periodic model validation - Model change management with robust validation and promotion of changes - Updates to production models, and the updating and running of production models to generate results - Peer review of models, adoption of a model governance policy and standards, and oversight and monitoring compliance with standards

Describe the worst credit exposure (WCE)

- The largest (worst) credit exposure at some level of confidence c - Defined as the largest value such that:

State the two key characteristics of lending risk

- The notional amount at risk at any time during the lending period is usually known with a degree of certainty - Only one party takes lending risk

State two aspects that differentiate counterparty risk from traditional credit risk

- The value of a derivatives contract in the future is highly uncertain - Since the value of a derivatives contract can be positive or negative, counterparty risk is typically bilateral

Maes and Dann (2016) and other economists have built a stylized optimization model showing that, under stylized economic assumptions, what relationship exists between the importance of the tools of resiliency and contingency planning? How does this conclusion change when moving from a stylized economic framework to reality?

- Theory: in their theoretical framework, both tools play an equally important role - Reality: from an individual firm perspective, resiliency actions may be underweighted due to behavioral biases or the inability to plan for the unimaginable or unknowable risks

Describe the elements of managing risk exposure in ALM

- This involves using traditional ALM metrics, such as duration and convexity, to set risk limits and rebalance the portfolio when needed - For example, suppose we had a criteria that the percentage difference between the duration of the assets and liabilities cannot exceed 2.5% - If this risk limit is ever breached, then we would rebalance the portfolio by purchasing or selling assets to realign the asset and liability durations

List the characteristics of a good benchmark

- Unambiguous: components and constituents should be well-defined - Investable: should be possible to buy the components of a benchmark and track it - Measurable: must be able to quantify the benchmark on a frequent basis - Appropriate: should be consistent with an investor's style and objectives - Reflective of current investment opinion (positive, negative, and neutral) - Specified in advance: known by all participants before the period begins

Define unknowable risks and state two ways to cope with them

- Unknowable risks - defined as situations "where even the events cannot be identified in advance. Neither events nor probabilities are known" - Ways to cope with unknowable risks: * Limiting leverage * Having enough capital and liquidity to absorb unknowable losses if they should occur

Why are metrics for credit exposure more complex than traditional single-horizon risk measures such as VaR?

- Unlike traditional single-horizon risk measures such as VaR, credit exposure needs to be defined over multiple time horizons to fully understand the impact of the time and specifics of the underlying contracts - Counterparty risk is looked at from both a pricing and risk management perspective, which requires different metrics - In looking at counterparty risk at a portfolio level, it is important to understand the effective exposure with respect to each counterparty

Describe the elements of measuring risk exposure in ALM

- Use metrics such as duration, convexity, and scenario testing to measure sensitivity of assets/liabilities to changes in interest rates, equity levels, etc. - One can also simulate the risk distribution of the assets and liabilities under stochastic simulation, and compute risk measures such as VaR and CTE

Describe characteristics of companies with strong measurement capabilities

- Use of corporate scorecards relative to predetermined tolerances - "What if" scenario testing - Reports to management on risk tolerances and objectives not being met - Well understood, proven EC models (updated/tested frequently) - Risk/return measures and EC models cover various horizon lengths - Reports and models capture correlations across risk categories - EC model identifies where risks provide hedges vs. high concentrations - Ability to determine effectiveness of reinsurance and hedging

State uses of deterministic scenarios

- Useful for sensitivity testing - Aids in developing management strategies and operational plans - Aids in product development - Aids in financial planning - Regulators can specify particular scenarios to be tested by all firms operating under their authority

State ideal characteristics of a risk model

- Uses a statistical distribution that fits historical data well both in the central portion of the data set and in the tail - Should not be constrained by history - Should make use of previous extreme values to offer information on the probability and magnitude of potential values more extreme than those seen previously

Describe issues with VaR

- VaR provides no information about the severity of the tail - VaR may cause people to act as though tail events don't exist * Rephrase potential for loss in tail * Use TVaR - a coherent risk measure (advantage over VaR) - VaR and TVaR assume constant correlations * Risks tend to become more correlated in the tail * Better choice: copulas, which allow correlations to change - Additivity problem: usually can't just add up sub-risks

List the 3 conditions that define a copula

1. C must be increasing in each u - A joint CDF must be non-decreasing 2. C(1, 1, ..., u_i, ..., 1) = u - Ensures each marginal distribution probability is U(0,1) 3. Rectangular inequality: C must increase if all u_i increase

Describe each line in the Three Lines of Defense Model

1. 1st line: front line operating management (risk owners/managers) - Management controls and internal control measures - Responsible for sections 2-5 of the Framework 2. 2nd line: internal monitoring and oversight (monitor, control, compliance) - Varies greatly (more developed/independent for large companies) - Specialty expertise groups: compliance, legal, risk management, security, quality, etc. - Responsibilities: * Assisting 1st line in design of controls and training * Defining how to monitor and measure control success * Escalate critical issues, emerging risks, and outliers * Provide risk management frameworks * Identifying shifts in implicit risk appetite 3. 3rd line: internal audit - Independent assurance to Board and senior management - Does NOT design/implement controls

State the two key causes of a liquidity crisis for insurers

1. Accelerated payment obligations 2. Reduction in available liquid funds

List the broad areas of actuarial work affected by climate change

1. Actuarial modeling - General modeling considerations - Investment assumptions - Mortality and morbidity assumptions - General insurance claims considerations 2. Product management 3. Risk and capital management - ERM frameworks - Capital adequacy - Relationship with the pension fund sponsor 4. Investment management 5. Climate-related risk disclosures

List 3 climate indices published by actuaries

1. Actuaries Climate Index - AAA, CIA, CAS, and SOA 2. Australian Actuaries Climate Index - Actuaries Institute of Australia 3. European Climate Index - Actuarial Association of Europe

Describe drawbacks of the traditional actuarial approach to modeling policyholder behavior

1. Aggregate-level modeling - does not consider socio-demographic, attitudinal, or behavioral factors - Causal factors behind lapse decisions are lost - Removing variation in data hurts the model 2. Rational approach - does not account for social, cognitive, and emotional factors

Reasons why NAIC RBC Model influences states that haven't adopted the Model Act

1. All companies filing an Orange blank (medical) must calculate Health RBC for annual statement 2. Regulators are familiar with the RBC concept and express concerns when TAC/ACL ratio is below 200% 3. Quasi-regulatory agencies like BCBS have embraced Health RBC ratios and may require these levels from companies associated with them

State the two ways to define the price of a financial instrument according to ERM-124-15

1. An expected value of future cash flows, incorporating some adjustment for the risk that is being taken (the risk premium) - The author calls this the actuarial price 2. The cost of an associated hedging strategy - This is the risk-neutral price

List the key variables in the cost of capital technique

1. Appropriate capital amounts (Ct ) 2. Period applicable to the capital (n) 3. CoC rate (rt = risk premium in excess of the risk-free rate) 4. Discount rate/curve (dt ) 5. Probability distribution of the uncertain cash flows Possible interdependencies: - As capital percentile increases, CoC rate decreases - As period increases: * Capital amount increases * Discount rate may get more complex (require yield curve approach)

State the two types of liquidity risk

1. Asset liquidity risk 2. Funding liquidity risk

List the key SPV benefits to sponsoring firms

1. Asset ownership by multiple parties and easier transfer 2. Minimal red tape: cheap and easy to set up (24 hours!) 3. Clarity of documentation: easy to limit certain activities/transactions 4. Freedom of jurisdiction: sponsor can incorporate the SPV in the most attractive region 5. Tax benefits: SPV assets are exempt from certain direct taxes 6. Legal protection: sponsor has limited liability for failed projects financed by the SPV 7. Isolation of financial risk: SPV is a bankruptcy remote "orphan company" 8. Meeting regulatory requirements: easier for sponsor to meet regulatory requirements after moving assets off balance sheet

Define the following biases: 1. Availability heuristic 2. Present bias 3. Optimism bias

1. Availability heuristic - decisions may be influenced by readily recallable (salient) events 2. Present bias - individuals value near-term rewards over more distant rewards (even above what would be implied by rational financial discounting) 3. Optimism bias - tendency to overestimate the probability of positive events and underestimate the probability of negative events

List the steps to secondary guarantee risk hedging

1. Baseline SG cost = average PV of SG claims over 1000 scenarios 2. Convert PV to annual fee (bps of AV) 3. Run market variable sensitivities to see a range of possible hedge payoffs 4. Purchase derivative instruments to mute market sensitivities

Describe and compare 4 examples of equity index models

1. Black-Scholes - assumes lognormal returns with constant σ and dividend yield 2. Heston - generalizes Black-Scholes to use stochastic volatility (more realistic) 3. Stochastic volatility with jumps (SVJ) - generalized Heston with jumps - Jumps are Poisson with a feedback process (volatility affects likelihood of jumps) - Produces very accurate returns but mathematically complex 4. Regime-switching model - Black-Scholes with random changes in parameters - Based on 2-state Markov chain - Advantages: * Significantly fatter tails of stock returns * Can increase downside risk with no upside improvement - Disadvantage: no asset available to hedge the risk in regime switch

List the key steps to set up a risk appetite framework

1. Bottom-up analysis of the company's risk profile 2. Interviews with the Board regarding risk tolerance 3. Alignment of risk appetite with the company's goal and strategy 4. Formalization ofr isk appetite statement with Board's approval 5. Establishment of risk policies, limits, and monitoring consistent with risk appetite 6. Design/implementation of risk-mitigation plan to be consistent with risk appetite 7. Communication with local senior management for their buy-in

Compare buffer and floor crediting structures for hybrid annuities

1. Buffer design - once the buffer is breached, losses are uncapped and the hybrid contract has a negative return 2. Floor design - negative returns immediately lead to losses, but losses are limited to the floor

List reasons to use coinsurance

1. Capital relief, especially with savings-oriented products 2. Signaling effect (reinsurer's blessing) 3. Overcome constraints with reinsurer's help - Different currencies - Different, alternative asset classes - Hedging, derivative strategies - Restrictions from regulators, management, or rating agencies

State the key principles for analyzing risk transfer

1. Ceding company retains less risk 2. Reasonable relationship between risk transferred and capital benefit 3. Plausible chance that reinsurer loss ≥ ceding company's capital benefit - Statutory accounting context

Components of Underwriting Risk (H2)

1. Claims fluctuation risk 2. Other underwriting risk

List 3 ways to create an integrated risk management strategy

1. Clearly define a market risk budget 2. Evaluate economic objectives vs. insurance constraints 3. Determine how and where ALM fits into the overall risk management framework

Provide examples of how actuarial work is exposed to climate change

1. Climatic impacts - Direct; heatwaves, storms, floods, rising sea levels, bush fires, droughts - Indirect: pollution, water/food supply, disease 2. Socioeconomic impacts - Social: migration, healthcare, emergency/social services, consumer behavior - Economic: GDP growth, investor preferences, infrastructure investment, employment, housing, energy, taxation

Describe how the strength of risk management can influence required BCAR levels

1. Companies can maintain BCAR levels near their rating guideline only if they have: - Strong RM capabilities - Low relative exposure to volatility 2. For companies with weak RM: - BCAR requirement is higher even if low relative exposure to volatility - BCAR requirement increases rapidly as volatility exposure increases

State three advantages of ordinary least squares (OLS) regression

1. Conceptually simple 2. Easy to implement 3. Computationally fast

State the five important considerations for the outlook of the hybrid annuity market

1. Consumer appeal 2. Access to new distribution 3. Balancing product profile 4. Offsetting VA guarantee risks 5. Sustainable design in a low interest rate environment

Describe and compare the following processes for generating default-free interest rates: 1. Continuous-time single-factor models 2. 2-factor models 3. Affine models

1. Continuous-time single-factor models - stochastic differential equation for the short rate - Market price of risk = Expected Return / Standard Deviation - Allows pricing of discount bonds and interest rate-contingent claims 2. 2-factor models - more realistic: allow different market price of risk for each factor 3. Affine models - assumes interest rates are driven by stochastic processes - Advantages: * Efficient parameter estimation with MLE * Can be used for RW and RN * Well-defined calibration procedures * Can sustain realistic correlations between yields - Disadvantage: models are linear (may not be realistic)

List the 5 COSO Framework sections and list at least 2 principles under each

1. Control environment - Principle #1: Demonstrates commitment to integrity and ethical values - Principle #2: Exercise oversight responsibility - Principle #3: Establishes structure, authority, and responsibility - Principle #4: Demonstrates commitment to competence - Principle #5: Enforces accountability 2. Risk Assessment - Principle #6: Specifies suitable objectives - Principle #7: Identifies and analyzes risk - Principle #8: Assesses fraud risk - Principle #9: identifies and analyzes significant change 3. Control Activities - Principle #10: Selects and develops control activities - Principle #11: Selects and develops general controls over IT - Principle #12: Deploys through policies and procedures 4. Information and Communication - Principle #13: Uses relevant information - Principle #14: Communicates internally - Principle #15: Communicates externally 5. Monitoring Acitivities - Principle #16: Conducts ongoing and/or separate evaluations - Principle #17: Evaluates and communicates deficiencies

List the components of required liquidity

1. Credit rating downgrade impact (lapses, decline in sales) 2. Normal operational cash flow volatility (model using historical data) 3. Catastrophic risk (model with stress scenarios) 4. Funding commitments for PE or real estate 5. Interest rate risk (disintermediation impact) 6. Adverse mortality and morbidity experience (new markets/products especially) 7. Correlation among the above factors (not all liquidity risks will occur at the same time) - Can lower overall liquidity requirement using correlation matrix - Required liquidity = sum of individual liquidity components - diversification 8. Budget net cash flow = amount of NCF the company expects within the horizon

List and describe the 5 key Risk Areas according to AM Best

1. Credit risk - counterparty exposure for all creditors (agents, reinsurers, bond issuers, etc.) 2. Market risk - liquidity events, asset/liability mismatches, and investment risks 3. Underwriting risk - product development, regulatory, reserves, pricing, mortality, lapses, etc. 4. Operational risk - reputation or franchise value loss due to management change, business interruption, fraud, and many other sources 5. Strategic risk - adverse business decisions or lack of response to industry changes

Identify the 3 broad types of quantitative credit models

1. Credit scoring - translates an entity's features (e.g. accounting ratios) into a credit score - Parametric approaches: * Probit and logit models * Altman's Z-score ( a type of discriminant analysis) - Non-parametric approaches * The k-nearest neighbor approach * Support vector machines 2. Structural models - model the value of the entity rather than using accounting ratios - Examples: Merton and KMV models 3. Reduced form models - use the credit rating derived from some other approach to derive a probability of default

Describe the 3 key areas of ERM according to AM Best

1. Culture - environment throughout an organization that embeds risk awareness and accountability in daily operations 2. Identification and management - ability to consistently identify key risks and establish uniform controls across the entire organization 3. Measurement - use of sophisticated tools to quantify and report on risks, general economic conditions, and key events on a regular basis

Provide the policyholders' hierarchy of liquidity

1. Current bank deposits (e.g. ATM) - no significant costs incurred 2. Short-term savings account - usually no cost unless it is a large sum 3. Overdraft, credit cards - costly and often available only for short periods of time 4. Sale of securities - risk of transaction costs, capital gains/losses, and taxes 5. Family/friends - common under economic hardship, but creates family dependence 6. Increase mortgage balance - increases bankruptcy risk and interest costs 7. Policy loans - easy but does not stop future premiums, and loan grows with interest 8. Stop premiums - does not provide cash, may lose insurance 9. Surrender life insurance - surrender charges; lose coverage and tax benefits; delayed payout 10. Sell the house - cost of renting, may lose significant portion of sale to taxes

Describe the general validation framework for the IFRS 17 risk adjustment

1. Data - raw and transformed 2. Assumptions - requires a control framework - Derived consistently year-to-year - Credible for the purpose - Subject to governance (peer review, sign-offs) 3. Process - activity steps at each level of risk adjustment calculation - Documentation - procedures, dictionaries - Process management - organization-wide requirements for process management - Controls - for levels of identified risk (IT security, etc.) - Audit trail - for regulators and auditors 4. Computation model - mechanics/features of contracts and assumptions - Stochastic models: expected values, impact of simplifications - Projected trends vs. trends in historical data - Reasonableness of projected cash flows - Explanation of changes in results from period to period 5. Results - reported and intermediate

State the 7 step data quality process

1. Data extraction 2. Data profiling 3. Generalized cleansing and de-duping 4. Data standardization 5. Enrichment 6. Approvals 7. Quality monitoring

Describe the three risk factors of credit risk

1. Default risk - The risk of default by the counterparty - Measured with the probability of default (PD) 2. Credit exposure risk - The risk of fluctuations in the market value of the claim on the counterparty - At default, this is known as exposure at default (EAD) 3. Recovery risk - The uncertainty in the fraction of the claim recovered after default - Recovery Rate = 1 - Loss GIven Default (LGD)

State and describe the four kinds of mortality risk factors

1. Demographic risk factors - Includes age, gender, ethnicity, and chromosones 2. Socio-economic risk factors - Includes income, education, occupation, and environment - Studies show a positive correlation between healthcare spending per capital and life expectancy. Typically once healthcare spending exceeds some level (e.g. $3,000 per capita per year), then life expectancy gains from additional spending are minimal 3. Behavioral risk factors - Includes tobacco/drug use, alcohol consumption, eating habits, and physical (in)activity 4. Biomedical risk factors - Includes blood pressure, blood glucose, body mass index (BMI), cholesterol, and impaired kidney function

State the three main types of internal controls

1. Detective - designed to detect errors or irregularities that may have occurred 2. Corrective - designed to correct errors or irregularities that have been detected 3. Preventive - designed to keep errors or irregularities from occurring within the business process

List the 3 steps of the RMS model of mortality improvement drivers

1. Develop a model of future mortality risk drivers (vitagions) 2. Understand the correlations of the risk drivers 3. Understand impacts on population sub-segments

Describe challenges modelers face early in a pandemic

1. Differences within the human population: - Individual differences in health status, immune response, and social behavior - Regional differences: * Population density * Age distribution * Household structure * Mobility patterns * Weather, pollutants * Heathcare services 2. Biased and limited data - Skewed toward severe cases - Data will be lacking on people with mild or no symptoms - Data samples are often too small, which creates uncertainty 3. Regional differences in case reporting - Some regions require lab tests; others rely on a doctor's judgement - Reliability of lab tests (false positives/negatives) - Cause of death attribution: virus vs. other conditions

State the factors that reduce the effectiveness of the mortality-longevity offset

1. Different ages - life insureds are usually younger (45-55 vs. 65-75 for SPIAs) - More typical correlation: -82% (still significant!) - More typical optimal hedge ratio = 10:1 (need less SPIA relative to term) 2. Different countries - many (re)insurers have business in different countries - Correlations break down due to incidence of smoking, age distribution, etc. - Typical US term life vs. UK SPIA correlation = -58% (also still significant!) 3. Different products - different products have different cashflow characteristics - Term/SPIA offset is different from a WL/group annuity offset

List the three steps for spreadsheet data quality management

1. Discovery 2. Triage 3. Control

Define level risk with the risk rating method

1. Divide the population as a whole into a number of homogeneous groups - Choose time period (longer = more data, but too long hides trends) 2. Model mortality of each group as a function of risk factors - GLM approaches: probit, logit - Postcode rating approach 3. Analyze the structure of the group of lives of interest (e.g. a portfolio of annuitants ) 4. Use risk factor exposures to infer the underlying mortality of the group Drawback: individuals may not conform to their risk factor stereotypes

List ways to identify and remove legal and regulatory uncertainties for derivatives

1. Documentation required to create legally enforceable agreements (statute of frauds) 2. Capacity of parties to enter into transactions (ultra vires) 3. Enforceability of bilateral close-out netting and collateral arrangements in bankruptcy 4. Enforceability of multi-branch netting arrangements in bankruptcy 5. Legality/enforceability of derivatives transactions

State the 8 components of ERM

1. Internal environment 2. Objective setting 3. Event identification 4. Risk assessment 5. Risk response 6. Control activities 7. Information and communication 8. Monitoring

State the five questions to ask when developing a series of risk-based controls

1. Does the control consider a failure that may rise to the level of a material weakness? 2. Can the control be relied upon to either prevent or detect (in a timely manner) a material misstatement of the filed financial statements? 3. Has the control been updated recently to reflect the current business process? 4. Has your organization considered remediation actions resulting from a fraudulent activity, findings from external and internal audits, and other control self-assessment processes? 5. Is the control a key component of your continuous controls monitoring (CCM) initiative?

State the practical reasons for using Pearson correlation

1. Easy to calculate 2. Easy to communicate (many are familiar with it) 3. Determines dependence structure of normal (elliptical) variables 4. Straightforward to construct a symmetric, PSD correlation matrix

List the features of a good risk response

1. Economical (benefits > costs) 2. Matches risk as closely as possible (balance with #1) 3. Simple as possible (higher complexity = higher chance of mistakes) 4. Active, not just informative (better to avoid breaches) 5. Retain unless risk is significant

What are the 3 levels of establishing a risk appetite framework?

