CFA Level III - Equity

Considerations when determining the appropriate level of risk (3)

(1) Implementation constraints: constraints on short positions or leverage may limit mgr's ability to under or overweight (limiting information ratio) (2) limited diversification opportunities: increasing risk leads decreasing marginal increases in expected returns (which is why efficient frontier is concave) (3) leverage could solve for #2, but too much leverage will eventually bring a reduction of expected compounded return in a multi-period setting

Merits of long-only investing

(1) LT risk premiums - risk premiums, e.g. mkt risk premium, are earned by investors going long the securities; short sellers will suffer negative returns over the long term - investors that have shorter time horizons may prefer some short exposure (2) capacity and scalability - of long-only strategy is set by liquidity of underlying securities; capacity of shorting is set by availability of securities to borrow to facilitate short selling -> long-only likely has more capacity (3) limited legal liability -> maximum loss of long is amount they paid for security; short has unlimited potential loss (4) regulations - some countries can ban short selling (5) transactional complexity - long-only has to tell a broker to buy and subsequently sell - short seller has to find shares to borrow, provide collateral, face risk of shares being recalled at inopportune time, find prime broker, etc (6) costs higher for long/short funds - both mgmt fees and operational expenses

(2) ways to quantify acceptable level of deviation from the benchmark

(1) active share (2) active risk

Benefits to overall portfolio of equity (4)

(1) capital appreciation - the main driver of LT equity returns (i.e. price returns) - equities on average have higher returns than debt during periods of strong growth but underperform in weak economies (2) dividend income - dividend yields are more stable than return due to price change (3) diversification - less than perfect correlation with other asset classes (causing port SD to be less than wtd sum of individual asset SDs) - but in crisis, correlations tend to increase, reducing benefit; also, standard deviations could increase, further reducing expected reduction in portfolio risk (4) potential to hedge inflation - company may be able to charge customers more when input costs rise due to inflation - most equities are positively correlated with inflation, but in hyperinflation negative correlations are observed

VaR based measures (3)

(1) conditional VaR (CVaR) - expected loss given VaR has been exceeded (aka expected tail loss or expected shortfall) (2) incremental VaR (IVaR) - the change in VaR from adding a new position to a portfolio (3) marginal VaR (MVaR) - impact on VaR of a very small change in position size -> in a diversified portfolio, shows contribution of each asset to total VaR

Process to create quantitative active investment strategy and pitfalls of quantitative investing

(1) define the mkt opportunity (investment thesis) (2) acquire and process the data - build databases - track companies over time (eg M&A, bankruptcies), company fundamentals, earnings / macro variables / sentiment indicators (3) back-test the strategy - apply the strategy to historical data to assess performance -> if there is strong relationship b/w factor exposure and subsequent performance then the factor has high predictive power - this correlation coefficient is the factor's information coefficient (IC) - calc factor score, which gives distance in SD away from average; e.g. for earnings yield, factor score = (stock's earnings yield - industry earnings yield) / mkt SD of earnings yield - calc IC - Pearson IC assumes a linear relationship b/w factor exposure and holding period return (i.e. subsequent returns), & IC is the correlation; ranges from -1 to +1, or -100% to +100% - Spearman Rank IC - Pearson IC is sensitive to outliers; Spearman Rank IC addresses this and is superior to Pearson IC - it is the IC of the rank of factor scores and rank of subsequent performance -> this analysis shows how far earnings yield (for example) are from avg earnings yield (factor score) and if this is related to future security performance (IC) - an IC score of even 5 or 6% is strong and indicates predictive power -> outliers can cause the Pearson IC to be biased, so should use Spearman to assess performance of the factor -> Info coefficient > 0.5 is deemed to be strong (4) evaluate the strategy - out-of-sample testing where you apply model to data different than what was used to build the model to confirm model robustness (5) portfolio construction - consider risk (individual variance of positions and correlation b/w) and trading costs Pitfalls: - survivorship bias - if back-tests are only applied to existing companies it will overlook cos that have failed in past - look-ahead bias - used info in model to give trading signals when info was not available - e.g using December financials to analyze January results, but likely that December financial info was not available during January so was not actually available to act as a trading signal - data-mining / overfitting - testing a lot of data until the analyst finds data that suggests the particular strategy works - make the data fit a pre-conceived relationship or when model is biased towards patterns that only occurred in the past - unrealistic turnover assumptions - constraints on turnover may limit ability to follow a strategy - lack of availability of stock to borrow - may limit ability to follow short selling strategy - transaction costs