1. Enterprise risk tolerance - Define organization's willingness/ability to take risk - Based on input from Board and senior executives - Unique to each company 2. Risk appetite for each risk category - Allocate enterprise risk tolerance to specific categories/activities - Look for high risk-adjusted returns and competitive advantages 3. Risk limits - Most granular level used for business operation - Translates risk tolerance into risk-monitoring measures

Describe 2 ways to estimate copula parameters

1. Estimate Kendall's tau from a data set - Set t = f(θ) and solve for θ 2. Maximum likelihood estimation - MLEs = parameters that maximize the joint PDF loglikelihood function

State the three uses of models

1. Explain a system 2. Study the effects of different parts of a system 3. Derive estimates and guide decisions

List 2 methods that are not likely to meet IFRS 17's criteria for a risk adjustment method

1. Explicit risk margins (e.g. 5%) added to individual assumptions - Difficult to translate to a confidence level - Confidence level is likely to vary over time 2. Adjustments to discount rates - Discount rates usually have little relevance to insurance cash flows - There is no accepted method for adjusting the discount rate

State four ways ORSA reports can add significant value to the regulatory process

1. Facilitates insurer-regulator dialogue 2. Contains information regarding insurer strategies, risks, controls, and results that can be used to obtain a general understanding of the whole organization 3. Contains a summary of all the enterprise risks 4. Provides reasonably consistent information from all insurers in a similar timeframe

State the four relevant risk questions for factors A-D

1. Factor A: how vulnerable are you to a certain risk? 2. Factor B: how exposed are you (here and now)? 3. Factor C: how much effort do you put in mitigation? 4. Factor D: if the risk occurs and mitigation fails, what will be the impact?

Provide a list of federal involvement in reinsurance (before Dodd-Frank)

1. Federal Crop Insurance Corporation (FCIC) 2. Price-Anderson Nuclear Industries Indemnification Act 3. Overseas Private Investment Corporation (OPIC) - political risk insurance for overseas US businesses 4. Aviation War Risk Program: allows the FAA to supply aircraft in times of war 5. National Flood Insurance Program (NFIP) 6. Terrorism Risk Insurance Act (TRIA): established the Terrorism Risk Insurance Program (TRIP) - Backstops commercial P&C losses resulting from certified acts of terrorism 7. Treasury's Surety Bond Branch: certifies insurers/reinsurers ("sureties") to directly write and reinsure federal surety bonds - Surety bond - contract where one party (the surety) guarantees the performance of obligations of a 2nd party (the principal) to the 3rd party (the obligee)

Describe critical principles implicit in the Model

1. First line: business/process owners - Manage highest impact risks day-to-day - Primarily front- and mid-line managers 2. Second line: supports management with expertise, process excellence, and management monitoring - Separate from first line but still under senior management - Management/oversight function that owns many aspects of the management of risk 3. Third line: provides independent assurance to senior management and Board - Has a primary reporting line to the Board - Evaluates first and second lines' efforts vs. expectations - NOT a management function (separates it from the second line!)

Describe whether the posterior distribution gives larger weight to the company data or the prior for: 1. A large prior sample size N 2. A small prior sample size N

1. For large N, the posterior distribution will give larger weight to the prior 2. For small N, the posterior distribution will give larger weight to the company data

Describe whether the posterior distribution gives larger weight to the company data or the prior for: 1. Large σ 2. Small σ

1. For large σ, the posterior distribution will give larger weight to the company data 2. For small σ, the posterior distribution will give larger weight to the prior

State the three types of initial events

1. Global events 2. Regional events 3. Company-specific events

Describe the different channels where resiliency adds value for firms

1. Helps avoid downside of default: proper risk management can help reduce the chance of bankruptcy. Bankruptcy has many costs (e.g. legal, frictional, brand/reputation) and the chance of facing these costs can be lowered with resiliency 2. Opens up opportunities: a strong position during a downturn allows the management of the resilient firm to take advantage of mispriced assets or competitors' weaknesses 3. Contingent management actions during a crisis have a much higher cost compared to resilience: executing contingency plans during a crisis often comes at a high cost and typically cannot fully recover to pre-crisis conditions 4. Contingency plans cannot make up for lost value: resiliency can help proactively manage risk and reduce the chance that value is lost in the first place

State the types of operational risks

1. High frequency / low severity - Examples: banks can make many small mistakes regularly when doing millions of transactions per day - Could use a non-life reserve method 2. Low frequency / high severity - Can result from poor project management, fraud, etc. - EVT can be useful to model skewed, fat tails - Important to model links between operational risks and with other risks (ex. fraud more likely in a down economy)

List the business goals helped by financial reinsurance

1. Improving timing and/or level of earnings 2. Improving solvency margins or regulatory capital 3. Improving IRR, ROE, or capital leverage 4. Supporting growth 5. Unlocking capital 6. Finance new business, joint ventures, and acquisitions 7. Replacing valuation conservatism with economic reality

Why is policyholder behavior getting more attention in the insurance industry?

1. Increased product flexibility 2. More investment components 3. Volatility in financial markets 4. Sophisticated financial reporting and regulation 5. Social and technological advances 6. Emerging behavioral economics research

State the 3 ORSA sections

1. Insurers Risk Management Framework 2. Insurers Assessment of Risk Exposure 3. Group Assessment of Risk Capital and Prospective Solvency Assessment

Summarize insurer crisis risk management "lessons learned" from COVID-19

1. Insurers should build resilience upfront (in the form of sufficient capital and liquidity resources) combined with limiting risk accumulations 2. Scenario analysis is important for both setting and assessing risk appetite in a dynamically-changing environment - Scenario analysis can be improved by testing exposure to contract wording challenges and further considering how public perception may change when wordings are challenged 3. In regards to the insurability of pandemic risk, the industry has learned that business interruption is not an insurable risk 4. Insurers need to improve policy wordings to avoid ambiguity and increase transparency/enforceability 5. Consider the impact of large, systemic risk accumulations in other areas besides pandemic risk (e.g. cyber risk)

List the benefits of economic capital analysis

1. Insurers: need capital to take risk and absorb experience fluctuations 2. Management: needs to understand capital requirement for the business - Should be able to justify capital to rating agencies - Want credit for diversification, credit management, etc. 3. Shareholders: want an adequate return, which is the insurer's cost of capital - Want to balance between too little and too much capital - Need to understand if company is adequately capitalized or over-capitalized 4. Regulators: want solvency - Want to understand on a realistic basis how well-capitalized companies are

List different sources of risk for life insurers that are typically within the scope of an ALM program

1. Interest rate risk 2. Liquidity risk 3. Credit risk 4. Currency risk 5. Market risk - Associated with losses in market value of non-fixed income (NFI) assets, such as equity or real estate

State the responses to the following operational risks: - Internal and external fraud - Employment practices and workplace safety - Clients, products, and business practices

1. Internal and external fraud - Scope is huge: from petty theft to rogue traders - Mitigation should reflect the size of the risk 2. Employment practices and workplace safety - Awareness of regulatory and legal obligation - Culture (e.g. avoid lack of diversity on company's Board) - Check employees' desks, seats, etc. to make sure they are safe and appropriate 3. Clients, products, and business practices - Main goal: ensure clients are not mis-sold products - Encourage a "know your client" mindset - Click through processes to verify client's knowledge - Discourage selling inappropriate products with culture, bonus deferrals, and clawbacks

Provide a list of operational risks

1. Internal and external fraud 2. Employment practices and workplace safety 3. Clients, products, and business practices 4. Damage to physical assets 5. Business disruption and system failure 6. Execution, delivery, and process management 7. Crime risk 8. Technology risk 9. Cyber risk 10. Regulatory risk 11. People risk 12. Legal risk 13. Model risk 14. Data risk 15. Reputational risk 16. Project risk 17. Strategic risk

List the different types of surplus relief arrangements and their pros/cons

1. Internal capital - Pros: good for large companies with lots of free surplus - Cons: too risky or not feasible for small companies 2. YRT reinsurance - Pros: readily available, lowers mortality and counterparty risk, rating agencies favor - Cons: no lapse risk coverage, less surplus relief than other reinsurance 3. Offshore captive reinsurance - Pros: lower reserve/capital requirements, tax benefits - Cons: regulators limit amount ceded, requires LOC 4. Letters of credit (LOC) - Pros: readily available for short-terms - Cons: mismatch risk, pricing risk, getting expensive 5. Bank lending - Pros: useful as a temporary or long-term solution - Cons: requires bank partnership 6. AXX Redundant Reserve Securitizations - Pros: more surplus relief than YRT reinsurance, avoids LOCs, tax benefits (IRR increases) - Cons: must define redundant reserve level, complex, hard to explain to investors

List a few ways insurers can increase the return on their asset portfolio during a low interest rate environment

1. Investing in bonds with lower credit quality 2. Increasing allocation to riskier asset classes such as equities, real estate, etc. 3. Increasing yield-to-maturity in an upward-sloping term structure - Sell shorter duration assets with lower yields, and buy longer duration assets with higher yields 4. Transfer risk to new policyholders by replacing sales of traditional guaranteed products with unit-linked products 5. Hedging (i.e. swaps, floors)

List the 6 high level steps in GSAM's ALM and SAA process

1. Investment Objectives and Constraints 2. Asset Universe and Assumptions 3. Liability Cash Flow and Replicating Portfolio 4. Risk Measures 5. Risk/Return Trade-Offs 6. ALM and SAA

Discuss recent regulatory changes to SPVs (circa 2011)

1. Key accounting change: it's no longer "automatic" that an SPV = OBSV - IFRS: holds SPVs on balance sheet if SPV is "controlled" by the main entity: 1. Sponsor benefits from undertaking activities on behalf of the SPV 2. Sponsor effectively controls the SPV 3. Sponsor has the majority of the risks of the SPV 4. Sponsor receives the majority of the benefits of the SPV - US GAAP: depends on principles around control of SPV's underlying assets 2. Regulatory changes for SPVs and OBSVs: - Tightening of covenants in lending documentation - Firmer legal risk management practices by banks and regulators - Increased emphasis on counterparty risk in capital market structures - Increase in pre-packs, debt-for-equity conversion, and disputes over valuation 3. Basel III: a new global regulatory standard on bank capital adequacy and liquidity - Pillar I: requires banks to hold higher capital (also higher quality and more liquid) - Pillar II: enhanced supervisory review for risk management and capital planning - Pillar III: enhanced risk disclosure and market discipline

Why is modeling operational risk severity difficult?

1. Lack of sufficient data 2. Poor data quality 3. The existence of truncated or censored data 4. Sensitivity to low probability/high severity loss events 5. Classification issues 6. The need to incorporate both internal and external data

Describe 5 challenges in understanding policyholder behavior

1. Lack of time - Time is spent preparing data, not analyzing it 2. Information asymmetry - Insurers have incomplete information about policyholder - Policyholders have incomplete understanding of products 3. Curse of knowledge - Significant gap between policyholder and actuarial product views 4. Data challenges - Lack of data, credibility, and causal linkages - Systems weren't designed with policyholder behavior in mind - Data not in required format 5. Human decision-making is not fully understood (but improving) - Behavioral economics, big data, computing power, dynamic modeling

Describe the two major potential impediments to the insurability of pandemic risks

1. Large potential loss accumulation and minimal diversification - Epidemic and pandemic risks are characterized by accumulations potentially large enough to raise the question of insurability - High correlations / low diversification associated with pandemic risk 2. Issue of potentially external moral hazard caused by distorted incentives (e.g. for public policy decisions) - Losses from pandemics occur over time, which means governmental and personal decisions may be taken to contain or suppress the pandemic

List and descrtibe the 5 drivers ("Vitagions") of mortality improvement

1. Lifestyle - behaviors that influence life expectancy (smoking, diet, exercise) 2. Health environment - sanitation, pollution, awareness of healthcare issues 3. Medical intervention - current/future medical treatments (statins, aspirin) 4. Regenerative medicine - techniques for repairing and renewing cells and organs (stem-cell therapy, nanomedicine) 5. Anti-aging processes - therapies that involve changing the "biological clock" of cells

State the steps for multivariate credit migration models

1. Link migration probabilities to changes in firm asset values using Merton - Change in rating = function of change in assets and asset volatility 2. Calculate correlations between assets of different firms 3. Model a range of company-specific indices 4. Simulate indices and independent firm factors 5. Calculate change in individual bond values, then sum to get total 6. Calculate risk measures (e.g. VaR)

List the recommended disclosures for derivatives

1. Management's attitude to financial risks, how instruments are used, and how risks are monitored and controlled 2. Accounting policies 3. Analysis of positions at the balance sheet date 4. Analysis of the credit risk inherent in those positions 5. Dealers only: additional information about the extent of their activities in financial instruments

Describe the genereal differences in use of market-consistent vs. real-world ESGs

1. Market-consistent valuation (risk-neutral) - Used to price derivatives and embedded options in insurance contracts - Mostly concerned with mathematical relationships among financial instruments - Less concerned with forward-looking expectations - Used to value securities that are not traded in markets or lack a closed-form solution * e.g. VAs, which are subject to policyholder behavior 2. Real-world risk management - Used to assess business risk, regulatory capital, and rating agency requirements - More concerned with forward-looking paths of economic variables

Describe three methods for measuring the interest rate risk of NFI assets backing liabilities

1. Model real estate and equities as bonds with a fixed equity risk premium when calculating their duration 2. Perform sensitivity analysis for various deterministic interest rate and equity return scenarios 3. Use a stochastic model that generates economic scenarios, and then obtain a probability distribution of surplus

Describe the following trading strategies: 1. Modeled trading strategies 2. Self-financing trading strategies

1. Modeled trading strategies - establish a mechanism of buying and selling securities 2. Self-financing trading strategies - do not require the injection of additional capital during their execution

List three different bases companies can use for measuring risk exposure

1. Protect economic surplus by minimizing PV(AssetCFs) - PV(LiabilityCFs) 2. Protect market value by minimizing MV(Assets) - MV(Liabilities) 3. Protect accounting results by minimizing BV(Assets) - MV(Liabilities)

List the recommended derivative accounting practices

1. MtM each period, with changes in market value recorded as income 2. Account for derivatives consistently with the risks hedged 3. Any derivatives held for purposes other than risk management should be MtM 4. Amounts due to/from counterparties should only be offset when there is a legal right to offset or when enforceable netting arrangements are in place

List the weaknesses in the formulaic approaches of economic capital recognized by regulators

1. No link between required capital and the effectiveness of risk management/mitigation 2. Doesn't cover all risks 3. Not transparent 4. Doesn't adjust to changes in the financial environment 5. No diversification credit

Identify the reasons why the direct approach is not practical in the context of hedging

1. No liquid market for the put options needed - ILA guarantee duration >> OTC maturities (5 years max, usually) 2. Option dealer may not want to sell deep ITM options - Rollup benefit may be 70% or more ITM 3. Hedge basis = uncommon index or mixture of hedge bases 4. Options may be too expensive due to liquidity mismatches - Financial derivatives may be settled daily, while ILA guarantees could take decades 5. Demand/supply effects on options that are unrelated to the hedged liability - Implied volatilities in options can increase

List the properties of a coherent allocation method

1. No undercut - a sub-portfolio's allocation should be no more than its standalone capital requirement 2. Symmetry - if the risk of 2 sub-portfolios is the same, then their allocation should be the same 3. Risk-free allocation - capital allocated to a risk-free line of business should be zero

Describe key items of the Dodd-Frank Act of 2010

1. Non-admitted and Reinsurance Reform Act (NRRA): enhances uniformity in state solvency regulation - Designates reinsurer's domicile state regulator as the sole state regulator - Prohibits other states from denying credit for reinsurance if an insurer's domicile state has granted it 2. Made the FIO Director a member of FSOC - Authorized to recommend "non-bank financial company" to be subject to Fed supervision and enhanced prudential standards 3. Authorized FIO to develop federal policy related to international insurance - FIO represents the US in the IAIS * Director of FIO serves as a member of the IAIS Executive Committee * IAIS initiatives: ComFrame, global insurance capital standards, standards for G-SIIs - FIO Director is on the Steering Committee of the EU-US Insurance Project - FIO's Modernization Report recommends: more oversight of captives and covered agreement with EU

State the five key metrics to consider when measuring results for internal based controls

1. Number of incidents per period 2. Average value of incidences identified per period 3. Estimate of total value of incidences identified per period 4. Average hourly rate of person remediating incidents per period 5. Percentage of transactions tested per period

State three drivers of spreads

1. Order-processing costs 2. Asymmetric-information costs 3. Inventory carrying costs

What are the two broad classes of financial products with counterparty risk? Which class is usually more significant in nature and why is this the case?

1. Over-the-counter derivatives - Interest rate swaps - Forex forwards - Credit default swaps 2. Securities financing transactions - Repos and reverse repos - Securities borrowing and lending In general, credit exposure for OTC derivatives contracts is more significant due to the: - Size of the market - Diversity of the instruments - Bilateral nature of derivatives transactions - Large possible fluctuations in the magnitude and sign of value

State the steps/mathematical foundation of hedging

1. Perform dynamic hedging using Monte Carlo - Alternative "direct approach": use options, but not as practical 2. Use fund mapping to translate underlying funds into a combination of hedgeable indices - Ensures the hedge has a good fit to the item being hedged 3. Rebalance hedge instruments to match Greeks

Describe the TCFD definition of climate-related risks

1. Physical risks - event-driven (acute) or long-term (chronic) shifts in climate - Adverse impact on asset values or investment returns - Mortality/morbidity impact of increasing heatwaves - Property damage/liabilities resulting from windstorms, floods, droughts, etc. - Disruption in health/social services, forced migration, infrastructure damage 2. Transition risks - result from transitioning to lower carbon economy - Policy risk - actions that either constrain/promote climate adaptation - Technology risk - Market risk - shifts in supply/demand for commodities, products, and services 3. Legal and reputation risks - Legal risk - climate change-related litigation and liability insurance * Failure of organizations to mitigate the impacts of climate change * Failure to adapt to climate change * Lack of disclosure around material financial risks - Reputation risk * Changing public perceptions on climate impact/policies * Actuaries who fail to advise appropriately

Define methods for managing credit risk, in the context of derivatives

1. Policies and procedures - Internal controls used before entering transactions - Documentation of credit risk mitigation provisions - Credit enhancement structures 2. Counterparty credit evaluation 3. Risk limits for counterparties 4. Collateral (require more for higher risk counterparties)

List the 4 primary responses to market and economic risk

1. Policies, procedures, and limits - Must have clear policies - Linked to diversification, counterparty risk, and operational risk 2. Diversification - hold a range of investments to limit exposure to poor investments - Measure with benchmarks and/or factor analysis 3. Investment strategy - easiest approach but varies by type of firm - Pensions: high market risk (very important) - Insurers: manage market risk within regulatory constraints - Banks: low market risk 4. Hedge with derivatives and options - Futures, forwards, puts, etc.

State the responses to demographic (life) and non-life risk

1. Premium rating - underwriting is key for demographic and non-life - Underwriting benefits should exceed costs - Varies by product: life, annuities, groups of lives - Non-life only: can use policyholder experience to rate future premiums 2. Risk transfer - reinsurance is the fundamental way - Proportional: good for high frequency claims (allows more NB and diversification) - Excess-of-loss is good for low frequency / high severity - Alternatives to reinsurance: securitization (e.g. cat bonds) - Pensions can use annuity swaps (buyout vs. buy-in) 3. Risk modification - more relevant to non-life - Alarms, cameras, promoting behavioral changes 4. Diversification of business mix - Geographic and/or risk diversification (e.g. mix life and longevity risks) - Avoid concentrations

Summarize the cash flow effects under pure coinsurance, coinsurance with funds withheld, modified coinsurance, and blended coinsurance

1. Pure coinsurance - Largest initial cash outlay for ceding company (= Initial Consideration - EA) 2. Coinsurance with funds withheld (Co FW) - Ceding company retains assets, sets up FW liability * Greatly reduces upfront cash flow - Ceding company pays reinsurer investment income on FW - Reinsurer pays ceding company change in FW 3. Modified coinsurance (ModCo) - Ceding company retains assets and reserves - Reinsurer pays ModCo Adj (= Change in ModCo Res - ModCo InvInc) - Exact same cash flow profile as Co FW 4. Co/ModCo - Changing mix = changing net cash flow - Can reduce ModCo % to further minimize cash flow - If ModCo Adj = Initial Consideration - EA, results in least overall net cash flow

Describe and compare the following processes for generating default-free interest rates: 1. Quadratic models 2. Nelson-Siegel

1. Quadratic models - assumes short rate is a quadratic function of state variables - Advantages: * Can model non-constant volatility and negative correlation * Captures nonlinearity - Disadvantages: * Difficult to fit targets of yields and returns * Estimation is difficult and time-consuming 2. Nelson-Siegel - stochastic model of level, slope, and curvature of term structure - Widely used for forecasting interest rates - May do better in linking the term strcture to macroeconomic factors - Dynamics of the stochastic process need to be understood when using

Describe existing industry views on risk appetite

1. Rating agency view (S&P) - Favorable ERM rating requires clean/formal linkage between risk appetite and strategic planning - During the financial crisis, some insurers' risk limits did not reflect their risk appetite 2. Solvency II view - Risk management should be integrated into organizational structure - SCR set to survive 1-in-200 year event (VaR99.5) 3. CEIOPS view - Clearly defined and well-documented risk management strategy - Adequate written policies 4. Current insurance company views - Most agree risk appetite is important in strategic planning - Companies seek to protect and create value from risk appetite

State the 4 general responses to risk

1. Reduce likelihood and/or severity of risk - Diversification, hedging, robust systems/processes 2. Remove the risk completely - Example: use exchange-traded options instead of OTC 3. Transfer the risk to another party - Non-capital markets: insurance, reinsurance, captives - Capital markets: securitization * Puts market price on risk * Faster than other ways of raising capital 4. Accept the risk - Some risks are too trivial or not worth the cost - Business plan may call for retention or self-insurance

List at least 5 essential features of a comprehensive ESG

1. Reflects relevant view of the economy 2. Includes extreme but plausible results 3. Reflects realistic market dynamics and relevant historical facts 4. Balances practicality and completeness 5. Consistent across RW and RN modes 6. Meets regulator and auditor requirements 7. Sufficient detail for extensive validation 8. Accommodates many types of calibration 9. Coputationally efficient and numerically stable

Briefly describe 3 parameter estimation methods

1. Regression analysis - Used to estimate the parameters of a dependency relationship - ERM-104 focuses on linear approaches 2. Maximum likelihood estimation (MLE) - Commonly used in estimating frequency and severity distributions - Resulting parameters are called maximum likelihood estimators (MLEs) 3. Model-free methods - Commonly used by actuaries (e.g. claim development factors)

State and briefly describe the 7 types of scenarios discussed in ERM-120

1. Reverse Scenarios - identify a scenario that is expected to give rise to a particular amount of financial loss 2. Historical Scenarios - based on experience during an observation period 3. Synthetic Scenarios - describe hypothetical conditions that have not been observed 4. Company-Specific Scenarios - specify events that are tailored to the specific mix of risk exposures of a firm or of a particular portfolio within the firm 5. Single-Event Scenarios - describe the effect of a single event 6. Multi-Event Scenarios - consider that multiple factors contribute to any future scenario 7. Global Scenarios - cover the effects of insurers and other financial institutions on a global level

List the 6 core ERM processes

1. Risk governance 2. Risk and capital measurement 3. Risk budgeting 4. Liquidity risk management 5. ALM and SAA 6. Risk reporting

State the three drivers of antiselection

1. Risk histories elicited online rather than with teleinterviews 2. Applicant access to medical records 3. Selecting one specific technician to do all of the paramedicals (i.e. tests)

State the four primary benefits of segregation of duties (SoD)

1. Risk of a deliberate fraud is mitigated (as the collusion of two or more persons would be required in order to circumvent controls) 2. Risk of legitimate errors is mitigated as the likelihood of error detection is increased 3. Cost of corrective actions are mitigated as the likelihood of early detections is increased 4. Organization's reputation for integrity and quality is enhanced through a system of checks and balances

Describe how to use the output of the RMS model of mortality improvement drivers

1. Run many simulations of future mortality using the correlations 2. Calculate PV of product cash flows under each scenario 3. Plot two products/ PV of cash flows to see the liability correlation relationship

List the general approaches to measuring risks

1. Scenario-based model - can be deterministic or stochastic - Specific scenarios can cover multiple risk drivers - Scenarios should reflect risk correlations 2. Static-factor model - linear combination of risk factors (e.g. US RBC) 3. Stochastic-factor model - calculate VaR / CTE from aggregate loss distribution 4. Covariance model - Evaluate individual risks with stochastic factor model - Aggregate risks assuming a multivariate normal distribution

List the features and uses of SPVs

1. Securitization of loans - Investors purchase bonds in tranches - Bond proceeds finance SPV assets 2. Asset transfer - especially assets that would be difficult to transfer otherwise 3. Financing a parent company without investing directly in the parent - Frequently used to finance large infrastructure projects 4. Relocation of risk from sponsor to SPV - SPV is "bankruptcy remote" 5. Financial engineering - manipulate capital ratios, manage regulatory requirements, hide losses, etc. 6. Raising capital at more favorable borrowing rates

State the three critical corporate internal controls

1. Segregation of Duties (SoD) 2. System access 3. Delegation of Authority (DoA)

Describe how to calculate the 3-month required liquidity for operational cash flow volatility

1. Simulate weekly net cash flow using historical results 2. Determine the mean and 99.5th percentile of the NCF distribution 3. Extra liquidity needed for 1-in-200 year weekly event = 99.5th percentile - mean 4. Extra liquidity needed for 1-in-200 year 3-month event:

List the types of pure captives

1. Single parent owner/insured (most basic) 2. Industrial insured group captives - single and multi-industry 3. Risk retention groups (RRGs) - type of captive created by LRRA 4. Association captives - formed by an association for the benefit of its members

State the three major types of factors that are driving the slowdown of mortality improvement

1. Some major causes of death (e.g. motor vehicle deaths) had lower average annual rates of improvement recently 2. Ischemic Heart Disease (HD) has had higher rates of mortality improvement, but since these deaths account for a lower share of total deaths now, the improvement gains are having smaller impacts on overall mortality experience 3. Some causes are rising, meaning the number of people dying is increasing

State the three steps of the model life cycle

1. Specification phase 2. Implementation phase 3. Production phase

List the uses of reinsurance

1. Stabilizing underwriting results (reduce underwriting volatility risk) 2. Increasing underwriting capacity - Increase surplus for regulatory accounting purposes - Increase market share, maximize capabilities (distribution, underwriting, etc.) - Manage accounting ratios (e.g. net written premium to surplus) - Allows reinsurer to earn a share of profits without full insurance infrastructure 3. Supporting entry into and exit from insurance markets - Reinsurer can assume a portion of new risk or discontinued line - Reinsurer may have advice/expertise on underwriting, pricing, etc. 4. Capital allocation among affiliates (intra-affiliate reinsurance) - Pooling arrangements: pool/allocate underwriting experience for groups - Efficient capital allocation: tailor risk profile of separate legal entities - Captive reinsurance: used to manage statutory reserve levels (has been scrutinized) 5. Achieving risk target concentration or diversification - Especially helpful for smaller insurers

Outline the 3 phases of outbreak modeling

1. Start of outbreak - Primary goal: quickly assess need for and impact of significant mitigation efforts - Range of outcomes is very wide - Policymakers will want to know the worst case 2. Early stages of outbreak - Modeling shifts to specific regions and hospital demand - Assumptions/forecasts are refined as more data emerges 3. Deceleration of the outbreak - Modeling shifts to estimating effort of relaxed mitigation efforts

Identify two forms of financial reinsurance

1. Stat capital relief 2. VIF financing

State the four perspectives of the COSO internal control integrated framework

1. Strategic 2. Operational 3. Financial 4. Compliance

State the 5 components of active cyber risk management

1. Technology 2. Training 3. Risk limit 4. Risk monitoring 5. Cyber insurance

State tips 1 through 5 of the 10 Tips for Implementing Risk-based Controls

1. The focus should be on the business process and any sub-processes rather than just the audit process 2. The control should be focused on the end-to-end process and its dependencies rather than just on transactions. Although the control should address the accuracy of a transaction, a risk-based control addresses the total business practice and not just a single transaction 3. The expected outcome is to identify and mitigate risk as well as determine opportunities for process improvements within the operation 4. There should be a focus on risk management rather than solely on current policies and procedures. Current policies and procedures may be outdated or incorrect 5. The goal should be on continual risk assessment coverage through a continuous controls monitoring (CCM) process

Describe risk appetite and performance measurement methods

1. The gap between current profile and risk tolerance - Makes sure the business is safe compared to its risk tolerance - Can reflect at the top (group) level, risk category level, and/or risk limit level 2. Risk-adjusted return (RAROC) vs. expected RAROC 3. Risk-adjusted value: economic value added (EVA) vs. expected EVA - EVA = earnings - opportunity costs * capital allocated (the case study focuses on 2 and 3) 4. Encourages considerations of cost of capital and risk appetite

State two key differences between VaR and PFE

1. Time horizon: PFE may be defined at a point far in the horizon (e.g. several years) whereas VaR typically refers to a short (e.g. 10-day) horizon 2. Sign: PFE refers to a number that will normally be associated with a gain (exposure) whereas traditional VaR refers to a loss

Describe the 3 approaches for reserving for high claim frequency classes

1. Total Loss Ratio Method (simplest approach) - Looks at the total earned premium for a particular year and asumes that a particular proportion of those premiums will ultimately fund claims 2. Chain Ladder Method (dominant approach) - Projects reported claims forward with link ratios until claims are complete - 2 ways to apply: a. Apply to cumulative claims: places more weight on earlier claims b. Apply to incremental claims: may get volatile if long period since the claim occurred 3. Bornhuetter-Ferguson Method (combines the above approaches) - Key advantage: updates for experience

State the three levels to assessing the counterparty risk of a single transaction

1. Trade level 2. Counterparty level 3. Portfolio level

State the two most important tools for insurers' ERM frameworks to build resiliency and contingency actions against potential crises

1. Underwriting and accumulation control 2. Internal models and scenario analysis

List the 5 IAA risk categories for modeling

1. Underwriting risk 2. Credit risk 3. Market risk 4. Operational risk 5. Liquidity risk

What are two different methods for determining the carve-out point for liabilities?

1. Use the latest tenor where liability cash flows can be effectively minimized with available fixed income assets - The remaining amount of assets is invested in NFI securities 2. Determine the amount of exposure to NFI assets first, then determine the carve-out point based on how many years of cash flows the NFI assets can support - Example: Allocate 30% of the portoflio to NFI assets, which can support long-term liability cashflows that occur past 40 years from now

State two key similarities between credit risk and options

1. Volatility is an important, key aspect 2. Pricing is complex

List the 3 risk components for modeling

1. Volatility risk (diversifiable - sell more) 2. Uncertainty risk (non-diversifiable) - Made up of model risk and parameter risk 3. Extreme events (calamity)

Likelihood is a function of which two aspects?

1. Vulnerability 2. Situational exposure

List the summary of modeling decisions

1. What risk metric? VaR or CTE? 2. Stochastic analysis or stress testing? 3. Real-world or risk-neutral? 4. How much diversification to reflect? 5. Time horizon: one-year vs. multi-year? 6. Allow negative interim surplus? 7. Account for future new business? How much?

List some basic questions that one should ask before performing ALM

1. What sources of financial risk fall within the scope of ALM? 2. Which risk exposure matters, and which does not? 3. On what basis should risk be measured and managed? 4. What asset and liabilities should be included and which, if any, should be excluded? 5. At what aggregation level should ALM be performed?

State the responses to the following operational risks: - Regulatory risk

10 Regulatory risk - breaches can have serious costs and reputation implications - Culture: communicate regulatory obligations firm-wide and train staff - In-house compliance departments - Can subscribe to alert services - Lobbying for/against future regulatory changes

State the responses to the following operational risks: - People risk

11. People risk - very important in financial services industry - Indirect employment-related risks * Avoid recruiting the wrong people * Retain key talent with promotions, interesting work, flexible work arrangements * Deal with poor employees appropriately * Train employees and help them stay current with professional qualifications - Adverse selection - insurer fails to place insureds in correct risk class * Mitigate with underwriting * Set premium so lower risk members don't try to avoid the risk category - Moral hazard - client acts adversely after insurance or loan is in place * Make the consequences of bankruptcy or fraud unattractive * Only investigage possible insurance fraud if the savings outweigh the costs * Fraud increases in times of ecomomic stress - Agency risk - has lead to some of the biggest recent financial disasters * Sticks: rules for dealing with bad behavior * Carrots: incentives to encourage good behavior - Bias - deliberate and unintentional * Reporting checking: should be done by someone competent and independent * Compare one underwriter's decisions to others

State the responses to the following operational risks: - Legal risk - Model risk - Data risk

12. Legal risk - Similar to regulatory risk; the solution is to stay informed - Can occur on a case-by-case basis - Safest solution: seek legal advice if there is doubt over what legal action to take 13. Model risk - risk that a model has been incorrectly implemented - Minimize with a rigorous, documented process for model coding - Clear audit trail - Ensure that all models are designed for their intended use - Test a range of models 14. Data risk - another type of execution, delivery, and process management risk - Errors include: incorrect and/or omitted entries - Limit the data that can be input: * Restrict input format of date, number, and text fields * Check if dates entered are reasonable * Recheck data that is transferred from one system to another * Look for large frequencies of single values

State the responses to the following operational risks: - Reputation risk - Project risk - Strategic risk

15. Reputation risk - Difficult to respond to in advance - Best solution: sound ERM framework - rebuild reputation as quickly as possible if damaged 16. Project risk - Manage with a comprehensive ERM framework 17. Strategic risk - Constantly review whether the firm is correctly positioned in its market - Decide whether to compete on price or product differentiation - Decide whether to focus on a single product or a range of products

State the general liquidity risk considerations

2 general types of liquidity risk: funding and market It's critical to understand the liabilities before modeling the assets - Liability options: accelerate/postpone payments, lapse, etc. - Assets: timing, certainty, and liquidity of cash flows Modeling challenges: limited data and very company-specific - Pensions tend to have the least liquidity risk - Insurers; varies by product - Banks tend to have the most liquidity risk

Describe how to compute the following common metrics to measure liquidity risk: 1. Liquidity Ratio 2. Excess (Deficit) of Liquidity

2. Liquidity Ratio = Available Liquidity Resources / Liquidity Needs 2. Excess (Deficit) of Liquidity = Available Liquidity Resources - Liquidity Needs

Define arbitrage

Arbitrage - the opportunity to make a riskless profit by exploiting price differences of identical or similar financial instruments, on different markets, or in different forms

List the 3 definitions and 3 focuses of risk appetite

3 definitions of risk appetite: 1. Total risk exposure undertaken to achieve objectives 2. Total impact of risk accepted in pursuit of objectives 3. Maximum amount of risk willing to accept in pursuit of objectives 3 focuses of risk appetite: 1. Protect and create value for the business - Analyze risk/return trade-off, maximize RAROC 2. Consistency between risk appetite and risk limits 3. Integration with business strategy and corporate culture - Guides risk-taking activities and facilitates risk monitoring

List the types of demographic risk

4 types of mortality risk: 1. Level risk - the risk of misestimating the underlying level of mortality - Experience rating - estimate past level mortality of a group using experience data - Risk rating - use the characteristics of lives in a gorup to estimate level mortality - Blend with credibility (Z * Experience + (1 - Z) * Risk Rating) 2. Volatility risk - actual deaths will differ period to period even if estimates are good 3. Catastrophe risk - risk of a large, temporary increase in mortality - Wars, pandemics, etc. 4. Trend risk - risk that future mortality level is different than expected All apply to longevity except catastrophe

State the responses to the following operational risks: - Damage to physical assets - Business disruption and system failure - Execution, delivery, and process management

4. Damage to physical assets - Limit with insurance if it is cost-effective - Avoid business locations with a high risk of natural disasters, crime, etc. - Use security guards and cameras 5. Business disruption and system failure - Contingency plans for alternative business locations if business is disrupted - Back up data regularly (ideally off-site) - Ensure key personnel can work from home - Consequential loss insurance - compensate for lost profits 6. Execution, delivery, and process management - Comprehensive documentation of processes is essential - Error logs that identify process failures - Reporting of "near-misses" so that lessons can be learned - Stress test new procedures

State tips 6 through 10 of the 10 Tips for Implementing Risk-based Controls

6. Risk-based internal controls facilitate change since they should be updated when there is a significant change to the business process or if the control is found to inadequately mitigate a potential risk 7. This approach should set the foundation for implementing operational metrics and analytics 8. Risk-based controls can identify risks and business process gaps across financial operations 9. Risk-based controls can help prevent and detect fraud since they should represent the end-to-end business process 10. Risk-based controls should always be developed by the business process owners, but approved by management with well-defined implementation and remediation plans

State the responses to the following operational risks: - Crime risk - Technology risk - Cyber risk

7. Crime risk - Responses vary by type of crime - Hacking, adverse selection, moral hazard (overlaps with people risk) 8. Technology risk - Key decision: outsource vs. keep in-house - Important to have a dedicated central IT resource - Back up data - Keep software up-to-date - Risks of new software - Balance bespoke (custom) IT systems with off-the-shelf solutions 9. Cyber risk - Keep software up-to-date - Firewalls, anti-virus software, etc. - Ethical hacking: "get paid to break in" - Control employee access and use of peripheral devices - Cyber risk insurance

Describe what the carveout strategy is

A carve-out strategy explicitly separates out the long-term liability cash flows after a certain number of years (called the carve-out point) - Standard fixed income assets would be used to minimize the short-term liability cashflows that occur before the carve-out period - NFI assets would be used to back the long-term liability cashflows that occur after the carve-out period - This method allows an insurer to explicitly measure and manage the risk exposure associated with using NFI assets to back the long-term liabilities

Describe a hybrid model in the context of cyber risk

A hybrid model uses both limited experience data and expert opinion to evaluate cyber risk exposure

Describe a line of credit

A loan arrangement with a bank allowing the customer to borrow up to a prespecified amount

Define benchmarking

A model validation procedure comparing model results to the results of similar models, or to other external criteria

Describe effective expected exposure (EEE)

A non-decreasing expected exposure

Define model developer

A person who is responsible for designing, constructing, and maintaining the model

Define model steward

A person who oversees the model's development and use. A model steward is responsible for assuring that the model developers, model users, and model reviewers are carrying out the responsibilities outlined in the model governance policy

Define model reviewer

A person who provides an independent review of the reasonableness of various aspects of the modeling process. A reviewer will be someone distinct from the person who performed the work being reviewed, and will possess adequate knowledge and expertise regarding the model and its uses

Define model risk ranking

A process that may be used to consolidate the individual model risk scores and sort all models or a subset of the models to determine a relevant risk ranking for each model in the enterprise. An overall score relative to the severity of the model's risk could be determined for each model relative to the other enterprise models

Define model

A representation of relationships among variables, entities, or events using statistical, economic, mathematical, or scientific concepts and equations

Define scenario (according to ERM-120)

A scenario is a possible future environment, either at a point in time or over a period of time

Define sensitivity (according to ERM-120)

A sensitivity is the effect of a set of alternative assumptions regarding a future environment

Define stress testing (according to ERM-120)

A stress test is a projection of the financial condition of a firm or economy under a specific set of severely adverse conditions

List key features of ERM impacted by climate risk

Climate-related risk should be incorporated into each of the following: - Governance and an ERM framework - Risk management policy - Risk tolerance statement - Risk responsiveness and feedback loop - Scenario analysis and ORSA

State the EIOPA data quality requirement definitions for: - Accurate data - Complete data - Appropriate data

Accurate data - Free from material mistakes, errors, and omissions - Timely and consistent recording - High levels of confidence - Credibility demonstrated through usage in the decision-making process Complete data - Allows recognition of the main homogeneous risk groups - Sufficient granularity to identify trends and full understanding of the underlying risks - Sufficient level of historical detail is available Appropriate data - Fit for purpose - No bias - Relevant to the portfolio of risks of the insurer

Describe considerations for international (multi-economy) ESGs

Additional needs for a multi-economy ESG: - Cross-economy calibration and validation requirements - Processes for foreign exchange rates - Global correlation matrices Additional challenges for a multi-economy ESG: - ESG variables may behave materially different in each economy - Historical data varies greatly (short histories, structural changes) - History may not be useful for modeling future co-movements - Arbitrage-free requires a shared stochastic process across economies

List the advantages and challenges of Monte Carlo valuation

Advantages of Monte Carlo: - Works for any embedded financial derivative - Can relax lognormal assumption and others Practical challenges with Monte Carlo: - Computationally intensive - Require judgement to create scenarios and set parameters - Calibrating to risk-neutral constraints may result in unrealistic distribution shapes - Must balance model accurage with cost of increased complexity

State the key advantages and disadvantages of using financial reinsurance vs. debt and equity

Advantages of financial reinsurance: 1. Reduces required capital or increases RBC ratio 2. Doesn't increase financial leverage ratios like debt 3. Non-dilutive, unlike equity 4. Private transaction - lower costs, more flexibility, and less legal documentation Disadvantages of financial reinsurance: 1. Introduces counterparty risk 2. May require substantially more management restrictions 3. May require substantial administrative and systems changes

List the types of alternative risk financing mechanisms

All alternative risk financing mechanisms must be legal entities - Insurance risk financing techniques (main focus of ERM-122) 1. Risk purchasing groups (RPGs) 2. Self insurance pools ("pools") 3. Captive insurance companies ("captives") - Non-insurance risk financing techniques include: 1. Trusts 2. Capital markets solutions: securitizations, hedging, etc.