(2) considerations in choosing a benchmark

(1) determine desired market and risk exposures - market exposure - eg, broad mkt exposure vs focused on specific sectors, domestic vs. international, developed / emerging / frontier markets, etc - risk factor exposure - expected sensitivity to various risk factors (mkt risk aka beta, firm size, growth vs. value, prior returns / momentum -> aka the the factors in multi-factor models) (2) identify methods used in constructing and maintaining an index - which stocks to include? exhaustive (every stock in defined universe, e.g. total US market index) or selective (subset, e.g. S&P 500) - index weighting method? mkt cap weighting, price weighting, equal weighting, fundamental weighting

(4) components of portfolio construction

(1) factor weightings - exposures to rewarded risks / factors that differ from benchmark via over or under weighting (active return due to differences in beta) (2) alpha skill - identify mispricings via unique skills and strategies of the mgr (e.g. factor timing which identifies when a rewarded factor (like value vs growth) might under or outperform or timing unrewarded factors like geo or industry sector exposures, commodity prices, security selection) (3) position sizing - balance confidence in alpha and factor insights with mitigating idiosyncratic risks coming from concentrated positions (4) breadth of expertise - ability to combine the first 3 elements; breadth defined as # of independent decisions the mgr makes each year -> higher breadth implies higher ability to outperform / generate consistent active returns

(3) approaches to construction of passive equity portfolio

(1) full replication - hold all securities in index - PROS: closely matches index return (before transaction costs) - CONS: can be costly when there are a large # of stocks and limited liquidity and must be regularly reconstituted / rebalanced (2) stratified sampling - hold a sample of securities - index divided into risk strata (subsets) based on risk characteristics and random samples of stock within each strata are chosen; weights of stocks for each strata are such that risk factor exposure matches index - reduces cost of full replication but approximates risk factor exposures - the more criteria used in constructing strata, the smaller the tracking error - tracking error declines as size of sample increases, but as more stocks are added, the added stocks will be less liquid, increasing effect of transaction costs - strata must be mutually exclusive and exhaustive -> used when you want to track a portfolio with a large # of constituents or when dealing with low level of AUM (3) optimization - mean-variance analysis to minimize tracking error or maximize a desirable result (eg returns) (more technical and quantitative approach to maximize desirable characteristics) - eg minimizing tracking error, subject to a constraint on portfolio vol - ADV: reduces tracking error relative to stratified sampling -> explicitly accounts for covariances of constituent stock returns rather than relying on a characteristic (eg sector) - DISADV: based on historical relationships, which can change; minimizing tracking error may make portfolio not mean-variance efficient - full replication preferred for index with smaller # of similar liquid stocks and when portfolio large enough to follow full replication - stratified sampling and optimization best with many heterogenous, thinly traded, small cap stocks -> but don't track index as closely so would have higher tracking error - can use a BLENDED approach - full replication for more liquid assets and stratified sampling or optimization for thinly traded

(2) types of risk constraints

(1) heuristic risk constraints - based on experience or general ideas of good practice but lack empirical evidence (e.g. arbitrary limits on exposures to individual positions, sectors, or regions, limits on leverage, or measures designed to control illiquidity / turnover of portfolio) - more likely to be used by fundamental mgrs (hold concentrated portfolios, which increases estimation error and makes formal risk constraints less useful) (2) formal risk constraints - statistical in nature / directly linked to distribution of returns -> formal risk constraints require forecasts of return on distributions, which introduces estimation error (e.g. limits on volatility, active risk, skewness, VaR-based measures) - more likely to be used by quantitative mgrs

When does an asset contribute more to portfolio's active variance?