Define risk factor

Anything that affects the chance of a disease or injury, but does not necessarily cause death

Describe the Merton model (structural credit model)

Applies the Black-Scholes framework in the context of credit risk: - Assumes firm value follows a lognormal random walk - Debt = strike price of a call and put owned by shareholders - If assets > debt, shareholders own the gain - If assets < debt, shareholders' loss is floored (default on debt) Drawbacks: - Same as Black-Scholes (ignores transaction costs, assumes lognormal returns, etc.) - Does not link debt level and expected returns

List the steps and advantages/disadvantages of the Multivariate Stress Tests for EC

Apply the previous methods to more than 1 risk at a time: - Immediate stress: specify multiple risk factors to apply instantaneously - Projection scenarios: model multiple risk factors in a scenario using ESGs Advantages: - Captures risk interactions (cross-Greek risk) - Can calculate EC for a range of different confidence levels - Evaluate pros and cons of alternative risk management strategies - Can simultaneously reflect both "normal" and tail risk Disadvantages: - Increases modeling complexity and computational requirements - Practically impossible to specify in advance all the scenarios that should be used

Describe non-parametric credit scoring models

Approaches that attempt to determine credit groups that firms belong to: - The k-Nearest Neighbor Approach * Drawbacks: ** Determining the number of neighbors is not straightforward ** Choice of dimensions is difficult ** High volume of calculations required (long computer run time) - Support Vector Machines (SVMs) * Each dimension model could represent a different accounting ratio * Flexible, but can create over-fitting problems if non-linear versions are used

List the steps for the discrete marginal contributions approach

Approximates the continuous marginal contribution approach 1. Determine the capital of the total portfolio, excluding the portfolio in question 2. Initial marginal contribution for each risk = Aggregate VaR - VaR from Step #1 3. Repeat Steps #1 and #2 for each risk 4. Sum all of the initial marginal contributions 5. Scale the initial marginal contributions by a factor of (Aggregate VaR / Step #4)

Define "arbitrage", and briefly describe a test to ensure that an ESG is arbitrage-free

Arbitrage - ability to buy/sell securities in a way that generates a risk-less profit - Undesirable in ESGs, especially if used for market-consistent pricing/validation ESGs used for RN should be arbitrage-free - Verify arbitrage-free with marginale tests * Average PV of RN scenarios = time zero price of asset - Often required by regulators

Describe the Solvency II use test for internal models

As part of the internal model approval process (IMAP): - Senior management must: * Understand model and how it fits with business and risk-management framework * Understand model limitations and reflect in decision-making * Ensure timely calculation and avoid time lags for decision-making - How to help senior management understand the model * Board could learn methodologies (e.g. copulas) at a high level * Simplify model until Board can understand it ** Advantages: easier calibration, less maintenance, faster ** Disadvantage: may not handle tail dependency and asymmetry

Compare the cost of capital method to quantile methods for calculating the IFRS 17 risk adjustment

Aspects of the CoC method that are different from quantile methods: - More complex - Based on a higher percentile - Reflects how risk changes with liabilities over time - Based on a probability distribution of insurance cash flows - Reflects duration and uncertainty of cash flows - Accounts for the cost of securing the cash flows

Describe how risk management influences AM Best's financial strength ratings

Assessing balance sheet strength involves analysis of BCAR and: - Regulatory filings and GAAP/IFRS financial statements - Capital structure/sources/uses, leverage, fixed charge coverage, and liquidity Future balance sheet strength is assessed by looking at operating performance and business profile - Profitability ultimately drives capital - Strong performers have earnings that are relatively consistent/sustainable * Will have higher ratings and/or lower BCAR requirements vs. weak performers - Weak performers have consistent losses or high volatility

Describe the characteristics of financial time series

Asset returns sometimes follow short-term trends but not long-term trends Characteristics of Short-Term Trends: - Volatility clustering - large or small returns can occur in groups * As general volatility increases, it affects all stocks and asset classes - Leptokurtic (thick tail) - extreme returns occur close together - Correlations exist between stocks and asset classes but aren't stable - Cross-correlation - the change in a stock at time t has no effect on any other stock at time t + 1 * May exist for absolute or squared returns

Describe the 3 pillars within Basel II

Basel II - capital standard for European banks 1. Pillar I - minimum capital based on credit, operational, and market risk 2. Pillar II - supervisory review 3. Pillar III - market discipline

State the risk aggregation approach for Solvency II and US RBC capital requirements

Basic SCR = var-covar approach - Problem: assumes constant (Gaussian) correlations * Assumes zero tail dependence (not good for jointly fat tails) * Understates VaR NAIC RBC is more extreme: correlations are either 0 or 1 - Assuming total independence underestimates RC - Assuming total dependence overestimates RC

Compare the basic reproduction number with the effective reproduction number

Basic reproduction number: R0 - Average number of persons that an infected person will infect - Assumes 100% of the population is always susceptible to infection - If R0 < 1, new infections will decline over time; else infections will grow - All else equal, as R0 increases, outbreak speed increases * Other factors also affect outbreak speed (e.g. infectious period) * Slow and fast outbreaks can have the same R0 Effective reproduction number: Re ≤ R0 - Like R0 but accounts for recovery/immunity, mitigation efforts, etc. - Re = R0 initially, then becomes less than R0 - If 50% of the population has recovered, Re = 0.50 * R0

Discuss managing market risk in derivative portfolios

Before aggregating, must determine the net exposure! 1. Delta risk - sensitivity to changes in the underlying security's price - "absolute price" or "rate risk" - Delta varies between -1 and 1 2. Gamma risk (convexity) - sensitivity of delta to changes in the underlying security's price - Higher nonlinearity (convexity) means higher risk - If gamma is nonzero, a portfolio will not be "delta hedged" after the underlying's price changes 3. Vega risk - sensitivity to changes in the underlying's volatility - Common with options 4. Theta risk - exposure to a change in value as time passes - Common with options 5. Basis or correlation risk - exposure to differences in derivative and hedge performance 6. Rho risk - sensitivity to changes in discount rate used to PV cash flows

Compare and contrast pure coinsurance with YRT reinsurance

Benefits covered: - Coinsurance covers ALL, not just mortality/morbidity Reinsurance premiums: - Coinsurance ceded premium = QS% * Policyholder Premium - YRT premiums independent of policyholder premium Reinsurance reserve credit: - Coinsurance ceded reserve = QS% * policy reserve - YRT credit = unearned YRT premium reserve Capital and solvency margin relief: - Much greater under coinsurance since more risks are transferred Ceding commission: - Higher with coinsurance (more profit potential for reinsurer)

Describe how AM Best views principles-based solvency regimes

Best strongly supports principles-based solvency regimes Common themes shared by solvency requirements and Best's rating approach: - RM is part of a balanced quantitative and qualitative review - Support for the development of internal capital models - RM and capital modeling are not "one size fits all"

Describe AM Best's evolving approach to ratings with respect to risk management

Best will consider allowing lower BCAR for insurers with: - Superior traditional RM relative to their risk profile - Superior capital management, financial flexibility, and cost-efficient access to capital in distressed scenarios - Strong ERM characteristics - Strong EC modeling capabilities * Providing a denominator for risk-adjusted returns * Provides understanding for correlations holistically * Best may put more weight on internal EC models

Describe how the ERM process influences AM Best's rating components

Best's rating components generally follow the ERM process 1. Decisions: lines, segments, territories, limits, etc. 2. Measurement and monitoring of: earnings, key risks, and correlations 3. Impact on: capital, EC, probability of default/downgrade When an insurer's risk management capabilities exceed its risk profile, it has a positive effect on ratings (else negative)

List the steps for bootstrapping

Construct an implied spot rate curve from the gross redemption yields of coupon-paying bonds: 1. Calculate the "dirty" price of a coupon-paying bond with the shortest maturity 2. Solve for the implied spot rate for the shortest term: s1 3. Repeat step 1 for the next-shortest maturity 4. Repeat step 2 to solve for s2, substituting in s1 for the year 1 cash flow 5. Continue this process for the remaining maturities

Describe the modeling approaches for market and economic risks

Bootstrapping - construct implied spot rate curve from observed bond yields - Drawback: can overstate bond returns if bond yields are trending down Factor-based approaches - Model lognormal of bond price as a function of risk-free rates, credit spreads, etc. - Account for heteroskedasticity with ARCH or GARCH - Drawback: gets more complex as more factors are added Principal Component Analysis (PCA) - Good alternative to factor-based approaches - Starts with the factor-based approach, then chooses only the components (factors) that explain most of the variance in the dependent variable * Select most important single factor first, then second most important, etc. * Uses correlation eigenvalues to determine importance - Drawback: assumes normal distribution for covariances * Ignores skewness and kurtosis * Copulas may be better (if enough data)

Describe credit default swaps, including key parties and drawbacks

CDS = "insurance" bought to protect against the default of a bond - Unlike other insurance, there is no insurable interest - Conceptually like a default-triggered put option with a strike price at the bond's par value Key parties: - Protection buyer pays an ongoing premium to the protection seller * Premium = fee based on notional value of reference entity bond - If the reference entity defaults (must be defined), can settle in 2 ways: 1. Physical settlement: seller pays par value to buyer, seller receives the bond 2. Cash settlement: seller pays cash = par - MV Drawbacks: - OTC: no exchange protection and no standardized structure - Criticism for allowing anonymous investors to drive bond prices down

List the properties of a coherent risk measure and indicate which quantile method satisfies all criteria

CTE meets all the criteria for a coherent risk measure: 1. Monotonicity: if x ≤ y, then f(x) ≤ f(y) 2. Subadditivity: f(x+y) ≤ f(x) + f(y) - VaR fails this - sometimes VaR(X + Y) > VaR (X) + VaR(Y) - Important for reflecting diversification benefits 3. Homogeneity: f(ax) ≤ a * f(x) for a constant a > 0 4. Translation invariance: f(x + a) = f(x) + a for a constant a

Describe the continuous marginal contributions (Euler Method) approach

Calculates the derivative of the total portfolio risk with respect to each individual portfolio's risk - Primary drawback: can get negative contributions if negative correlations exist

Describe the purpose of ESG calibration and compare market-consistent with real-world parameterization

Calibration - setting parameters to produce the distributions/dynamics required for the application - Should reflect correlations across asset classes RW vs. RN/MC Calibration: - MC applications: ESGs must satisfy RN and arbitrage-free conditions - RW parameterization is an iterative process * Simple regression not effective since market variables are not independent * Require more robust tools like MLE and/or Kalman filtering

List the responses to foreign exchange risk

Can use many of the same instruments as IR risk (forwards, futures, options, etc.) Unique aspects of foreign exchange risk: - Usually undesirable: does not offer additional return, only risk - Risk exposure to hedge is not always straightforward - A firm with business in many countries may naturally be hedged - Overseas equities are often hedged either with some rule of thumb or not at all Important to establish the net level of exposure - Only the difference between a payable and receivable should be hedged in a particular currency

State why capital allocation is necessary and the factors to consider in allocation

Capital allocation is necessary to... - Measure profitability relative to risk - Price insurance products - Planning and control cycle (risk budgeting and return measurement) Capital may be allocated by a number of factors including: - Risks - Products / product groups - Geographical locations - Entities / organizational units

Define captives and list the 3 essential elements of captive insurance

Captive - an insurance company that is wholly owned and controlled by its insureds - Primary purpose: insure the risks of its owners - The captive's underwriting profits primarily benefit the insureds - Has a contractual obligation to pay future losses, not just administer payments The 3 essential elements of captive insurance: 1. Insureds put their own capital at risk - Unlike RPGs and pools 2. Operates outside of the commercial insurance marketplace 3. Must be used to achieve the insureds' risk financing objectives - "Insured" = source and owner of risk (sometimes different) - "Owner" = entity that bears ultimate financial impact (e.g. sponsor) - Risks insured by a captive do not have to be related

Discuss catastrophe bonds, including benefits to cedant and possible bond triggers

Cat bond = structured debt that transfers risk from a sponsor to investors - Issued through an SPV created by the sponsor (insurer) - Terms and conditions similar to a reinsurance contract - Investors' funds are held in a segregated collateral account in the SPV Benefits to cedant: - Diversification - Costs can be cheaper and more predictable than reinsurance - Manage credit risk with collateral Possible cat bond triggers: - Indemnity trigger (most common): actual incurred loss of the sponsor - Parametric trigger: a description of the physical characteristics of the catastrophe - Industry loss trigger: amount of insured loss incurred by the industry

Define catastrophe risk

Catastrophe risk - risk of a large, temporary increase in mortality - Wars, epidemics, etc. Use scenario analysis in the simplest cases or copulas in more complex - Scenario example: observe the impact of a 20% increase in mortality across all ages Not a type of longevity risk

Describe causal modeling, including CLD and outputs

Causal modeling - simulate non-linear relationships with Monte Carlo - Example: simulate yield curves, then simulate insurance losses Causal loop diagram (CLD) - nodes represent connections between variables - Positive feedback: nodes change in the same direction - Negative feedback: nodes change in different directions - Symbols: C = counteracting, // = delay, thick line = more important Outputs always respond to inptus in a "system" - Open system: outputs do NOT influence inputs - Closed system: outputs do influence inputs Drawbacks: can't rely on CLDs alone; may need copula or var-covar matrix for residual dependency

Describe credit triggers

Clauses in financial contracts that allow creditors to demand immediate payments if the credit rating of the borrow falls below some predetermined level

Describe climate risk's impact on capital adequacy

Climate risk might require: - More extensive reinsurance - Better monitoring of portfolios - Stricter limits on policies - Withdrawal from individual lines of business Rating agencies may introduce new climate measures setting ratings

Describe models for generating corporate bond yields

Corporate bond prices = f(general interest rates, credit expectations) 4 general classes of corporate bond model approaches 1. Structural models (Merton) - prices debt using a model based on the issuing firm's value 2. Reduced-form models - models time until default and intensity (e.g. using Poisson) - Linked to survival distribution theory - Cannot model rating changes (i.e. use for bonds with stable ratings) 3. Ratings-based models - models changes in the corporate bond spreads in response to rating changes - Does not allow for idiosyncratic bond price behavior - Closed-form version requires simplifying assumptions 4. Hybrid credit models - used for specific credit applications

Define counterparty risk

Counterparty risk is defined as the risk that a counterparty in a derivatives transaction will default prior to expiration of a trade and will therefore not make the current and future payments required by the contract

List the approaches for credit, market, and operational risks

Credit Risk - modeled consistently with banking standards - Default, credit migration/transitions, spread volatility - Reinsurance default risk - use stochastic Monte Carlo to model Market Risk - modeled consistently with banks and other financial institutions - Changing yields affects MV of liabilities, so you must use ALM - MV of liabilities must be technically derived because there is no tradeable market - Can use risk-neutral (RN) or real-world (RW) valuation Operational Risk (OR) - can be assessed qualitatively or quantitatively - Simple add-on model - aggregate anticipated OR costs - Stochastic frequency/severity model - use scenario analysis for main ORs

Describe credit exposure (CE)

Credit exposure is defined as the replacement value of the asset (floored at zero) on the target date

Discuss the characteristics of parameteric credit scoring models

Credit score = function of financial ratios, etc. Benefits of using financial ratios as independent variables: - Consistent basis for comparing firms of different sizes - Sensible comparisons over time (removes effect of inflation) Drawbacks: - Assumes a normal distribution * Financial ratios usually have complex distributions (rarely normal) * Distributions can be U-shaped, skewed, and have a lower bound of zero - Does not allow the inclusion of qualitative factors

Describe how to measure current and potential credit exposure

Current exposure = current market value (easier) - Captures cost of current default Potential exposure = estimate of the future replacement value (harder) - Captures cost of future default - Requires probability analysis and broad confidence intervals (e.g. 2 σ's) - 2 measures of potential exposure: 1. Expected exposure (mean PV) 2. Maximum or "worst case" exposure (e.g. VaR) - Simpler approach: use tables of factors that vary by derivative type and maturity

Define derivative legal risk and its sources

Derivative legal risk - risk of loss because a contract cannot be enforced by law - Developments that have reduced legal risk: * Widely-used standard documentation * Changes in the legal environment that have answered enforceability questions Sources of legal risk in derivatives: 1. Insufficient documentation 2. Insufficient capacity or authority to enter derivative - Greater in the UK due to ultra vires 3. Uncertain legality of contracts - Greatly reduced in the US by the Futures Trading Practices Act of 1992 4. Unenforceability in bankruptcy or insolvency - Close-out netting reduces risk - Automatic stay provisions increase risk

List 3 desirable characteristics of an ESG validation system, and list considerations for validating data

Desirable characteristics of an ESG validation system: - Automated (better than manual) - Repeatable and consistent - Create acceptance criteria before observing output (no "problem discovery" Considerations for validation data: - Accuracy or degree of confidence in the data - Completeness - Bias present (data should be appropriately unbiased) - Relevance

Define internal controls

Desired goals or conditions for a specific event cycle which, if achieved, minimize the potential that waste, loss, unauthorized use, or misappropriation will occur

Compare resiliency vs. contingency planning

Difference between resiliency and contingency planning is based on the timing and how the uncertainty unfolds: - Resiliency - characterized as actions taken before the risk event occurs - Contingency plans/actions - characterized as actions taken after the risk event occurs

State responses to direct interest rate exposure

Direct exposure - when the company needs to pay or receive ongoing interest payments - Forward rate agreements - OTC agreement to pay interest on a notional amount at some point in the future - Interest rate caps and floors - series of caplets or floorlets

Describe diversification with respect to economic capital

Diversification = extent to which EC < straight sum of the individual capital amounts - Increases as risk dependencies fall - Depends on the level of aggregation

Contrast diversifiers and hedgers

Diversification and hedging are two principle secondary techniques to manage risks: Diversifiers want to avoid high positive correlation - Prefer correlation numbers to be very small positives, zero, or even negative Hedgers prefer high positive correlation - Goal: find hedging instruments that are highly correlated with risk exposures - Insurance example: quota share or excess of loss reinsurance - Capital markets example: derivatives issued by banks that hedged indexed insurance contracts

State the drawbacks of the Black-Scholes model

Drawback: assumptions are not always appropriate in practice (but still a good "first cut") - Assumes asset returns are lognormal and follow a random walk - Assumes fixed values for all of the parameters - Requires that a perfect hedge is always available - Assumes zero transaction costs

Define economic capital and its drivers

EC has "2 meanings" (Available EC = Required EC + Excess Capital) 1. Available EC = realistic or market-consistent amount actually available 2. Required EC = realistic amount a business believes it needs to meet future risks Drivers of EC analysis: 1. Embedded value calculations 2. Regulatory pressure (Basel II, Solvency II, IAIS) Questions for decision-makers using EC analysis: 1. Type and scope of risks? 2. How to measure risk? Acceptable probability of ruin? 3. What decisions should underpin EC model development?

Describe how AM Best believes ERM should vary by insurer

ERM becomes more important as both of the following increase: - Size and complexity of the insurer - Relative risk and volatility Large, complex insurers must refine their ERM framework and EC modeling to... - Remain competitive - Build sustainable earnings and capital - Maintain high ratings Smaller insurers may require a less sophisticated ERM framework but should: - Foster a risk-aware culture - Improve ability to identify, monitor, and manage risk - Consider impact of risk correlations

Define "economic scenario generator"

ESG - a computer model of an economic environment that is used to produce simulations of the joint behavior of financial market values and economic variables - Output: interest rates, equity returns, exchange rates, etc.

Describe the importance of stylized facts in an ESG model, and provide examples of stylized facts

ESGs should reflect stylized facts about the economy - Based on historical analysis and sometimes judgement - Can account for things that basic statistics cannot (fat tails, path-dependence, etc.) Examples of stylized facts: - The yield curve is usually upward-sloping, especially as short-term yields fall - Short-term yields tend to fluctuate more than longer-term yields - Interest rates can be negative - Corporate credit spreads rise as credit quality falls - Default probabilities fluctuate with the general economy, firm- and industry-specific conditions - Equity return volatility fluctuates significantly but is generally higher than bond volatility - Market variable correlations are not stable over time and depend on the frequency of observation

Define economic capital and describe its main components

Economic capital = realistic amount of capital needed to cover losses at a certain risk tolerance level - Main components: risk measure, probability threshold, and time horizon

List the three primary valuation and risk management practices

End-users and dealers should adopt the same valuation and market risk management practices 1. Regularly MtM derivatives for risk management purposes 2. Periodically forecast cash and funding requirements 3. Independent, authoritative function to design and adhere to prudent risk limits

Discuss the trends in performance measurement with regards to EEV vs. ROE

European Embedded Value (EEV) is trending towards MCEV - Some companies use EC as the basis for cost of capital - EC approach allows companies to account for non-market risk - There is still "wide divergence" in capital approaches Traditional ROE is trending toward RAROC - Considering EC allows insurer to more accurately price business lines * Traditional ROE, etc. does not do this - Risk-adjusted measures allow comparisons across different business types (e.g. insurance vs. banking)

Define trend risk and two broad types of approaches

Expected future mortality level and the uncertainty of the expected level - Worsening mortality hurts life products (mortality risk) - Improving mortlaity hurts payout annuities (longevity risk) 2 broad types of approaches: 1. Non-parametric (P-Spline Approach) - Smooth historical mortality, then project - Drawback: no uncertainty level, so can't use it for large numbers 2. Parametric (model mortality as a function of factors) - Common factors: age (+), calendar year (-), year of birth (cohort effects) - Can be projected stochastically - All-cause mdoels are the most common, but cause of death models also exist - Dependent variable is usually ln(q) or logit because mortality is nonlinear

Describe the modeling for expected government bond returns

Expected return = Gross Redemption Yield (GRY) on a government bond with the term = projection period - Reflect appropriate term structure (shape of yield curve) - Rising yield curve may indicate higher risk for longer assets * Exception: investors with long liabilities

Describe Extreme Value Theory (EVT)

Extreme Value Theory (EVT) is a branch of statistics dealing with the extreme deviations from the median of probability distributions - The main result of EVT is that it characterizes the distribution of: * The sample maximum * The distribution of values above a given threshold - EVT quantifies the potential black swans hinted at by historical extremes - Makes use of previous extreme values to offer information on the probability and magnitude of potential values more extreme than seen previously

Compare FIA, VA, and Hybrid Annuity product designs

Fixed Indexed Annuity (FIA): - No downside risk - Participate indirectly in the performance of the underlying index - Fixed account is available - Guaranteed minimum crediting rate Variable Annuity (VA): - Unlimited downside risk - Invest directly in separate accounts/funds - Fixed account is often, but not always available - No guaranteed minimum crediting rate outside of the fixed account Hybrid Annuity: - Downside risk with some protection (buffer/floor) - Participate indirectly in the performance of the underlying index - Fixed account is not usually available - No guaranteed minimum crediting rate

Describe the following as they relate to data used in operational risk management (ORM) analysis: - Outliers - Homogeneity - Data sources

For ORM, outliers are the most relevant data (don't toss them!) - Differentiating outliers from relevant tail data is difficult in small samples Maintain homogeneity: don't aggregate routine execution errors and unauthorized activities External data sources: - External public data (media reports, legal settlements, etc.) * Usually well-documented * Chief disadvantage: reporting bias (not all losses are reported) * Best used as soft data for tail - Consortium data (supplied by member institutions on an anonymous basis) * Chief disadvantage: anonymity reduces descriptive information * Most useful when effective controls ensure data quality/consistency

State the intent of ORSA

For the insurer to provide a regulator with its own view of its risk

Discuss managing market risk of individual derivatives

Forward-based derivatives (i.e. forwards, futures, etc.) - Relatively straightforward: primarily has delta risk (little gamma risk) - Typical hedge = delta hedge or hold offsetting position in another forward Option-based derivatives - more dynamic and complex - Black-Scholes model has 5 factors that determine many option prices: 1. Price of the underlying 2. Exercise (strike) price of the option 3. Time to expiration of the option 4. Volatility of the underlying's price 5. Discount rate over the life of the option - Options have more exposure to gamma, vega, and theta risk - Dynamic hedging - continuously hedging an option with a position in the underlying * Two main risks: 1) gamma risk and 2) hedging cost could be higher than expected

Describe possible future advances in policyholder behavior modeling

Future possibilities based on developments in other industries: 1. "Smarter" databases containing data "microdata" 2. Exponential advances in computational power 3. Better understanding of decision-making processes and behaviors Approaches with the most potential for policyholder behavior: 1. Move from statistical averages to understanding household decision-making 2. Model the causal structure of decision-making, including the effect of biases

List the marginal approaches to allocating capital vs. game theory

Game Theory 1. Shapley Value 2. Aumann-Shapley Value Marginal Approaches 1. Pro-rata or linear (simplest) - uses Shapley Value 2. Discrete marginal contributions (approximates #3) - uses Shapley Value 3. Continuous marginal contributions 4. Myers-Read allocation method - special case of Aumann-Shapley Value Game theory is widely used for decision-making in conflict situations - in risk allocation, the conflict is how to share the diversification benefit

Describe how to incorporate parameter risk into simulation models

General 2-stage process: 1. In each scenario, simulate the parameters from the parameter-risk distributions 2. Simulate the process from the simulated parameters Applying the process to a trend example: 1. Simulate aggregate claims L 2. Simulate the trend factor J 3. Multiply J and L to create a distribution of K = J * L 4. CV(K) approaches CV(L) as the expected number of claims N gets larger For Pareto: 1. Simulate α_hat 2. Simulate claims based on α_hat In all cases, even if process risk diversifies away, parameter risk will not

List product management considerations with respect to climate risk

General considerations: - Premium increases in areas susceptible to flooding, fires, etc. - Balance increased granularity in pricing with the social benefits of pooling risk - Changing product design may result in unanticipated consumer behavior Products that align policyholders' interests with favorable climate outcomes: - Incentives that eliminate/control risk in low-carbon products - Provide capital for climate-related risk initatives

Describe how to measure tail dependence

Given that Y is very large, what is the probability that X is very large too?

What is the cost of trading?

Half of the bid-ask spread

Describe the differences between hard data and soft data

Hard data - Collected systematically on a prospective basis for a short period (e.g. 15 years) - Example: daily ocean wave height over a 5-year period - Yields many more data points than soft data - Not appropriate for estimating a 1-in-100 year event Soft data - Uses a less robust process or a proxy variable - Example: geological record of tsunami waves in the past 500 years - More valuable for risk analysis: superior when hard data is insufficient - Better for modeling catastrophic events - 500 soft data points over a 100-year period are better than millions over a 5-yr period Many risks are underestimated because hard data doesn't capture catastrophic events as often!

What are the two main implications of using hedging to manage counterparty risk?

Hedging can both: - Reduce the exposure - Provide a means to transact further with the counterparty

Describe considerations around climate-related risk disclosures for climate risk

How disclosures may be relevant to actuaries: 1. Actuaries will develop climate risk disclosures 2. Actuaries will use disclosures from firms invested in by insurers, etc. TCFD's 4-pillar approach for climate risk disclosure: 1. Governance - organization's governance around climate risks/opportunities 2. Strategy - actual/potential impacts of climate risks/opportunities 3. Risk management - how the organization identifies, assesses, and manages climate risks 4. Metrics and targets - metrics and targets used

Describe how equity index returns are different from bond returns and other features of equity index ESG models

How equity indexes are different from bonds: - Equity price index is random in perpetuity - More volatile - Can exhibit sudden jumps Features of equity index ESGs: - Capable of modeling jumps, stochastic volatility, and accurate returns - Common drawback: can't determine a unique price because of incomplete markets (i.e. a market where it's not possible to hedge away all risks) - Produce total returns (price changes plus dividends) * Dividend yield is modeled as a percentage of price to conform to no-arbitrage restrictions - Must account for co-movements in related indexes (e.g. large cap, mid-cap, etc.) * Important for VA applications - Most do not directly relate returns to macroeconomic factors

Describe the following requirements for Solvency II internal models: - Statistical Quality Standards - External Models and Data

IMAP Statistical Quality Standards - Diversification benefits must be based on some combination of: * Empirical/statistical analysis (may require a lot of work) * Expert judgement (explain and document in detail) - Diversification must hold on average and in extreme scenarios! IMAP External Models and Data Requirements - 3rd party ESGs - demonstrate understanding of dependency modeling, calibration, and limitations

What is the relationship between available capital and required capital? How does diversification affect required capital?

If available capital >>> total required capital: - Company can pay dividend or buy back shares - Could make an acquisition If available capital < total required capital: - Could raise capital or sell BUs Capital should be allocated after considering diversification benefits - Invest more in BUs with singificant diversification benefits Diversification benefits can arise... - Within products containing offsetting risks - Within BUs containing products with natural hedges (e.g. life/longevity) - At company level if there are different types of businesses Must understand capital fungibility - how easy is it to move capital?

Describe the relationship of climate risk with the pension fund sponsor

Impact on employer covenants - Environmentally unfriendly companies may be less profitable in the future - Affects ability to fund pension obligations Impact on mortality/morbidity of employees

List the steps and advantages/disadvantages of the Immediate Stress Approach for EC

Steps: 1. Apply a risk factor at time 0 2. EC = decrease in (A - L) resulting from the risk factor (floor at zero) Advantages: - Simple and easy to understand - Popular internationally: used for SST, Solvency II, ICA Disadvantages: - Ignores risk interdependencies (especially the single-factor approach) - Ignores risk mitigation strategies (management actions, dynamic hedging) - Ignores future cash flows external to the time 0 calculation of A and L

Describe how the extent of loss may be modeled, including factors affecting expected recovery and modeling considerations

In practice, the recovery rate is usually modeled, not the extent of loss Suppose a 1000 par bond defaults and investors ultimately recover 400 - recovery rate = 40% Can be modeled with 2 measures: 1. Price after default: better for short-term 2. Ultimate recovery: better for long-term (usually higher than 1) Factors that affect expected recovery: - Seniority of the obligation - Industry - Point in the economic cycle - Degree and type of collateral - Jurisdiction - Composition of the creditors Other modeling considerations: - Usually modeled with historical data (deterministic or stochastic) - Can be parameterized with beta or more complex distribution

What actions did insurers take following the financial crisis?

Increased liquidity, lowered risk - Almost doubled allocations to cash and short-term securities - Significantly reduced exposure to securitized assets Raised capital - Issued common and preferred equity - Issued unsecured debt, surplus notes, and subordinated debt - De-risked through changing asset allocations - Attempted to divest businesses

Define the following terms: - Incubation period - Infectious period - Generation time

Incubation period - Period between an individual's exposure to a virus and the onset of clinical symptoms Infectious period - Period during which an infected individual can transmit the virus to another person - May lag initial exposure: may or may not overlap with incubation period Generation time - Time between symptom onset in infected person #1 and symptom onset in person #2 - Short generation time and high R0 is a dangerous combination

State responses to indirect interest rate exposure

Indirect exposure - exposure to interest-sensitive liabilities 1. Cash flow matching - match A/L cash flows as closely as possible - Basic cash flow matching - e.g. match actual bond cash flows to liabilities * Drawback: limits investment strategy and risk-taking - Interest rate swaps - e.g. pay fixed/receive floating * Don't have to change underlying investments * No premium to pay (just an agreement) - Interest rate swaptions - option to enter a swap if rates cross a certain level * Better than a swap if willing to pay premium to hedge only downside risk 2. Redington's immunization: maintain A/L relationship after small interest rate changes a. PV of assets = PV of liabilities 2. Modified duration of assets = modified duration of liabilities 3. Asset convexity > liability convexity 3. Hedging using model points - match overall A/L with swaps only at certain key points on the yield curve

List at least 4 ways that insurers can improve customer satisfaction

Insurance customer satisfaction is much lower than personal banking and retail customer satisfaction! How to increase mid-market and millennial customer satisfaction: 1. Quick tutorial on life insurance with concise explanation of benefits 2. Individualize needs analysis and recommendations 3. Simpler application process (fewer questions and pre-filled info) 4. Real-time quotes, similar to auto insurance 5. Relevant quote alternatives for comparison 6. View of life insurance and related products (e.g. purchased by peers)

Describe the role of catastrophe risk management and dynamic hedging

Insurers have had success with more advanced risk management in 2 areas: 1. Catastrophe RM - top threat to P&C insurers - Key concern: rapidly rising exposures (property values) combined with increase in frequency of natural disasters (hurricanes, etc.) - Man-made events are expected to increase in frequency - Tools are more sophisticated: meteorological, seismological, statistical 2. Dynamic hedging for risks in retirement products (interest rate, ALM, and disintermediation risk) - Policyholder-based - policyholder behavior risk with options - Capital market-based - risk arising from investment guarantees

List risks related to operational risk

Internal processes or external catastrophes produce unexpected losses - Human capital risk (can't attract/retain qualified personnel) - Operational risk (bad controls, systems, 3rd parties, or physical damage) - Start-up risk (lack of discipline, control, data, volume) - Distribution channels risk (ineffective agents, etc.) - Fraud risk (embezzlement, etc.) - Strategic risk (bad planning, resource allocation, or adaptation) - Competitive risk (lost sales, profit) - Stress testing/scenarios (scenario modeling problems) - Business continuation (disasters, disruptions) - Acquisition of business (problems with pricing, sales, reserves, etc.)

Describe reverse stress testing

Involves asking what possible scenarios would bring a company to a specific solvency or loss level (e.g. what equity drop would lead to a $100M loss for our company)

Describe how the mitigation of counterparty risk is a double-edged sword

It will reduce overall risks but could potentially allow financial markets to develop too quickly or reach a dangerous size

State the challenges and practical issues with migration models

It's best to smooth out an entire economic cycle - Smooths out extreme years (e.g. 2009) - Requires Markov chain assumption, which may not be valid Simple approach: N-Year Default Probability = N * 1-Year Default Probability - Accuracy falls for large N (e.g. implies 100% default probability over 10 years) Ratings withdrawals: some firms will have their ratings withdrawn - Solution: reflect unrated probability Practical issues: - Low granularity (many firms per rating) - Cost to obtain ratings (benefits > costs?) - Inconsistencies by agency (S&P ≠ Moody's, etc.)

State the steps and formula for Kendall's tau correlation

Kendall's tau measures observations' tendency to move together

Describe the advantages and challenges of using more complex mechanistic models

Key advantage: provides more detail to policymakers - Can study the effect of social distancing for high-risk subgroups - Create subgroups of special importance * E.g. study healthcare workers, who are vital to other subgroups - Capture social relationships between subgroups (students, workers, retirees) Potential challenge: social relationships may vary by region - E.g. elderly may be more likely to live with younger generations in some regions - Region-specific parameters are likely required

Describe Maximum PFE

The highest/peak PFE value over time

Define valuation manual

The manual detailing statutory reserve requirements, including PBR, adopted by the NAIC or as amended thereafter

Compare and contrast risk retention groups and risk purchasing groups

Key similarities with RPGs: - Authorized under the same legislation (LRRA) - Increased underwriting capacity by reducing multi-state regulation Key differences with RPGs: - RRGs are licensed insurers, while RPGs are not (usually an LLC) - RRG insureds are encouraged to put their own capital at risk RRG requirements that make it a type of pure captive: - Only has to be licensed in domicile (other states can't require examinations, etc.) - Must have homogeneous risk (but insured not required to be an "industrial") - All insureds must be owners; all owners must be insureds

State the equation relating loss given default and recovery rates

LGD = 1 - Recovery Rate

Describe the range of institutions taking significant counterparty risk

Large derivatives player - Typically a large bank - Will have a vast number of OTC derivatives trades on their books - Will trade with each other and have many other clients - Coverage of all or many different asset classes (interest rate, foreign exchange, equity, commodities, credit derivatives) Medium derivatives player - Typically a smaller bank or financial institution such as a hedge fund or pension fund - Will have many OTC derivatives trades on their books - Will trade with a relatively large number of clients - Will cover several asset classes although may not be active in all of them Small derivatives player - Typically a large corporation with significant derivatives requirements or a small financial institution - Will have a few OTC derivatives trades on their books - Will trade with potentially only a few different clients - May be specialized to a single asset class

Describe the properties of the Weibull GEV

Less useful for large loss modeling since it's bounded at -1 / ξ - Shape parameter ξ < 0 - Distributions in the Weibull MDA include Weibull itself and the beta distribution

List the limitations of correlations and how copulas and multivariate methods address these problems

Limitations of correlations: - Introduces unacceptable distortion where the risks are not normally distributed - Assumes constant relationship and doesn't reflect stress conditions (e.g. in tail) - May significantly understate required capital Copulas and multivariate methods address the above problems - Copulas allow for non-linear relationships among many risks - Multivariate methods: * Define risk factor relationships explicitly * Simultaneously assess multiple product groups, geographic locations, etc. * Implicitly capture diversification

Describe risk adjustment validation considerations before aggregation with respect to the following items: - Main elements to validate - Statistical distribution for each risk

Main elements to validate: - Validations required for VaR, CTE, or CoC method * Form of statistical distributions used * Parameters of the statistical distributions chosen * Sufficiency of the number of simulations * Quantile chosen for the risk adjustment * Basis for entity's risk preference - Additional validations for the CoC method: * Scenarios' impact on remaining contract cash flows * Projected capital amounts * CoC rate used Validation of the statistical distribution for each risk: - Assess fit of chosen distribution - Verify that enough simulations were run

List the goals, uses, and restrictions under risk budgeting and capital allocation

Main goal: allocate total risk budget for entire company down to each area - Areas include lines, products, geographical areas, etc. - Budgeting process: managers decide in which areas to accept risk Uses: - Decision tool for each levels require more/less capital - Allocate responsibility and limits to managers (create disincentives) - Compare with risk actually taken - Manage risk appetite Must understand restrictions on capital: 1. Fungibility of capital - ability to absorb losses, regardless of where losses occur 2. Transferability of capital - BU's ability to transfer assets/liabilities to other areas (should take into account the time and costs of the transfer)

State the responses to liquidity risk

Main response: actively monitor liquidity needs - Scenario analysis - Consider limitations on transferring liquidity within an organization - Incentivize employees to manage liquidity Managing market liquidity risk: - Investment strategy consistent with liability cash flows - Swaps - Maintain a cushion of high-quality, liquid assets - Understand liability options (early withdrawal, etc.) Managing funding liquidity risk: - Diversification in the term and source of funding - Manage credit risk (linked to liquidity risk) - Gauge ability to raise capital, even if not presently needed - Contingency funding plan to provide liquidity in times of stress * LOCs, back-up liquidity, ability to issue new products

State the responses to systemic risk

Managing common counterparty risk - Simple solution: use a range of counterparties - Extreme solution: use exchange-traded derivatives that pool counterparties Regulators have the most responsibility for managing systemic feedback risk - Circuit breakers - halt stock trading after extreme market movements - Solvency regulation that avoids pro-cyclicality - Capital buffers that build up when times are good Managing systemic liquidity risk with regulation: - Government-provided funding for banks - Relax monetary policy by lowering interest rates Regulators can also separate businesses to mitigate spreading of catastrophic losses

Explain the margins used in managing future counterparty risk

Margin = deposits by exchange members that help offset losses caused by members' insolvency 1. Initial margin = the value of assets transferred to a margin account once a contract is opened - Percent of contract (higher volatility contract requires a higher percentage) - Set lower for spread positions than naked positions - Marking-to-market - process of adjusting margins based on the current cost to close the contract 2. Maintenance margin: if margin falls below maintenance, must pay variation margin 3. Variation margin = amount paid to bring the margin back up to the initial margin

Define liquidity, correlation, and funding risk, in the context of derivatives

Market liquidity risk - Cost of hedging larger derivative > cost of hedging smaller derivative Basis or correlation risk - Arises with imperfect hedges: derivative and hedge item are not perfectly correlated Investing and funding risk - Driven by cashflow mismatches and/or collateral payment requirements - Can be forecasted, but fluctuates with changes in the market

Define model user

The person who is responsible for running the model and using it to calculate results. This is a role that requires adequate knowledge of both the underlying business and of the model being used to analyze that business

Describe how to estimate Kendall's tau from a data sample

The sample estimate of τ = average sign over all possible distinct values of i and j < 1, where there are n (n - 1) / 2 total distinct pairs possible

State the challenges faced by risk retention groups

May become difficult to operate outside the commercial marketplace in the future! - US Government Accountability Office (GAO) has recommended: * More uniform regulatory requirements across all states * Corporate governance standards to establish insureds' authority over management * Expansion of regulatory and audit requirements for RRGs - GAO cited specific issues: * RRGs are allowed to use GAAP instead of US Stat when filing reports * RRG board members and executives can be non-insureds - The NAIC formed a working group (NRRA) to represent RRGs * Recommended different regulation for RRGs that operate as captives vs. commercial insurers * Pointed out that RRGs fail at a higher rate when regulated as commercial insurers

Define model classification scheme

Methodical identification of models according to model characteristics. For example, models may be classified according to risk rating (high, medium, low), model type (cash flow, regression), model purpose (valuation, pricing, RBC), accounting method (GAAP, stat, tax, IFRS), product type (life, annuity, health), etc.

Describe the modeling for expected corporate bond returns

Methods for adjusting for credit default risk: 1. Nominal spread = gross bond yield - risk-free rate - Easy and quick to calculate - Ignores important features but can be a reasonable approximation 2. Static spread = constant spread added to risk-free rates such that the PVCFs = market price - Reflects entire term structure (unlike the nominal spread) 3. Option-adjusted spread (OAS) - like static spread but solved-for stochastically - Better for securities with optionality Why historical spreads tend to be larger than experience: - Credit beta = compensation for higher volatility compared to risk-free rates - Investor loss aversion: market could have gotten worse (just didn't) - Includes other risks: liquidity, tax effects, etc.

List ways of combining internal and external loss data

Methods for frequency or severity: - Use data directly (usually not appropriate) - Scale the data for size through a normalization algorithm (usually not appropriate) - Scale the data for size through proportionality (usually best) * Assumes the same relative risk control profile across all categories Other methods for severity (both are bad) - Manually select losses from a database (loses probability value) - Use internal data for the body and external data for the tail (manipulation risk)

Describe environmental risk

Model similarly for other low frequency classes Key difference: more predictable than earthquakes, etc. - May occur from non-natural events - Firms with poor risk management are more likely to have an environmental event - Can reflect predictability in rates

Describe collateralized debt obligations

Modern CDOs are synthetic securitizations: special purpose vehicles (SPVs) - Portfolios of bonds, mortgages, etc. are divided into tranches - Investors purchase tranches and receive a return (riskier tranche = higher return) * Equity tranche: highest risk (absorbs defaults first) * Senior tranche: lowest risk (absorbs defaults last) * Mezzanine tranche: between equity and senior - Attachment point - aggregate loss at which a tranche begins absorbing defaults - Detatchment point - point at which the tranche is exhausted by default losses - SPV modeling considerations: * Choice of model and parameters is crucial for determining attachment points - Dependency, diversification, and concentration in underlying credit risks - Tail dependence

Describe how to apply the actuarial approach to operational risk management (ORM)

Monte Carlo simulation is often used Theoretical requirements for the aggregate distribution: 1. X's should be i.i.d. 2. Frequency and severity should be independent

Describe current and future captive regulation

NAIC Financial Analysis WG - Reviewing XXX/AXXX business ceded to captives - Will provide solvency recommendations Shadow Insurance = deals with affiliates to lower reserves and regulatory requirements - Investigated by NY DFS - 4 areas of concern: 1. Conditional LOCs 2. Two-step transactions 3. Hollow assets 4. Naked Parental guarantees The future: - Captive regulation will likely increase - Less reliance on captives will result - Increased transparency at the group level

Discuss the risk blindness for Solvency II and US RBC capital requirements

Neither protects against entire insurance industry collapse (worst case) - Solvency II is "blind" to risk beyond the VaR percentile * Worse for fat-tailed risks - RBC is worse: no specified risk metric or confidence level * Possible solution: CTE ** Less "blind" than VaR ** CTE is also sub-additive ** Requires smaller confidence levels for the same number of scenarios * Disadvantage of CTE: harder to convey results

Define antiselection

Not disclosing information known by the insurance applicant in order to get life insurance coverage at a lower premium rate than if the information had been revealed on the application

Describe the advantages and disadvantages of one-year vs. multi-year time horizons for EC

One-year method: EC = capital required to survive a shock within the next year - Advantage: avoids complex modeling - Disadvantages: * Ignores impact of cash flow changes beyond the next year * Cannot be used to calculate VaR * Ignores future responses to shock event (management actions, underwriting practices, hedging) - More popular than multi-year Multi-Year method EC is based on a stochastic scenario-based model - Advantage: deeper understanding of the long-term risk exposures - Disadvantages: * May ignore management actions (but not as much as the 1-yr approach) * Difficult to capture full range of possible regulatory and management actions

List 5 data "questions" early in a pandemic

Over time, modelers gain more data on the following (but know very little at first): - Virus transmission - how easily it is spread and under what conditions - Average length of the infectious period and how it varies across individuals - Virus's ability to survive outside of host (e.g. on surfaces) - Infection fatality rate (IFR) - Other factors contributing to IFR: age, gender, socioeconomics, comorbidities - Variation in symptoms - whether some people show little or no symptoms - Immunity in survivors - whether it exists and how long it lasts - Regional factors that influence transmission/fatality rate - Weather's impact on transmission, if any

Define model validation

The practice of performing an independent challenge and thorough assessment of the suitability and adequacy of the model based on tests carried out across multiple dimensions including design, data, assumptions, results, and governance

What is the credit valuation adjustment (CVA)?