(1) higher active weight (2) asset's active reruns are related to overall portfolio active returns

Index Weighting methods

(1) market-cap weighting: weight each portfolio stock by its mkt cap % of total mkt cap of stocks in index - the most common mkt cap weighting method is based on stock's free float (outstanding shares not closely held so available for trading by mkt participants) - can be thought of as liquidity-wtd index because large cap stocks tend to have the highest liquidity - S&P 500 is mkt value weighted (2) price weighting: weight each stock by its price - so hold equal number of shares, which gives higher share prices larger index weights - DJIA is price weighted (3) equal weighting: same amount of $ in each one - less concentration risk - has small cap bias so returns are more volatile (4) fundamental weighting: weighting by proportions of total index value of fundamental factor (like sales, income, dividends, book value) - mkt cap and price weighted funds will overweight overvalued securities and underweight undervalued securities b/c both use price in weighting scheme - equal weighting does not consider price, so will benefit from an overweighting of undervalued securities and underweighting of overvalued securities (fundamental weighting has this advantage as well)

(3) approaches to passive equity investing

(1) pooled investments - open ended mutual funds: low transaction costs and convenient fund structure , but trade only at mkt close and have to sell stocks in response to SH redemption (leads to taxable gains) - ETFs: can trade intraday, don't have to sell stocks for SH redemption (deliver in-kind delivery of stock), but higher transaction costs (from commission costs and bid-ask spreads) and possible illiquidity in some ETF secondary mkts (mutual funds are redeemed directly with the fund so have better liquidity) - ETFs can be bought using margin borrowing and investors can take short position in ETF (both unlike open-end mutual funds) (2) derivatives based strategies - use derivatives (options, futures, swaps) to recreate risk/return of an index - ADV: quickly adjust factor exposure at low cost (small premium paid for options, no explicit transaction cost for futures), liquid, easy to leverage portfolio - DISADV: finite expiration so must be rolled over, position limits, existing offerings of exchange traded derivs might not meet speciality portfolio needs, counterparty risk (3) separately managed equity index based portfolios - hold all stocks in index, which has higher costs (must build out infrastructure to do it)

(3) types of factor-based strategies and is it active or passive

(1) returns-oriented: - momentum (overweight stocks that have recently outperformed) - divided yield (overweight stocks with high dividends or div growth rates) - fundamental weighted (based on dividends, sales, income, etc) (2) risk-oriented: seek to reduce portfolio volatility (simple and provides risk reduction, but based on pasta data which may not be reflective of future) - can be seen by overweighting low volatility - eg volatilty weighting would be a risk-oriented strategy (3) diversification-oriented: equal-wtd and max diversification strategies (maximizing ratio of wtd avg volatility of the individual stocks to portfolio vol) - using fewer securities than the index would NOT be a diversification strategy -> note, factor-based is passive (rules-based, transparent, and replicable) but involves elements of active (emphasizing some factors and de-emphasizing others) -> classified as a hybrid investment style

(3) characteristics necessary for equity index to be used as a benchmark

(1) rules-based - rules for including / excluding, weighting, rebalancing frequency must be consistent, objective, and predictable so it can be replicated (2) transparent - rules are public and understandable (3) investable - investors can replicate the risk/return performance of the index

Types of bottom-up strategies (2) and sub-styles of each

(1) value-based - identify securities trading below intrinsic value - relative value: comparing valuation multiples to peers and buying cos with inexplicably low relative multiple - contrarian investing: purchasing or selling against prevailing mkt sentiment - often invest in cos with low/negative earnings or low dividends, expecting the stock to rebound when earnings rebound - contrarians would say investors tend to overweight recent trends - high-quality value: equal emphasis on intrinsic value and financial strength, quality mgmt, profitability ("Warren Buffet approach") - income investing: high dividend yields and positive dividend growth rates - deep-value investing: extremely low valuation relative to assets (e.g. low P/B) or industry peers, often due to financial distress - typically requires LT investment horizon - restructuring and distressed debt investing: invest prior to or during bankruptcy filing - goal to provide value via restructuring or company has enough assets to generate returns - typically requires LT investment horizon - differs from deep-value b/c may not have a low P/B - special situations: mispricings due to corporate events like divestitures, spin-offs, mergers (2) growth-based - invest in companies with revenues, earnings, or CFs that are expected to grow faster than industry or mkt - less concerned with high valuation multiples and more concerned about source and persistence of the growth rates of the company - GARP (Growth at a reasonable price) - look at PEG and pick lowest - value would look at P/E multiple, growth would look at forward EPS -> can look at both (hybrid strategy)

Absolute vs. relative risk

- Absolute risk - used when investment objective expressed in terms of total return w/o referencing any benchmark (e.g. cash + margin) -> focus on size and composition of absolute portfolio variance - relative risk - used when investment objective is performance relative to a market index

Active share and active risk by investment style (pure indexing, factor neutral, factor diversified, concentrated factor bets, concentrated stock picker); what is sector rotator investment approach

- Pure indexing - no active positions; portfolio equal to the benchmark -> zero active share and active risk - factor neutral - no active factor bets - idiosyncratic risk is low if diversified -> low active risk; active share low if diversified - factor diversified - balanced exposure to risk factors and minimized idiosyncratic risk through high # of securities in portfolio - low active risk (low idiosyncratic risk due to diversification); high active share from large amt of securities that are unlikely to be in benchmark - concentrated factor bets - targeted factor bets; idiosyncratic risk likely to be high - high active share and active risk - concentrated stock picker - targeted individual stocks - highest active share and active risk -> concentrated portfolio and high active share (0.90+) - sector rotator: would need large permitted deviations in sector weights and high target active risk (makes concentrated sector and cash bets) -> has high active share and either high or low active risk based on whether or not diversified - a concentrated stock picker would need large permitted deviations in individual security weights, high active risk, low sector deviations - a diversified multi-factor investor would not need such large deviations from index weights but would still need some flexibility to generate a moderate level of active risk and return - would have low single security risk, modest portfolio risk, and some flexibility in factor risk / sector deviations - closet indexer - narrow sector deviation, low targeted active risk

Active management - goal and (2) categories

- active investing seeks to outperform a passive benchmark (1) fundamental - subjective in nature (investment decision based on opinion / discretionary judgment) -> analyst skill / judgment to determine intrinsic value of securities - analyst conducts research on company, mkts, economy, financial statements, business model, mgmt team, and industry positioning to establish valuation (include non financial variables like quality of mgmt, competitive landscape, pricing power) - consideration of firm specific factors such as ESG indicate discretionary - relative to quantitative approach, there will be fewer positions and allocation of each will be larger (due to intensive research required on each position and depth of insights on company characteristics and competitive landscape) - rebalancing process involves continuous monitoring of positions for changes in weights and rebalances at any time according to current opinion - RISKS: at the individual company level - if analyst has misestimated intrinsic value or if mkt fails to recognize mispricing and security stays mispriced (2) quantitative - objective in nature (investment decision based on rules-based, data driven models) -> models generate systematic rules to select investments - analyze large amts of historical data to identify relationships b/w equity returns and factors (e.g. valuation, size, financial strength, industry sector) that have predictive power -> i.e. find historical relationships likely to persist - spread factor bet across large group of securities - take smaller positions in a larger number of holdings to reach desired exposure to risk factors -> weights determined by optimization to maximize alpha or information ratio - formal portfolio optimization used - rebalancing process: formal, regular automatic rebalancing strategy based on systematic rules at predetermined intervals (eg monthly or quarterly) - RISKS: at the portfolio level - if factor returns do not deliver performance predicted by model -> actrive managers can use both - the distinction is in how investment decision is made - if based on systematic rules, mgr classified as quantitative; if based on mgr opinion, classified as fundamental - the fundamental manager will use judgement in making investment decisions, whilst the quantitative manager will use judgement in building models, particularly in deciding which factors and signals are relevant - discretionary uses an informal approach to portfolio construction by only selecting attractive securities (discretionary) vs a more formal approach that considers are more diverse selection of securities (systematic) - loss aversion is more symptomatic of fundamental than quantitative