The reduction in price due to counterparty risk

Summarize the "surprises" when comparing pre-COVID-19 scenarios to reality

P&C difficulties related to business interruption property trigger provided the biggest surprise relative to pre-COVID-19 scenario analysis - Policy wordings and exclusions are being challenged, especially for business interruption coverage Public perception of the insurance industry - Uncertainty surrounding the interpretation of some less clear exclusions will likely leave a trail of litigation and lead to potential reputational damage to the insurance industry Regulatory restrictions / rapid regulatory change - Regulators have been far quicker to impose capital management restrictions (e.g. on banks and share buybacks) Impact on the real economy (because of extraordinary monetary/financial policy actions that were taken) Diversification assumptions challenged because of the global impact of COVID-19

Describe PCA-based approaches

PCA-based approaches can identify the main drivers of interest rate changes - Useful when there are many possible factors - Good for interest rates, which have a term structure (many shapes, movements) - Examples of principal components: yield curve level, steepness, etc. - Uses eigenvalues of covariance matrices to identify importance * Largest eigenvalue = most important component * 2nd-largest eigenvalue = 2nd most important (etc.) * Elements of eigenvectors show how points on the yield curve react - Choose a number of principal components that explains a sufficiently high proportion of past returns

Describe how to calculate Pearson's rho, as well as its limitations

Pearson's ρ = linear correlation coefficient - Key drawback: assumes linear relationship between X and Y - Misleading when X and Y are highly dependent but not linearly dependent

Explain the importance of understanding policyholder behavior in terms of how it impacts an insurance operation

Policyholder behavior risk: non-diversifiable and non-hedgeable - Elective, unlike mortality risk! Affects many areas of insurance operations 1. Product design: early mistakes were made - VA GMDB dollar-for-dollar withdrawals, LTCI lapse rates were set too high 2. Reserves and capital: premium paid, investment elections, fund transfers, etc. 3. ALM: liability durations change as policyholders react to changing interest rates 4. ERM: tail risk and other operational risks

Describe the properties of the Gumbel GEV

Popular for modeling large losses with less fat tails than the Fréchet GEV - Shape parameter ξ = 0 - A variety of distributions are in the Gumbel MDA, including Gumbel itself * Thin-tailed: normal and exponential * Fat-tailed: gamma, lognormal, and inverse Gaussian - Distributions in the Gumbel MDA have an infinite number of moments

Describe the positive and negative MtM cases

Positive MtM - When a counterparty defaults, they will be unable to make future commitments and hence an institution will have a claim on the positive MtM at the time of default - The amount of this MtM less any recovery value will represent the loss due to the default - Exposure is positive Negative MtM - In this case, an institution owes its counterparty through negative MtM and is still legally obliged to settle this amount. From a valuation perspective, this position is essentially unchanged - An institution does not gain or lose from their counterparty's default in this case - Exposure is floored at zero

List risks related to legal risk

Potential for losses/disruptions due to unenforceable contracts, lawsuits, or adverse judgements - Political risk (government policy changes insurance market, investments) - Regulatory risk (changes in laws/taxes or slow product approval) - Legal risk (non-conformance with laws, standards, etc.) - Distribution channels risk (e.g. agent breaks law) - Sovereign risk (foreign laws change) - Fraud risk (external)

List risks related to reputation risk

Potential negative impact on customer base or litigation resulting from negative publicity, whether true or not - Reputation risk (rating downgrade, etc.) - Quality/service delivery risk (bad business practices, management decisions) - Fraud risk (external)

Describe how predictive analytics have been applied to modeling policyholder behavior

Predictive analytics is an improvement over traditional actuarial models - Can operate on variables about the individual (age, education, etc.) - Can predict behaviors triggered by environmental changes Limitations of predictive analytics: - Fails to capture causal influences and non-quantitative factors - Fails to model decision shortcuts, psychological biases, etc. - Doesn't capture life events (e.g. marriage) that trigger a life insurance purchase

Describe the importance of modeling epidemics

Primary purpose: guide public policy decisions - Assess impact of strategic options: * Closures of schools, bars, and restaurants * Travel restrictions and stay-at-home orders * Social distancing - Assess demand for hospital capacity under different scenarios - Output can be married with economic models

List the types of sponsored captives and related programs

Primary types of sponsored captives - Basic form: parent sponsor owns a captive reinsuring similar subsidaries (insureds) - Cell captive - participant assets held in separate accounts * Rental captive - instead of paying in capital, insureds pay an access fee Other uses of sponsored captives - Incubators for small insureds - Off balance sheet risk financing for large insurers - Agency captives and producer-owned reinsurance companies - Special purpose vehicles (SPVs) - may be considered captives Captive facility - a commercial (re)insurer that provides captive programs - Not a sponsored captive, but has elements in common with a sponsored captive - Participants risk their own capital and manage their own risks - Participants do not have ownership or control of the facility

Define pro rata reinsurance and excess of loss reinsurance

Pro rata - proportional risk sharing between the cedant and reinsurer - Quota share (AKA coinsurance) - a fixed percentage of risk is ceded - Surplus share - cedant cedes a percentage of "surplus" risk above a fixed dollar level retained by the cedant Excess of loss reinsurance - non-proportional risk sharing - Reinsurance only applies if losses exceed cedant's retention level - Higher retention means lower reinsurance premiums - Can take one of three forms: 1. Aggregate (AKA stop loss) - applies if aggregate losses exceed retention during an effective period 2. Per occurrence - like aggregate but only applies for a specific covered event (e.g. natural disaster) 3. Per risk excess of loss - applies at the individual risk level (not aggregate)

Identify the types of parametric credit scoring models

Probit and Logit models (GLMs) - Independent variables = financial ratios, etc. - Dependent variables takes a value of 0 to 1 (default probability) Altman's Z-Score (a form of discriminant analysis) - If Z > 2.99: safe! - If Z < 1.80: there is a risk of distress

Describe sources of uncertainty that can and cannot be diversified away

Process risk can be diversified away by writing more business - Actual experience converges to true population parameters Other risks cannot be diversified with volume - Parameter risk (major topic of ERM-104) * Sampling risk - differences between population and sample used * Data bias - uncertainty in adjustments from experience to exposure period (e.g. claim trends) - Model misspecification - wrong model or distribution used * Insufficient parameter identification - failing to reflect correlations or parameter variation * Actuarial model risk - risk that entire framework is wrong

State the COSO definition of ERM

Process, effected by an entity's Board of Directors, management, and other personnel; applied in strategy setting and across the enterprise; designed to identify potential events that may affect the entity; and manage risk to be within its risk appetite, to provide reasonable assurance regarding the achievement of entity objectives

State the pros of a centralized vs. decentralized model governance framework

Pros for centralized structure: * Resources committed to model governance activities may be less likely to be pulled away for competing priorities - A centralized unit can be staffed with individuals possessing unique skills best suited for model governance roles and responsibilities Pros for decentralized structure: - Expertise with the existing models could reside with individuals within the functional areas, in consideration of the fact that knowledge transfer to a centralized unit involves a time and resource commitment Both approaches can be successful; the decision on how best to address model governance needs will depend on the company's own situation

State the risks to SGUL writers

Protection-oriented SGUL risks: - Substantial reserve (surplus) strain - Low profit margins and few levers to manage profits - Interest rate risk - Lapse risk (low lapses are bad, effects from secondary market) - Mortality risk from old age exposure - Minimum premium assumptions may be wrong Additional accumulation-oriented SGUL risks: - Policyholder behavior isk ey - Even more lapse risk - Shadow fund is more likely to be ITM if expected premium and lapses are wrong

Explain the differences between pure captives vs. sponsored captives

Pure captive - owned entirely by the insureds (single or group) - Often, will only write risk that is 100% related to its insureds (but not always) * Not pure if the captive writes unrelated risk for the purpose of making a profit - Single owner does not necessarily make a captive "pure" * Example: Allstate was once a commercial insurer wholly owned by Sears Sponsored captive - owned and controlled by parties unrelated to the insureds - Sponsor = owner of the captive - Participants = actual insureds - Do not necessarily require insureds to risk capital (e.g. rental captives) - Do not necessarily pool the risks of the participants (e.g. cell captives)

Describe hedging with instruments other than futures and forwards

Put options can floor the price the security - Put payoff = max(0, K - Xt) - Differences with futures/forwards: * Put options = insurance (limit downside only) * Put options require an upfront premium * Options only address asset risk: may not address all ALM risks Credit default swaps (CDS) - OTC derivatives that hedge bond default risk Out-performance option - pays off if the returns on one asset exceed those on another by more than a certain amount - Useful in pensions and insurance to protect investment returns relative to interest-sensitive liabilities

Describe risk metrics used to set risk appetite

Quantitative measures (must use stochastic or stress testing for tail risk) - VaR: value will be above/below X in Y% of scenarios * Capital/equity at risk: used for solvency, ratings, economic capital * Earnings at risk (EaR): GAAP/IFRS * MCEV (within X% of some target) - TVaR (CTE) is more appropriate for skewed variables (tail risk) * Average of tail values beyond VaR Qualitative measures - Credit ratings (e.g. maintain financial strength) - Risk preference (e.g. avoid regions with high cat risk) - Franchise value (e.g. reputation risk) Enterprise risk tolerance statement - specific statements on each of the above

How can RAROC and MCEV be integrated into new business budgeting?

RAROC and MCEV measure risk better than traditional measures like EV, profit margin, and combined ratios - RAROC = risk-adjusted return on capital - MCEV considers the following: * Cost of non-hedgeable risks (CNHR) * Time value of options and guarantees offered in the insurance contracts (TVOG) - Frictional cost of capital explicitly (FC) Shareholder value is created when a business unit generates: - RAROC > hurdle rate - MCEV earnings > target MCEV

Define risk-neutral (q-measure) vs. real-world (p-measure)

RN discounts risk-adjusted cash flows at risk-free rates - Assumptions: no arbitrage, derivatives are reproducible with market securities - Best for market-consistent pricing when market information is available RW discounts projected cash flows at RDR - More widely used for insurer EC * RN inputs are often not available - Advantage: easier to interpret distribution of results for VaR, etc. - Drawback: RDR can be difficult to estimate

Define model risk

The risk of adverse consequences from decisions made as a result of a model that does not adequately represent that which is being modeled

Compare the validation requirements for real-world ESGs vs. market-consistent ESGs

RW validation is more varied than RN and falls into 3 categories: 1. Point-in-time: verify the model's characteristics adequately match the market at t = 0 2. In-sample: checks how well the ESG reflects stylized facts 3. Out-of-sample validation or back-testing: compares model output at T to T + t Market-consistent (RN) scenario validation is more defined: - Market consistency: model prices should match current market prices - RN validation: * Expected returns for each asset class = risk-free rates * Can use a martingale or "1 = 1" test ** E[next value | current and previous values] = current value

State a situation where real-world and risk-neutral scenarios might be used together

Real-world and risk-neutral scenarios will often be applied together in applications that combine risk management metrics with pricing

Describe the demand and supply effects on actual futures prices

Reasons why the actual price of a futures ≠ expected futures price: - Normal backwardation - when the actual futures price is less than expected * High supply (demand for short positions > demand for long positions) * Occurs if income + convenience yield > costs - Contango - oppositive of normal backwardation * High demand (e.g. investors trying to gain exposure to a particular commodity) * Can occur if storage costs are very high Relationship between futures price and time until expiry: - Normal market - futures prices increase with expiry date - Inverted market - futures prices decrease with expiry date * Seasonal effects: investors expect higher supply of the asset in the future

Define reinsurance and the essential elements of a reinsurance contract

Reinsurance = "insurance of insurance companies" - Cedant pays a premium to another insurer (the reinsurer) - Reinsurer indemnifies the cedant - Contract is between the cedant and reinsurer * Cedant must honor underlying policies even if reinsurer defaults Essential elements of a reinsurance contract: 1. At least 2 insurance companies: cedant + reinsurer(s) 2. One or more underlying policies issued by the cedant to policyholders 3. Transfer of some insurance risk from the cedant to the reinsurer(s) Retrocession - reinsurer (retrocedant) cedes business to another reinsurer (retrocessionaire)

Describe the goals of measuring and assessing operational risk

Requirements for key decision-makers to understand exposure to operational risk: - Total aggregate risk exposure for each business at the target tolerance level (N%) - Aggregate expected loss = probability-weighted annual aggregate loss - Aggregate unexpected loss (N%) = aggregate risk faced by the business - Unexpected loss (N%) = total aggregate risk exposure (N%) - aggregate expected loss - Total cost of risk = expected loss + cost of capital * unexpected loss Uses of operational risk models: - Regulatory and economic capital - Inform rating agencies

Describe the Aumann-Shapley Value method

Requires a simulation technique like the RMK conditional risk algorithm: 1. Run many simulations (e.g. 10,000) 2. Sort results from best to worst (worst = requires the most capital) 3. Choose an acceptable probability of ruin (e.g. 2.275%) 4. Simulated capital requirement = average results around the 97.725th percentile - Example: take average of the 1% around the percentile (i.e. scenarios 9723 to 9822) Advantages and other notes: - Allows for fractional "players" (portfolios), unlike the Shapley Value - Myers-Read is a special case of Aumann-Shapley - Captures additional risk added by a small increase in size

Briefly describe the residual risk effort matrix and inherent risk control matrix

Residual risk effort matrix: - Used by risk management (remember the beginning of ERM-147 emphasized risk managers focusing on residual risk) - Matrix used by risk managers to focus activities on mitigating the risk areas with the highest exposures - Horizontal axis (x-axis): Effort (Factor C) - Vertical axis (y-axis): Vulnerability, Exposure, and Impact (Factors A/B/D) Inherent risk control matrix: - Used by internal audit (remember the beginning of ERM-147 said the internal auditor will want to have an understanding of inherent risk) - Horizontal (x-axis): level of current risk management / mitigation effectiveness (Factor A) - Vertical axis (y-axis): level of inherent risk (Factors B/C)

State the responses to environmental risk

Responses are similar to other low frequency non-life insurance risks: - Reinsurance - Diversification - Risk modification Should also analyze claims to learn how to reduce future risk

List the responses to credit risk (own credit risk and counterparty credit risk)

Responses for own credit risk - ability to borrow, write new business, etc. 1. Managing capital structure 2. Managing the mix of business Responses for counterparty credit risk (and possibly own credit risk indirectly) 1. Underwriting 2. Due diligence 3. Credit insurance 4. Risk transfer: CDS, CDOs, and CLNs

List types of market risk and related risks

Risk associated with adverse movement of interest rates, forex rates, or equity prices - Interest rate risk (adverse yield curve shifts, disintermediation) - Basis risk (mismatched asset/liability yield or hedge and hedged item) - Reinvestment risk (falling yields for future cash flow investment) - Currency risk (USD rises, foreign currency falls) - Asset concentration risk (increased market risk from lack of diversification) - Asset/liability matching risk (losses from duration or cash flow mismatches) * Closely related to interest rate risk - Asset market value risk (realized losses from selling at inopportune time) - Political and sovereign risk (adverse change in domestic or foreign policy)

List types of credit and related risks

Risk that borrower or counterparty will not meet obligations - Business credit risk (reinsurer, etc. fails to perform) - Invested asset credit risk (asset defaults, asset rating downgrade) - Political risk (changes in government policy) - Concentration risk (increased credit risk from lack of diversification)

Define residual risk

Risk that remains in an organization, taking into account all mitigating actions that have been taken, within budget constraints, in order to optimally manage the risk Residual risk is defined as the function of the (1) factor impact and the (2) factor likelihood

Define credit risk

The risk of financial loss owing to counterparty failure to perform its contractual obligations

Summarize the key differences between real-world and risk-neutral scenarios

Risk-neutral scenarios are typically used for pricing (in particular, when no closed-form solution exists or when cash flows depend on stochastic financial variables), and calibrated using pricing data Real-world scenarios are typically used for risk management, and are calibrated using historical data Additional differences include: - The difference between risk-neutral scenarios and real-world scenarios is not the individual scenarios themselves; it is the probability of those scenarios occurring! - Risk-neutral scenarios embed the risk premiums necessary to get appropriate market prices - A risk-neutral scenario set weights the bad scenarios more than a real-world scenario set - Risk-neutral scenarios are often difficult to interpret and look unreasonable if viewed relative to historical interest rates or returns; the primary criterion is the ability to fit today's market prices and address pricing purposes/valuing cash flows

Compare and contrast the time horizons for Solvency II and US RBC capital requirements

SCR uses a 1-year horizon - Not designed to view business on a multi-year/going-concern basis - May miss risks that develop over time RBC does not specify a time horizon at all - "Point-in-time" assessment based on a retrospective view - Moving toward a longer-term, liability run-off approach * Problems: more model uncertainty ** Requires very long-term assumptions ** Difficult to reflect company actions

Describe recent trends in life insurance sales

Sales have been declining and are at a 50-year low - Sharpest decline: middle-income and millennial market - Average US household is underinsured by 400K Average face amount has risen - Agents now focus on the higher net worth market Distribution channels are not designed to serve the mid-market

Describe scenario analysis

Scenario analysis is a process of asking "what if" by defining complex scenarios or trajectories based on the current, specific state of the world and involves: - Analyzing the impact on variables of interest such as solvency, liquidity, or earnings and reputation - Evaluating these against the institution's risk appetite - Ultimately, taking direct actions or developing contingency plans to increase reslience

Why is executing an ALM strategy based on segmentation bad?

Segmentation is when we explicitly back a block of liabilities with certain assets - Each line of business receives its own asset portfolio - This is not recommended, because it does not consider offsetting risk across different lines of businesses, and may end up increasing the company's overall risk exposure! - Best practice is to aggregate risk exposures and manage them at the total company level

Describe considerations for setting risk appetite for catastrophic risk

Set risk appetite for climate change risk, pandemics, natural disasters, etc. - Strongly linked to liquidity risk - Also consider solvency impact Map to net amount at risk limits - NAR = Death Benefit - Reserve Example: don't lose more than 20% of GAAP equity if the 1918 Spanish flu happens again

Describe characteristics of companies with strong ERM culture

Set the tone at the top (Board and senior management) - Embed risk awareness throughout the organization - Integrated risk management; responsibility with senior management - Board / senior management informed and understand key risks - Incentive compensation tied to risk management objectives Establish and clearly communicate risk management objectives - Supports overall corporate goals and stakeholders' expectations - BU management required to implement risk mangement practices Define roles and responsibilities - Segregate decision-makers from those who monitor risk - CRO coordinates risk management and establishes/facilitates framework - CEO responsible for executing corporate strategy - Board provides active oversight and challenges management - BU managers are directly responsible for managing risk Strategic decision-making process - Based on risk-adjusted returns/metrics consistent with the corporate risk profile - Integrate with financial planning and budgeting process - Management can demonstrate how risk/return decisions improve insurer value

Describe the Shapley Value (game theory) method

Shapley Value = the average of the "first in", "last in", and all "intermediate ins" - First in = pro-rata VaR - Last in = allocated VaR using the Discrete Marginal Contribution method - Intermediate in = average of all "ins" other than first and last Advantage: no scaling needed since portfolio values add up to total EC Disadvantages: - The number of "players" (portfolios" needs to be a whole number - Very calculation-intensive

Describe the Myers-Read allocation method

Shareholders hold a put option on the insurer's losses in excess of capital - Assume losses are normal and values the put option using Black-Scholes - If risk exposure increases, put option value increases - If capital increases, put option value decreases EC for a portfolio = [put option cost of the last unit added] * [expected loss] - Assumes put option cost is a constant percentage of expected losses (constant ruin) Advantage over other marginal methods: total capital = sum of the marginal increments - Total EC = sum of all portfolio EC amounts

List the responsibilities of an independent credit risk management function

Should be able to analyze derivatives and have the following responsibilities: 1. Approve credit exposure measurement standards 2. Set/monitor credit limits 3. Review credits and concentrations of credit risk internally before engaging with counterparty 4. Review and monitor risk reduction arrangements

Discuss how management should establish authority in derivatives transactions

Should designate who is authorized to commit their institutions to derivatives transactions - Management may delegate authority to certain individuals - Management may limit authority to certain types of transactions - Authority to commit should be communicated - There is no substitute for appropriate internal controls

Discuss credit migration models

Shows the probabilities of moving from current rating to another rating (or default) - ratings are like risk classes, default is like death, and unrated status is like a lapse - Firms with higher credit ratings have lower default probabilities - Relationship between credit ratings and default probabilities is nonlinear

Compare and contrast captives and commercial insurers

Similarities between captives and commercial insurers: - Captives are licensed insurance companies * Have a domicile state but can operate in other states * Can issue policies, buy reinsurance, and receive reinsurance commissions - Captives are regulated as insurers under state laws * Special purpose laws - laws developed specifically for captive insurers Differences between captives and commercial insurers: - Captives work outside the commercial market - Captives give risk owners more flexibility to achieve their goals - Captives may be riskier: have less capital and no state guaranty access - Captive insureds actively manage the company (unlike mutual policyholders)

Compare and contrast risk purchasing groups and self insurance pools

Similarities between pools and RPGs: - More access to insurance, more control, and lower costs - Strictly risk financing mechanisms: they do not take risk - Not licensed insurers (pools are "qualified self-insurers") Differences with RPGs: - Pools facilitate each member's retention of risk - Pools assume claim payment administration

Describe non-life insurance risk (in terms of similarities and differences to life insurance risk)

Similarities with life insurance risks: - Same 4 risk categories: level, volatility, catastrophe, and trend - Has same general aspects: incidence and intensity (claim size) Differences with life insurance risks: - Intensity of non-life claims is much less certain - Trend risk may be more difficult to define (also less important usually) * Usually follows economic cycles: use scenario analysis 2 broad groups of non-life insurance risks: 1. High frequency claims (e.g. auto) - Produce a lot of data, so they are more straightforward to analyze - Establish IBNR reserves 2. Low frequency claims (e.g. hurricanes) - Require EVT, copulas, etc. - IBNR reserves may be possible for imminent events

List ways of addressing common issues when modeling severity

Simple raw empirical data methods rarely yield enough data points - 99.5% level would require observing 5 events in 1000 years Fitting common parametric distributions is better but rarely feasible Fitting truncated loss data (not collected from $0) - Generalized Pareto require too much subjectivity and tail may not converge to 0 - Use Modified MLE process to fit to known distribution (e.g. lognormal) FItting heavy-tailed truncated data - Option #1: use distribution with more than 2 parameters (LNG or Burr) - Option #2: Fit tail portion using ALEC - Key problems: overfitting and estimation error Non-homogeneous data violate the i.i.d. assumption, which will produce false results

Describe the pro-rata or linear marginal contributions approach

Simplest approach: use the standalone EC amounts to allocate - Doesn't penalize highly-correlated portfolios - Doesn't reward portfolios that increase the overall diversification effect

Which state of the spike-and-slab model is associated with statistical significance of the regression coefficient? What is the value of li when we are in this state of the model?