Active return (incl. formula) and sources (3)

- active portfolios look to outperform a benchmark after all costs - excess return above a benchmark (ie active return) -> will be positive if mgr over weights securities that outperform benchmark and underweights securities that underperform - Active return = SUM [(active weight, aka portfolio weight - benchmark weight)(return for that security)] SOURCES: (1) active weightings (taking exposures different than benchmark) to rewarded factors (i.e. risks that give LT positive risk premiums - mkt risk / beta, size, value, liquidity) (2) tactical exposure to mispriced securities, sectors, and rewarded risks that generate alpha (i.e. return that can't be explained by LT exposure to rewarded factors) - alpha related to mgmt skill -> generates return by identifying mispricings - return = SUM(sensitivity to rewarded factor x return of rewarded factor) + alpha (aka return from unrewarded factors) (3) idiosyncratic risk - due to luck (i.e. not from mkt exposure or alpha) - comes from concentrated positions (a mgr who spreads the portfolio across many assets likely to minimize idiosyncratic risk and have lower portfolio vol) - active return = SUM (Sensitivity of PORTFOLIO to each rewarded factor - sensitivity of BENCHMARK to each rewarded factor) x (return of reach rewarded factor) + (alpha + idiosyncratic) -> factors expected to generate value (small cap, value, high P/B, etc) will have positive factor exposures, and the flips (eg large cap growth) will have negative

Active share, factor bets, active risk

- active share: how similar the portfolio is to the benchmark in terms of stock holdings - b/w 0 (same as benchmark) and 1 - factor bets: mgr makes factor bets when portfolio's exposure to a risk factor differs from benchmark - taking a factor bet necessitates increased active share, but higher active share does not always mean factor bets have been taken (you might hold stock A instead of B, but two stocks have same factor exposure) - active risk: measures extent to which active return (portfolio return - benchmark return) varies from period to period -> a consequence of active share and factor bets -> if you swap a security in the benchmark with a similar security, active share will increase but active risk will not as much b/c active risk is a measure of volatility of returns and it is likely that new security behaves in a similar way to the benchmark

Active risk & active share for active traders, indexers, and factor investing

- active traders: have high active risk and active share - these are stock pickers (includes pair trading) - indexing: low both - factor: in between these two extremes

Active risk - what it is and formula (2 ways)

- aka tracking error - standard deviation of active returns (portfolio returns minus benchmark returns) -> how consistent was the portfolio's performance relative to benchmark - active risk = SQRT [ (SUM (active return)^2 ) / (T - 1) ] Where T = # of return periods - a mgr can completely control active share b/c they control weights of securities, but cannot control active risk because predicted correlations / variances of securities may be different than realized - active risk = SQRT [ variance (SUM (Sensitivity of PORTFOLIO to each rewarded factor - sensitivity of BENCHMARK to each rewarded factor) x (return of each rewarded factor) + variance (idiosyncratic) ] -> this is v. Similar to extended active return formula that breaks out active return into factor, alpha, and idiosyncratic components (just alpha and idiosyncratic risk are combined into one term here) - active risk increases as factor exposure (active beta) and idiosyncratic risk (i.e. portfolio concentration) levels increase - high exposure to a risk factor leads to high level of active risk -> no net factor exposure will have active risk attributed entirely to active share - active risk attributable to active share is inversely related to # of securities in portfolio - active risk is affected by cross-correlation b/w securities while active share is not - eg if you underweight a pharma stock to overweight another pharma stock, active share will increase b/c weights of portfolio will be diff than weights of benchmark, but active risk will not substantially change if two pharma stocks have high correlation. However, if you underweight a pharma stock to overweight a stock w/ low correlation to it (eg a consumer stock) both active share and active risk will increase