Slab model, which is when li = 1

Discuss the international regulatory trends toward EC

Solvency II (Europe) - Philosophy: the most well-managed insurers are most likely to remain solvent - Solvency capital requirement (SCR) = required capital set at 99.5% VaR * Involves internal models (very EC-like) - Minimum Capital Requirement (MCR) = absolute minimum level * Uses formula approach * Regulators take urgent action if capital < MCR US Risk-Based Capital (RBC) - Traditionally uses a formulaic approach - Moving towards PBA and internal models (C-3 Phase II) Canada has been using a PBA with internal models IAIS is moving toward a common total balance sheet approach

Describe the 3 pillars within Solvency II

Solvency II - capital standard for European insurers 1. Pillar I - Quantitative capital requirements for SCR and MCR - Solvency capital requirement (SCR) - key solvency control level based on either: * European Standard Formula (same for all firms) * Internal model - firm-specific model subject to IMAP - Minimum capital requirement (MCR) 2. Pillar II - Qualitative risk management and supervisory requirements 3. Pillar III - Supervisory reporting and disclosure

Discuss the differences in core philosophies for Solvency II and US RBC capital requirements

Solvency II: top-down (One-Year VaR 99.5) - Uses a well-specified metric (VaR), time horizon (1-yr), and confidence level (99.5%) - Allows insurer-specific risk models (if meet qualifications) - Multivariate VaR reflects risk dependency structure US RBC: bottom-up (mix of factor- and principles-based) - No specified time horizon, risk metric, or confidence level - Components considered in isolation (no dependency structure) - Some factors moving toward PBA: C-3 Phase II, etc. - NAIC SMI recommends other updates (e.g. explicit cat risk charge for P&C)

Compare and contrast using ALM as a risk mitigation exercise vs. using ALM as a part of a strategic decision-making framework

Some companies execute ALM as a risk mitigation exercise - In this case, the goal is simply to track the risk exposure of the firm's surplus and make sure it is within specified risk limits - Example: make sure that asset duration is always within 1% of liability duration Other firms integrate ALM within a broader ERM and strategic decision-making framework - The goal is to achieve a certain financial objective, subject to risk tolerances and constraints - Example: asset portfolio must return 7% annually, while still being within 2% of the liability duration

Define Spearman sample correlation

Spearman correlation = Pearson correlation of X and Y ranks 1. Rank X's (1 = largest, 2 = 2nd largest, etc.) 2. Rank Y's (1 = largest, 2 = 2nd largest, etc.) 3. Calculate correlation of the ranks

Describe how to estimate Spearman's rho from a data sample

Spearman's ρ can be estimated from a discrete data sample of n observations of X and Y

Compare the two primary types of models for projecting outbreaks

Statistical models use correlations or patterns in data to forecast virus propagation - Example: IHME COVID-19 model - Common approach: focus on time series of virus-related deaths by region - Can develop a curve for each region Mechanistic models model dynamic processes underlying the virus propagation - Includes susceptible-infected-recovered (SIR) models * Projects population shifts through subgroups: susceptible -> infected -> recovered/dead * Example: Imperial College of London's COVID-19 model - Range from very simple to very complex * Simplest models assume equal change of infection for all * Complex models have more granular subgroups; some can even simulate individuals - Includes compartmental models (AKA "population" or "metapopulation" models) * Can be deterministic or stochastic and operate in discrete or continuous time

List the steps and advantages/disadvantages of the Projections Scenarios Approach for EC

Steps: 1. Create a scenario for each risk factor 2. At each time step, calculate the value of the assets and liabilities - Use more time steps if the liabilities are path-dependent - Develop the P&L and balance sheet at each time step 3. EC = net present value of the P&L results Advantages: - Captures a wider range of adverse risk scenarios - Good for relatively independent risk factors like longevity and market risks - Captures dynamic interactions and risk management strategies - Highly flexible for developing specific, realistic adverse scenarios Disadvantage: - May require nested stochastic techniques

Define stochastic analysis vs. stress tests

Stochastic required for VaR, TVaR, etc. - Requires known probability distribution Stress testing does not require a known probability distribution - Both must identify scenarios with large losses - Both are sensitive to subjective assumptions and judgement

What is the difference between strategic asset allocation and tactical asset allocation?

Strategic asset allocation (SAA) - Used to determine a long-term policy portfolio - Reflects the desired systematic risk exposure Tactical asset allocation (TAA) - Specifies the allowable deviation from SAA to chase α - Takes advantage of short-term market opportunities - Example of return objectives: 1. Earn minimum return required to fund statutory reserves 2. Earn enhanced margin - competitive return and a reasonable profit 3. Earn higher return on surplus account using riskier assets

State the liquidity modeling considerations

Stress and scenario testing (most common) 1. Rising interest rates 2. Ratings downgrade 3. Large operational loss 4. Loss of control over a key distribution channel 5. Impaired capital markets 6. Insurers only: - Large insurance claim from a single or related events - Sudden termination of a large reinsurance contract Other liquidity modeling considerations: - Allow for limits on transfers of assets - Model asset streams separately: do not rely on total asset returns - Reflect interactions between liquidity and other risks - Model stochastically if enough data is available - Correlations can change (increase in bad scenarios)

What is the most important resource to reduce the risk of nondisclosure?

Telephone interviews!

Discuss sidecars, including benefits to cedants

Temporary reinsurance vehicle that shares premiums and losses exclusively with an insurer - Financed by debt and/or equity issued to investors in capital markets - Usually shares cat risk pro rata - Investors' proceeds are held in a trust and used to pay losses on ceded insurance policies Benefits to cedants: - Additional source of reinsurance when the market is limited (e.g. after a natural disaster) - Cedants can write more business even if traditional reinsurance rates are high

What is the relationship between the efficient frontier and the capital market line?

The CML is tanget to the efficient frontier and equals Rf if risk = 0 Tangent point = theoretical optimal portfolio (optimal risk/reward combination)

State the First Fundamental Theorem of Asset Pricing

The First Fundamental Theorem of Asset Pricing: Price must be given as the expected discounted future cash flows, where the expectation is taken with respect to a risk-neutral measure

Describe the relationship between the COSO framework and the Three Lines of Defense Model

The Framework: COSO Internal Control - Integrated Framework - Provides a structure to consider risk and control * Outlines 5 sections and 17 principles - Assigns 1 section to senior management, but silent on others The Model: Three Lines of Defense Model - Clarifies roles and duties needed for the Framework - Assigns "lines of defense" to the other four sections

Review the SOA's stylized SIR Model

The SOA developed a stylized SIR (compartmental) model for illustrative purposes: - Projects in discrete 1-day steps - User enters a time 0 population state (e.g. 1 out of 100K people infected) - Can assume a homogeneous population (equal risk) or 2 risk-specific subgroups - R0 = product of 3 factors: 1. Transmission probability if infected person comes into contact with uninfected person 2. Number of social contacts per person each day 3. Average duration of infection (in days) Example: suppose the 3 factors are 1%, 30, and 10, respectively: R0 = 3 people - Social distancing can vary to none to some and also revert to none over time

Describe considerations for the capital amount chosen for the IFRS 17 risk adjustment under the cost of capital method

The amount of capital held depends on: - Level of security desired - Probability that unfavorable cash flows will consume capital - Level of risk aversion Insurance company capital exists to provide security against financial distress - Increases probability of solvency, which protects policyholders - As an insurer's scale increases, risk capital increases at a decreasing rate (diversification) - Many jurisdictions have minimum capital requirements - Capital used for risk adjustment purposes is more akin to EC

Describe expected positive exposure (PFE)

The average expected exposure through time

What is the midprice?

The average of the bid and ask quotes

Describe effective expected positive exposure (EEPE)

The average of the effective expected exposure (EEE)

Describe considerations for the scope and frequency of model validation

The breadth/depth of review and testing depends on a number of factors, including: - Riskiness of the model, based on the model risk ranking - Frequency of external audits and regulatory reviews - The trigger of a model review upon failure of one or more model controls - Model characteristics: * Model type (e.g. cash flow, regression) * Model purpose (e.g. valuation, pricing, RBC) * Accounting method (e.g. GAAP, stat, tax, IFRS) * Product type (e.g. life, annuity, health)

Define model validation scope

The depth and breadth of reviews and testing performed on various areas as part of the validation exercise

Describe expected credit exposure (ECE)

The expected value of the asset replacement value x (floored at zero) on a target date

Compare traditional and modern risks

The types of risk an individual is exposed to may vary depending on the degree of economic development: - Can split risks into traditional risks (less developed/poorer countries) and modern risks - Traditional risks include malnutrition, indoor air pollution, water, sanitation, and hygiene - Modern risks include tobacco use, physical inactivity/obesity, urban air quality, road traffic safety, and occupational risks

List the practical considerations when choosing hedging instruments

These are all desirable characteristics - satisfy as many as possible and make good trade-offs - Targeted exposure - hedging instruments should have a good match with the targeted hedged risks * Basis mismatch can increase significantly during crises - High liquidity - hedge will inevitably need to be rebalanced - Good credit quality - otherwise, counterparty risk may not be worth it - Competitively-priced hedges - examine a range of alternatives - Available under a wide range of economic conditions - OTC markets sometimes disappear

List the modeling considerations for operational risk

These mostly relate to low frequency / high severity risks - Scenario analysis is useful - Stochastic is best for calculating required capital - Bottom-up methods: * Bayesian networks (Bayesian probability models) * Simpler approach: assume required capital = a fixed percentage of income - Top-down methods * Operational Risk Income Volatility = Total Income Volatility - Income Volatility from Non-ORs ** Not forward-looking * Solution: use CAPM to strip out changes due to overall market movements * Difficult to disaggregate effects of various factors - Risk capital approach * OR Risk Capital = Total RC - RC for Credit and Market Risks * More forward-looking (better risk indicator) * Drawbacks: not straightforward to assess total RC and may ignore risk interactions

State the goal of ORSA

To fostor an effective level of ERM at all insurers, through which each insurer identifies, assesses, monitors, prioritizes, and reports on its material and relevant risks identified by the insurer, using techniques that are appropriate to the nature, scale, and complexity of the insurers' risks, in a manner that is adequate to support risk and capital decisions

Define the goal of hedging

To mitigate the risks the company does not want to bear so the company can retain the risks that bring optimal return

Identify other SGUL product designs (both traditional and non-traditional)

Traditional designs: 1. Protection-oriented SGUL - Minimizes premiums - Intended for policyholders that value death protection over accumulation 2. Accumulation-oriented SGUL - Lower charges and higher credited rates to build more AV - Higher premiums and higher lapses expected Non-traditional designs: 1. Indexed SGUL - Credited rate based on index performance 2. Variable SGUL - AV linked to separate account assets - M&E charge instead of pricing spread

Define alternative risk transfer vs. traditional risk transfer

Traditional risk transfer (TRT) refers to the heavily-regulated commercial insurance industry - Capacity limited by traditional underwriting and heavy regulation Alternative risk transfer (ART) uses alternative risk financing mechanisms - Operates outside of the commercial insurance industry - More sophisticated, innovative risk takers - Insureds retain much more risk (sometimes all risk) - Insureds have high net worth to supply risk capacity

Describe traditional forms of underwriting and how end-to-end underwriting is different

Traditional underwriting methods: - Full underwriting (lowest mortality cost) * Medical/para-medical exam * Collection of fluids (blood/urine) * Rx, MVR, and MIB - Non-medical underwriting (higher mortality cost) * No medical/para-medical exam * No fluid collection * Rx, MVR, and MIB End-to-end approach: - Supplements non-medical underwriting with predictive analytics - Can bring non-medical mortality cost down to fully-underwritten levels - More customer friendly (less invasive) - Faster, cheaper

Define treaty reinsurance and facultative reinsurance

Treaty reinsurance - risks in the agreement are automatically ceded - More common and less labor-intensive than facultative - "Follow the fortunes" mindset * Reinsurers have trusted cedants to make sound underwriting and claims decisions * Works well when the interests of the cedant and reinsurer align (benefit mutually) Facultative reinsurance (not automatic) - Cedant can offer individual risks (policies) to the reinsurer - Reinsurer can underwrite each risk and choose which to accept - More administrative burden than treaty reinsurance - More likely to be used for large risks that may be excluded from treaty agreements

State the components of credit risk and key credit risk modeling considerations

Two components of credit default risk: 1. Probability of default 2. Loss given default: magnitude of loss given that default has occurred - Usually recoveries are modeled instead Key credit risk modeling considerations: - Individual credit risks are often interrelated (joint risks) - Credit risk can be highly interrelated with other risks (e.g. market, liquidity) - Must balance accuracy of model with its cost - Models can be qualitative or quantitative (more common) * Qualitative: can be used to assign ratings * Quantitative: input financial variables, and output score, default probability, etc.

Provide the basic formulas for both Solvency II and US RBC capital requirements and identify the types of risks captured in each

Types of risks captured by each: - Basic SCR: underwriting, market, credit, and operational risks * Aggregates SCRs from individual risks with correlation factors (var-covar approach) - ACL RBC: asset, underwriting, and other risk * Focuses on the most common/material risks for life, health, and P&C * Does not necessarily capture every single exposure

State the key global trends in reinsurance

US insurers are net purchasers of reinsurance - 62% of premium ceded by US insurers goes to non-US reinsurers - 7 non-US jurisdictions used (Bermuda is #1) - 83% of all ceded premium concentrated in 4 reinsurers (only 1 is in the US) No new US-based reinsurers in 2 decades (26 total) Global life and health reinsurance is more concentrated than P&C - 90% of L&H held by the top 10 reinsurers of the world - 50% of P&C in top 10 Global reinsurance capital is increasing, and reinsurance premiums are decreasing Significant growth in reinsurance alternatives (e.g. cat bonds)

Compare the cost of capital method for the IFRS 17 risk adjustment to market-consistent valuation

Under a market-consistent approach: - Capital may be based on EC or regulatory capital - Capital reflects financial and other risks relevant to IFRS 17 - May use a shorter time horizon (e.g. 1 year) - CoC rate usually reflects frictional costs that aren't relevant to IFRS 17: * Cost of double taxation on investment returns * Agency costs - misalignment of investors' and insurer management's interests * Financial distress costs - e.g. loss of reputation - CoC rate may be prescribed (e.g. 6% for Solvency II and SST)

Discuss underwriting, due diligence, and possible outcomes

Underwriting - Banks underwrite mortgages, insurers underwrite policies, etc. - Methods: GLM, credit scoring, discriminant analysis - Also done for OTC derivative collateral Due diligence (non-standard form of underwriting) - Assess the party that will provide goods or services - Consider financial strength, firm's operations, etc. - Consider incidential credit risk (outside firm's core area) Possible outcomes - Don't issue loan or insurance policy - Don't use counterparty or structure payments to limit credit risk - Use credit insurance for incidental credit risk

Discuss the considerations for measuring underwriting risk

Underwriting risk - mortality, morbidity, longevity, lapse, other policyholder options - Volatility risk (mortality, morbidity, longevity) decreases as policy volume increases * Stochastic binomial model: generate a random number s, 0 ≤ s ≤ 1 ** If s ≤ q: dead! ** Else, generate another s: if s ≤ q: lapse! - Systematic components * Misestimation of the mean (parameter risk) * Trend risk - risk that mortality trend differs from what is expected - Catastrophic risk - low frequency, high severity (pandemics, earthquakes) * Usually mitigated with reinsurance - Policyholder option risks usually depend on economic conditions (requires stochastic or stress modeling) * Includes surrender, lapse, and annuitization risks

Identify the considerations and advantages/disadvantages of qualitative credit models

Used by credit agencies to assign ratings to securities and companies Considerations for qualitative models 1. Seniority of debt: higher seniority = lower default risk 2. Collateral (more collateral = less default risk, higher collateral liquidity = less default risk) 3. Term of debt influences approach: long- vs. short-term tebt Advantage over quantitative models: allow for more factors Drawbacks: - More subjective, less consistent - May not adapt to changing economic environment - Subject to anchoring bias

Identify the credit portfolio models

Used to account for jointly-fat tails (increasing correlations in the tail) 1. Multivariate structural models (Multivariate Merton) - Assume multivariate lognormal distribution OR enter individual default probabilities in a copula 2. Multivariate Credit Migration Models - Multivariate form of CreditMetrics 3. Common shock models - Calculate probability of default associated with Poisson shocks 4. Time-until-default models - Model default probability using bond survival functions

State uses and common deficiencies of internal models

Uses of Internal Models: - To assess financial condition - To determine regulatory capital requirements - To examine risk management competency - To develop and refine corporate strategies Common Deficiencies of Internal Models - Lack of sufficiently heavy tails - Parameter calibration - Lack of consideration of recent trends

Define systemic risks

Usually extensions of market risk to the broader economy - Feedback risk implies serial correlation * Implies arbitrage, but probably only temporarily, so exclude from modeling - Contagion risk - certain series can be closely linked for extreme values * Possible solution: copulas (if sufficient data)

Identify the 3 "Greeks" to assess VA guarantee risks

VA guarantee risks can be assessed with the following "Greeks" 1. Delta (MV risk) - rate of guarantee change for changes in the MV of underlying assets 2. Rho (IR risk) - rate of change in liability as interest rates change 3. Vega (volatility) - rate of change in liability as volatility changes Hedging these risks can lower required EC!

Define VaR vs. TVaR (CTE)

VaR = quantile associated with probability of ruin - Expresses shareholder view of risk (walk away on ruin) TVaR = average of worst losses - Expresses regulator view of risk (protect policyholders) - Accounts for shape of tail

Compare the confidence level method with CTE for purposes of the IFRS 17 risk adjustment

VaR is easier to calculate/communicate VaR is a better choice when: - Tail is unknown/unreliable/volatile * CTE assumptions may be difficult/impossible to validate - Distribution is not highly skewed - Diversification benefits are not significant/important CTE is a better choice when: - Distribution is highly skewed - Capturing tail info is important - Diversification benefits are important (coherency desired) Neither captures info about the rest of the distribution (e.g. favorable outcomes)

Describe validation for risk adjustment aggregation/allocation

Validate risk adjustment subadditivity at all levels: RiskAdj(A + B) ≤ RiskAdj(A) + RiskAdj(B) Ex-ante aggregation of risk adjustments: - Give specific attention to: * Consistency of assumptions between aggregated elements * Being exhaustive but avoiding double counting * Qualitative analysis of the global risk adjustment value - Verify homogeneity/consistency/granularity of components being aggregated Ex-post aggregation of risk adjustments: - Aggregated risk adjustment validations: * Consistency of calculation methods used for each pre-aggregation risk adjustment * Diversification effects - Validate the chosen dependence structure (directly or indirectly) - Top-down and bottom-up recalculations ensure consistency between ex-post and ex-ante approaches

Describe risk adjustment validation considerations before aggregation with respect to the following items: - Capital amount for stand-alone risks - Aggregate capital amount - Projection of capital amount

Validation of capital amount for standalone risks: - Deterministic: validate mean result ≈ stochastic mean - Stochastic: verify enough simulations were run - Validate data groupings Validation of aggregate capital amount: - Diversification benefit impact of each individual risk - Reasonableness of diversification benefit based on risk groupings - Reverse scenario: review diversification associated with a specific stress level - Verify correlations used to aggregate risks Validation of the projection of capital amount: - Projection factor (ratios, drivers, etc. instead of exact calculations) * Bad drivers: level premiums when risks are increasing, GAAP/Stat liabilities that don't reflect contracts' actual risks - Constant projection ratio * Review actual ratio over projection period and by scenario * Check that diversification benefit is actually consistent - Length of projection should be long enough to include all material risks

Describe risk adjustment validation considerations before aggregation with respect to the following items: - Rate of return on capital - Definition of confidence level

Validation of the rate of return on capital: - Consistency/appropriateness of capital allocation and rate by line of business - Consistencies between capital amount and rate of return - Qualitative vs/ quantitative metrics for determining rate of return Validation of the definition of confidence level: - Formal review of documented risk tolerance - Independent reviews of confidence level, time horizon, and risk adjustment method - Should be aligned with the entity's business model risk appetite statement

Describe considerations for validating the final risk adjustment result

Validation through sensitivities: - Analyze how the risk adjustment changes if: * Using a different statistical distribution * Significant parameters are increased/decreased individually * Using a different dependence structure (e.g. different copula) Validation through analysis of change in risk adjustment - Break down the change from previous reporting period Validation through comparison to benchmarks - Use consistency checks/ratios to detect errors * E.g. compare ratio of risk adjustment to net PV of insurance cash flows over time - Quantitative impact studies - compare to calculations using financial data from organizations writing similar business Validation through proxies (e.g. Solvency II) - Re-project using the proxy and compare the resulting risk adjustment

Describe how AM Best views volatility in the rating process

Volatility is a leading indicator of future balance sheet strength! ERM should allow companies to find the best risk/reward balance - Balance stakeholder expectations (shareholders, policyholders, regulators, rating agencies, etc.) - Some insurers should reduce volatility (purchase reinsurance, change business mix, refine liabilities, etc.) - Others should accept current volatility and focus on boosting returns (price actions, expense reductions, changes to reinsurance, etc.)

Give a description of the spike-and-slab prior

When using a spike-and-slab prior, the prior distribution for the coefficients is a mixture of two densities, one with a small variance (spike) and one with a much larger variance (slab) - Both distributions are normal with mean 0, and the variance of the slab model is higher than the variance of the spike model - Ex: Slab ~ N(0, 1) and Spike ~ N(0, 0.1)

Credit Risk of Options

Whereas forward contracts and swaps have bilateral default risk, options have unilateral credit risk - The buyer of an option bears all the credit risk - Like forward contracts, European options have no payments until expiration - Options have no current credit risk until expiration, but have large potential credit risk that is equal to the option's market value

Explain why information is necessary but not sufficient to improve policyholder decision-making

Why providing information isn't enough: - More information can help but may not be properly framed - It is very difficult to increase knowledge - Educated people don't necessarily make rational choices Possible solutions: - Incentives should be properly structured and evident - Decision environment should counteract biases ("de-biasing") - Use defaults that reflect expert knowledge while still allowing individual choice

Describe the Board vs. management role in risk management

- Board's role: oversee and monitor risk management and reporting - Management's role: day-to-day risk management * CEO = critical link between the Board and management * CEO should have ERM responsibilities and promote risk management as a core competency

List the questions the CRO should ask about

- Board's understanding of risk tolerance (very important to establish) - Management's incentives - Flow of risk information - Insurer's performance on a risk-adjusted basis

List risk identification techniques

- Brainstorming - Independent group analysis - Surveys, gap analysis, and delphi - Interviews - Working groups

State the pros of the bootstrap method

- Can include fat tails, jumps, or any departure from the normal distribution - Accounts for correlations across series (because one draw consistents of the simultaneous returns for N assets)

State the cons of Monte Carlo Methods

- Can involve costly investments in intellectual and systems development - Requires substantially more computing power than simpler methods - Measuring VaR using these methods requires hours to run, but computing power is improving this

Advantages of VaR

- Can quantify the potential loss in simple terms, and is easily understood by managers - A versatile measure of risk that can help determine how to allocate capital - Even if it is inaccurate at times, the risk manager can still use VaR to help make decisions as long as it is not used in isolation (better to pair it with stress testing)

State the cons of Monte Carlo Risk Neutral (RN) valuation

- Cannot price options accurately when the holder can exercise early - The distribution of prices must be finely measured to price options with sharp discontinuities

List the aspects to address in the risk management policy

- Clear risk management philosophy - Relationship between risk management and mission/values/objectives - How risk management is embedded in activities (pricing, reserving, etc.) - Governance and oversight (responsibilities of Board, management, other areas) - Behavioral expectations - Scope of activities covered - Supervisory requirements - NB acquisition - Risk categories and definitions

List the good characteristics of a CRO (appointed by CEO)

- Has Board's support - Works with CFO to integrate earnings and risk management - Highly visible and accountable - Facilitates risk debate and coordinates risk activities

Explain the importance of a common risk language in the insurer

- Competing risk language is bad! * Inhibits management buy-in * Confuses people; prevents ERM culture * Reinforces silos * Form over function: May fail to identify "real" risks * Creates inefficiencies and duplication * Leads to inconsistent risk measurement - Best practices for a common risk management language: * Universally understood top-down risk rating system * Standard templates and risk categories * Thresholds for reporting and escalation

State Best Practices in Identifying and Assessing Emerging Risks

- Conduct emerging risk reviews - Integrate emerging risk review into the strategic planning process - Identify all assumptions and carry out disciplined assumption testing - Challenge conventional thought processes and expectations - Apply new and developing methodologies to better understand and predict risk

Describe risk-focused process analysis

- Construct flow charts for every process in the organization - Analyze the points where risks/failures occur * Describe all risks in detail * Identify links across processes - Get input from all key areas * Advertising and sales * Premium collection * Investment and capital raising * Hiring staff and payroll

Two Time Perspectives for Measuring Credit Risk

- Current credit risk: risk associated with credit events in the immediate future - Potential credit risk: risk that the counterparty will default at a later date (despite being currently solvent)

Describe and state the notation for individual VaR

- Definition: The VaR of one component taken in isolation - Notation:

Describe and state the formula for undiversified VaR

- Definition: The sum of individual VaRs (assumes perfect, positive correlation) - Formula:

Describe and state the notation for portfolio/diversified VaR

- Definition: VaR taking into account diversification benefits between components - Notation:

Describe and state the formula for incremental VaR

- Definition: change in VaR owing to a new position - Formula:

Describe and state the notation for marginal VaR

- Definition: change in portfolio VaR resulting from taking an additional dollar of exposure to a given component - Notation:

Describe and state the notation for component VaR

- Definition: how much the portfolio VaR would change if the given component was deleted - Notation:

State the steps to develop a culture implementation plan

- Develop a risk management behavioral model - Get support of senior management and develop their risk awareness - Embed/reinforce behaviors in frameworks and processes - Timeframe should be realistic

Describe the shortcomings of scenario analysis (according to ERM-107)

- Difficult to estimate many of the strategic risks due to insufficient data - The models backing the quantification of scenarios are built on rather restrictive assumptions - Fat-tails or regime shifts may not be adequately modeled (or may not be modeled at all) - There is a strong tendency for people to focus on numbers at the expense of more important qualitative aspects - Can lull corporate executives and regulators alike into a false sense of security

List the key considerations in performance management and reward systems

- ERM implementation will fail if the managers don't have incentives - Key considerations: * Size of incentive should motivate the targeted individual * Include senior management at a minimum * Use clear, activity-based measurements (milestones, financial stats, VaR, etc.)

State the benefits of scenario analysis (according to ERM-107)

- Encourages managers and other employees to think through the diverse pieces of the puzzle. Allows managers to organize these fragments into a cohesive pattern in the future business environment - Allows managers to discuss the implications on the effectiveness of current strategies while evaluating alternative strategic options - Can help make blind spots visible and uncover areas where further knowledge and insight are needed - Helps managers evaluate the robustness of strategic alternatives when operating in an uncertain future business context - Provides executive decision-makers a way to think about the need for corporate response capabilities if they face unexpected events - Company strategists get to focus on how central external environmental factors can frame competitive conditions - Can apply structured thinking around possible future business scenarios

Difficulties with using Credit VaR

- Estimating credit VaR is more complicated than market VaR because credit events are rare and recovery rates are hard to estimate - Credit risk is less easily aggregated than market risk because the correlations between the credit risks of counterparties must be considered

State the pros of Monte Carlo Methods

- Flexibility - The most powerful approach to VaR - Can account for a wide range of risks - Can account for non-linear exposures and complex pricing patterns - Can have simulations extended for a longer time horizon - Can be used for operational risk management

State the cons of the bootstrap method

- For small sample sizes M, the bootstrapped distribution may be a poor approximation of the actual one - This method relies heavily on the assumption that returns are independent

Describe the four modes of scanning

- Formal search - obtain information on the relevance of specific issues and input this information into the planning and decision-making processes - Conditional viewing - track pre-selected information from particular sources - Informal search - actively look for information through unfocused and unstructured efforts - Undirected viewing - scan many diverse sources of information with no specific informational needs in mind

State the pros of Monte Carlo Risk Neutral (RN) valuation

- General - Can be applied to options that have price-dependent paths or complicated payoffs at expiration (e.g. arithmetic asian options)

Describe high reliability organizations (HROs)

- HROs comprise a particular group of organizations that have the common characteristic that even minor mistakes might result in a fatal outcome - These organizations have specific practices in place to prevent these problems, while still retaining sufficient flexibility to deal with unexpected situations

List the steps to proactively embed risk management behaviors

- Include proactive principles in risk management strategy and policies - Set corporate risk goals for senior managers - Define behaviors in roles and succession/talent development - Training programs

Extensions/Supplements to VaR

- Incremental VaR: Measures the incremental effect of an asset on the VaR of a portfolio by measuring the difference between the portfolio's VaR with and without that asset - Cash flow at risk (CFAR): The minimum cash-flow loss that we expect to be exceeded with a given probability over a specified time period - Earnings at risk (EAR): Defined analogously to CFAR, but measures risk to accounting earnings - Tail Value at Risk (TVAR, CTE): "VaR" + "The expected loss in excess of VaR"

Describe when learning may be preferred over selection

- Low complexity projects with a low cost of learning - Long-time horizon projects with larger budgets

State the cons of simulation methods

- More prone to model risk owing to the need to prespecify the distribution - Much slower and less transparent than analytical methods - Create sampling variation in the measurement of VaR - Greater precision comes at the expense of a vastly increased number of replications, which slows down the process

List the key features of ERM

- Overall Governance Structure - Risk Management Policy - Risk Tolerance Statement - Feedback Loop

List the steps of independent group analysis

- Participants silently write down ideas on risks (no collaboration!) - Facilitator aggregates ideas - Facilitator leads a discussion where risk gets defended and justified - Group then ranks risks anonymously - Facilitator combines rankings to give an objective ranking of the risks

What are the components of a PESTEL analysis?

- Political events - Economic developments - Social trends - Technological innovations - Environmental issues - Legal issues

Credit Risk of Forward Contracts

- Prior to expiration of the forward contract, no current credit risk exists - There is potential credit risk for the payments to be made at expiration - The market value of a forward at a given time reflects its potential credit risk

State the pros of simulation methods

- Quite flexible - They can either postulate a stochastic process or resample from historical data - Allow full valuation on the target date

State reasons to understand Emerging Risk Management

- Reduce volatility - Reduce the probability of failure - Add value - Reveal potential opportunities - Encourage formulation of more effective strategies

Describe the risk committee - keys to effectiveness and responsibilities

- Risk committee - established by the Board * Includes risk, audit, financial reporting, and compliance staff * More effective if: ** Diverse, inquisitive, experienced individuals ** Members allowed to challenge risk assessment processes ** Reporting is effective - Typical risk committee charter responsibilities: * Framework effectiveness * Compliance with supervisory requirements * Establish independent risk function to execute committee's mandate

List 4 broad areas of risk identification

- Risk identification tools - Risk identification techniques - Assessing risk nature - Risk register

Define risk register and list factors that each entry should have

- Risk register - a central document with details of all risks faced by the organization * Record risks here once identified by a previous method - Factors that each entry in a risk register should have: * Unique identifier * Risk category * Date of assessment * Quantifiable? * Likelihood and severity * Exposure period

List risk identification tools

- SWOT - Risk checklists - Risk prompt lists - Risk taxonomy - Risk trigger questions - Case studies - Risk-focused process analysis

Describe when selection may be preferred over learning

- The degree of complexity and unknowability of the terrain is high - The cost of learning is relatively high - Time is of paramount importance to ensure at least one early success

Describe the difficulties in designing foolproof contingency plans

- The environment is outside the view of corporate management and therefore cannot be planned for - Management can be lulled by the illusion of having robust contingency plans in place - Contingency planning can make things worse if it leads to tunnel vision and ignorance * Plans should not restrict responses only to actions built into an existing repertoire!

Describe the Ho-Lee Model

- The function θ(t) is chosen so that today's bond prices fit the current term structure - No-arbitrage model

List the strategic choices in risk tolerance statement

- Use the same time horizon as the corporate strategy (3-5 years) - Set boundaries for how much risk the insurer is prepared to accept - Make a clear link between risk tolerances and limits - Base risk tolerance on insurer-specific circumstances - Focus on which risks to take, not on how much risk to take

State the ways that the insurer's risk function can become involved in new activities

- Utilize actuarial skills to identify/assess risks - Strategy team should assess risks - Prepare risk assessments before launch - Engage relevant supervisors

Limitations of VaR

- VaR can be difficult to estimate - Different estimation methods can give different values - Can lull one into a false sense of security - Understimates magnitude and frequency of worst-case scenarios - VaR for individual positions does not generally aggregate in a simple way to portfolio VaR - Fails to incorporate positive results into its risk profile, thus providing an incomplete picture of overall exposures

What is the meaning of θ and κ in the stoachastic differential equation (SDE): dr = κ(θ - r)dt + σ(r^γ)dz

- κ is the mean reversion speed - θ is the long-term interest rate

State the components of Porter's five-forces model

1. The threat of new entrants 2. The bargaining power of buyers 3. The bargaining power of suppliers 4. The threat of substitute products or services 5. The intensity of competition in the industry

State two alternative viewpoints that explain fat tails

1. The true distribution is stationary and indeed contains fat tails 2. The distribution does change through time

State the disadvantages of brainstorming

1. "Free riders" - individuals that show up but don't contribute 2. Requires single location for everyone 3. Participants' responses may be influenced by prior contributions 4. Can lead to a lack of completeness

Value at Risk (VaR): Measurement Elements

1. A probability level (typically 5% or 1%) 2. The time period 3. The approach to modeling the loss distribution (i.e. the technique used to estimate VaR)

State the four main acceleration methods discussed in Jorion Ch. 12

1. Antithetic Variable Technique 2. Control Variate Technique 3. Importance Sampling Technique 4. Stratified Sampling Technique

State three guidelines when creating and using the RCD tool

1. Captures known risk categories: the RCD tool should be all-inclusive in that it must have a category/sub-category for all individual risks that are known to the organization - Typically impossible to be fully comprehensive since many individual risks cannot be known in advance - Specific category names/classification are not as important as making sure all known risks are captured in the analysis 2. Internal communication: internal to the organization, the RCD tool should be used to create a uniform risk language, so there is consistent enterprise-wide understanding of risk definition and categorization - Define risks consistently by source for reliable risk identification and risk quantification processes 3. External Communication: when communicating risk-related matters to external stakeholders, the RCD tool should be mapped/translated into the risk terminology used by the external stakeholders to avoid miscommunication

List the 3 aspects of developing a risk behavior model

1. Focus on choosing/managing risk, not eliminating it! 2. Encourage people to speak up 3. Empower people and give them skills needed for desired behaviors

State the 6 Characteristics of Emerging Risks

1. High level of uncertainty 2. Lack of consensus 3. Uncertain relevance 4. Difficult to communicate 5. Difficult to assign ownership 6. Systemic or business practice issues

State the common characteristics of emerging risks

1. High uncertainty: little information, so frequency/severity is difficult to assess 2. Difficult to quantify 3. No industry position: no single insurer wants to make the first move 4. Difficult to communicate: danger of reacting to phantom risks 5. Supervisory involvement is often necessary

State the five steps for the qualitative risk assessment (QRA)

1. Identify participants 2. Provide advance communication 3. Interview participants 4. Score likelihood/severity 5. Select key risks

Describe the five major steps to arrange scenario planning

1. Identify the key environmental risk factors 2. Elaborate some of the major themes since they may characterize alternative developments in future competitive market conditions 3. Elaborate on the major themes outlined previously and describe a few environmental scenarios that arise as the consequence of different assumptions about the risk factors and the underpinning themes that are relevant for the corporate strategy 4. Evaluate the consequences of the key strategic risk factors within the themes that shape the alternative scenarios in view of essential strategic concerns 5. Formulate new strategic alternatives, if required, and evaluate them given the different scenarios

List the steps of the typical roadmap to establish risk tolerance

1. Measure qualitative and quantitative risk separately 2. Merge qualitative and quantitative measurements into current risk appetite 3. Define and assess ongoing risk appetite 4. Develop metrics to monitor 5. Monitor activities relative to defined appetite

State two drawbacks of quasi-Monte Carlo (QMC) methods

1. Since the draws are not independent, accuracy can't be assessed easily 2. For high-dimensionality problems, some QMC sequences tend to cycle. This leads to a decrease in performance

State the three types of risk categories used for the RCD tool

1. Strategic - a category of risks related to items of strategic importance - Often, these are differentiators of success/failure versus competitors 2. Operational - a category of risks related to item of routine operations - Typically includes risks related to human resources, technology, processes, and disasters 3. Financial - a category of risks related to external markets and prices - Typically includes risks related to the economic environment, stock market, bond market, and commodity prices

What are the four components of a SWOT analysis?

1. Strengths (positive internal) 2. Weaknesses (negative internal) 3. Opportunities (positive external) 4. Threats (negative external)

Methods for estimating VaR

1. The analytical or variance-covariance method 2. The historical method 3. Monte Carlo simulation

Credit Risk of Swaps

A swap is similar to a series of forward contracts - However, because a swap has periodic payments, there will also be current credit risk at a series of points during the contract's life - The swap's market value can be calculated at any time and reflects the potential credit risk For most swaps, the potential credit risk is the largest during the middle period of the swap's life - At the beginning of the swap, credit risk is small because the risk of default is low - At the end of the swap, credit risk is low because most of the swap payments have already been made (i.e. low exposure at default)

State the advantages and disadvantages of the historical method

Advantage: - Nonparametric Disadvantages: - Relies completely on events of the past - However, in many cases, the past is not a perfect predictor of the future

State advantages and disadvantages of multivariate GARCH

Advantage: - Provides internally-consistent risk estimates for a portfolio of assets Disadvantages: - There are lots of parameters to estimate - Need to make sure the covariance matrix is positive definite for all t

State the advantages and disadvantages of the variance-covariance method

Advantage: - Simple and easy to calculate Disadvantage: - Assumes returns are normal, which is inappropriate for portfolios that contain options or distributions with fat tails

What does GARCH stand for?

Generalized Autoregressive Conditional Heteroskedastic

What is the main advantage of establishing a common risk vocabulary?

It ensures a more consistent way of looking upon and analyzing risks across the organization

State the pros and cons of EWMA

Pros: - Easy to implement - Only relies on one parameter - so estimation is simpler than GARCH - More robust to estimation error than other models - The estimator is recursive. The forecast is based on the previous forecast and the latest innovation Cons: - Setting the decay factor through optimization is challenging - Different values of the decay factor create incompatibilities across the covariance terms and may lead to unreasonable values for correlations - The model doesn't allow for mean reversion, but in practice we do observe mean reversion in monthly risk forecasts

State the advantages and disadvantages of EWMA for modeling correlations

Pros: - Much simpler than GARCH - Parsimonious Cons: - Lacks flexibility - The EWMA implied covariance matrix is often not positive definite

State the pros and cons of MA

Pros: - Simple - Widely employed Cons: - Ignores the dynamic ordering of observations - Ghosting Features - Uncertainty for how to set M

State the pros and cons of GARCH

Pros: - Univariate GARCH provides a parsimonious model with few parameters that seem to fit the data quite well - These models are the mainstay of time-series analysis of financial markets that systematically display volatility clustering - The estimator is recursive. The forecast is based on the previous forecast and the latest innovation Cons: - Nonlinearity, difficult parameter estimation - The basic GARCH model is symmetric

Describe selection and learning

Selection: - Selection relies on conducting multiple parallel trials, understanding that many projects will ultimately be abandoned in the process of finding one solution - Speed and low cost are emphasized at the expense of learning Learning - Relies on a cyclical, experiental, and sequential learning process - Trial and error learning

What is meant by a steady state GARCH model?

Steady-state is when the initial variance is equal to the long-term variance h

State the pieces of SWOT analysis

Strengths / Weaknesses - INTERNAL: market share, economies of scale, leadership, solvency Opportunities / Threats - EXTERNAL: competitors, demand, demographics, market liquidity

Describe volatility forecasts under the EWMA model

The expected forecast of future volatility values is equal to the current volatility estimate

Describe the importance of the row total and column total for influence matrix analysis

The row total indicates: - How the risk factors affect all of the other risk factors - Captures the importance of the specific risk factor and its relative ranking The column total indicates: - How much each risk factor is influenced by all of the other risk factors - Called the passive score

Describe risk tolerances vs. risk limits

Tolerance = broader risk exposure that the Board accepts - Set to achieve business strategy - Must be translatable into risk limits - Parameters used to describe risk tolerance: * LOBs the insurer will or won't accept * Earnings volatility * Desired capital for supervisory requirements, desired rating, and/or EC * Maximum aggregate risk exposure Limit = specific threshold that can be monitored (e.g. ~KRIs) - Counterparty credit limits for investments and reinsurers - Target credit quality for reinsurers/assets - Concentration limits - Underwriting/pricing limits - Reserve adequacy - Liquidity benchmarks

Value at Risk (VaR): Definition

VaR is a probability-based measure of loss potential for a company or fund - Defined as an estimate of the loss that we expect to be exceeded with a given level of probability over a specified time period Example: VaR for a portfolio is $1.5M for one day with probability of 0.05 - This means there is a 5% chance the portfolio will lose at least $1.5M in a single day

Define emerging risks

those risks an organization has not yet recognized or those which are known to exist, but are not well-understood