Types of equity income (dividends, securities lending, writing options, dividend capture)

- dividend income: - optional stock dividend - investor can choose b/w cash pmt or stock dividend (i.e. additional shares) - special dividend - one-time cash pmt - securities lending - lending a security to a short seller in return for a fee and collateral / cash that can be invested (short seller has to borrow a security to deliver it to the buyer when the short is made) - CONS: short selling drives down price of underlying (which you are ultimately getting back); must be concerned with credit worthiness of borrower (to return security and to compensate for any dividends received during loan period); lender loses right to vote; administrative costs of securities lending program would reduce total income earned - index funds and large institutional portfolios are best situated for it - writing options - dividend capture - buy stock right before ex-dividend date, hold through ex-dividend date (entitling investor to receive dividend pmt), and then sell stock -> profitable if stock price declines by less than amt of the dividend

Active equity mgmt - factor based strategies (incl. hedged portfolio approach, factor timing), activist, statistical arbitrage, event-driven strategies

- factor based strategies are used by quantitative strategies to identify factors that have a positive association with LT positive risk premium (small cap stocks, value, price momentum, growth, quality) - value added is when mgr can identify which factors are rewarded (receive LT positive risk premium) vs. unrewarded (don't offer persistent return) HEDGED PORTFOLIO APPROACH - rank investable stock universe by factor - divide into quantiles - long the best quantile and short the worst (eg long the smallest 10% of companies and short the largest 10%) - it is hedged b/c it is long and short, but does not necessarily remove mkt risk Cons to hedged portfolio approach: - middle quantiles lost - best performing cos could be in middle - assumes relationship b/w factor and stock return is linear - i.e. as factor increases, expected returns increase by same amount -> any nonlinear relationship not captured - portfolio can appear diversified when using multiple factors to select securities, but if factors are highly correlated, diversification will be less than expected - restrictions on short selling may limit ability to implement strategy - hedged portfolio is not a "pure" factor portfolio b/c it will typically have exposures to other risk factors - factor mimicking: long/short portfolio that is dollar neutral with a unit exposure to a chosen factor and exposure of 0 to other factors - invest in many positions w/o regard to short selling constraints and costs -> can be expensive FACTOR TIMING - equity style rotation: mgr believes different factors work well at different times and will allocate to a factor when its expected to outperform ACTIVIST STRATEGY: take <10% share in company and then pushing for changes (via public proposal, negotiation w/ mgmt, or proxy) that enhance value of stake - target companies with slower earnings or revenue growth (eg underperforming on ROA or asset turnover), negative share price momentum, weak corp governance - defense mechanisms: multi-class share structures with multiple votes to founders; poison pill that allows existing SHs to buy more shares at a discount, diluting shares of activist; staggered board STATISTICAL ARBITRAGE: - quantitative - use technical stock price and volume data to find pricing inefficiencies - pairs trading: identifies 2 securities in same industry that historically are highly correlated - when this relationship breaks down, buys underperforming security and shorts outperforming one -> profits if mean reversion occurs -> risk is that there is a fundamental breakdown from historical correlation (eg fundamental changes in mgmt and strategy) that will cause them to not mean revert - another pairs trading method is to short the spread (ratio of two rated stock prices) when the spread b/w the two is high, and long the spread when it is low - mkt microstructure-based: take advantage of supply/demand imbalances that last for a few seconds (done by HFT) EVENT-DRIVEN - exploit mkt inefficiencies that occur around corp events such as M&A, earnings or restructuring announcements, share buybacks, special dividends, spinoffs - risk arbitrage: buy shares of target company after M&A has been announced - stock price will go up, but not to offer price b/c of risk that deal doesn't go through -> profit is the margin (if a stock transaction, buy shares of target and short shares of acquirer in same ratio as the proposed share exchange of the deal)

Factor-based indexes and adv/disadv vs mkt cap wtd

- factor index can be single factor (eg CAPM) or multifactor (eg mkt risk / beta, firm size, book-to-mkt, profitability, growth rate of assets) - replicate risk/return characteristics of an index by creating portfolio with same exposures to a set of risk factors - high exposure to specific risk factor allow you to augment or replace market cap weighted index based on belief about future returns to various risk factors -> look to increase returns by exploiting out of favor factors (concentrates risk exposure relative to mkt cap wtd) - ADV: are rules based so have lower operating costs than actively managed funds, but still allows for different factor exposures based on investor's expectations (allows you to increase returns relative to mkt cap index) - DISADV: higher costs than passive cap-weighted index funds due to mgmt fees and trading commissions; decision re: factor selection, weighting, and rebalancing are often transparent, allowing others to mimic and potentially reducing oppty for higher returns - single factor models (eg CAPM)

Long-short strategy - gross vs. net exposure

- gross exposure - sum of value of long positions + absolute value of short position - net exposure - difference b/w value of long positions and value of shorts -> both expressed as a % of investor's capital E.g. you raise $100M of capital and invest $80M in longs and short sell $30M - gross exposure is 110% (80% + 30%) and net exposure is 50% (80% - 30%) - will have a cash balance of $50M.. $20M of uninvested on long side and $30M from short sale proceeds

Shareholder engagement - what it is and costs

- investors interacting with companies to potentially favorably impact stock price - includes participating in calls with the company and voting on corporate issues - would involve corporate strategy, capital allocation, corporate governance, mgmt comp, board composition - free riders who don't incur the costs of engagement still benefit - costs: (1) requires an investment of time and resources - active mgrs will do so to improve performance, passive mgrs will look to reduce these costs (passive managers, like index funds, try to minimize costs, so SH engagement will likely increase costs and lower returns); larger investors can more easily absorb these costs as they can spread it out over more assets (2) focus on ST goals like increasing stock price at expense of LT goals (3) acquisition of MNPI, increasing risk of insider trading (4) potential conflicts of interest

mkt impact costs and Slippage + factors that affect it

- market impact cost = implicit cost related to price movement caused by mgrs executing trades in the mkt (eg a mgr buying shares may force prices up, thereby eroding alpha) - slippage = the market impact cost of a single trade. Measured as the difference b/w execution price and the midpoint of the quoted mkt bid/ask spread at the time the trade was first entered FACTORS: - AUM vs. mkt cap of securities - lower trading volume of smaller cap securities is a liquidity barrier to managers with higher AUM (position may be limited to a portion of the avg daily trading volume) -> a fund with a focus on large cap can support a higher level of AUM than one focused on small cap - smaller positions are likely to have less market impact costs b/c lower need for liquidity from mkts - high portfolio turnover and shorter investment horizons increase slippage -> trading more frequently implies higher impact cost b/c you will get some slippage every time you trade - higher mkt volatility leads to higher slippage - information content - manager's trade acts as a signal to the market and encourages other mkt participants to carry out same trade -> increases slippage

Active share - what it is and formula

- measures the degree to which the # and sizing of the positions in the mgr's portfolio are different from those of the benchmark - Active share = (1/2) SUM[ ABS VALUE (weight of security in portfolio - weight of security in benchmark) ] - ranges from 0 (hold benchmark weights) to 1 (holds portfolio of stocks that aren't in benchmark) - equal to the total overweighting or total underweighting - (1 - active share) = % overlap b/w portfolio and the benchmark - if two portfolios have the same benchmark and only invest in benchmark securities, the portfolio with fewer securities will have a higher concentration and therefore a higher Active Share -> i.e active share inversely proportional to # of securities in portfolio - investors assess the fees paid per active share - eg a fund with an active share of 0.2 would be considered expensive vs a fund with active share of 0.5 if both charging the same fees -> willing to pay more fees for high active share (indicator of greater active mgmt) - a mgr can completely control active share b/c can control position sizes

Types of equity costs (mgmt fees, performance fees, admin fees, mkting and distribution fees, trading costs, investment strategy costs)

- mgmt fees - % of AUM and cover comp, research and analysis, hardware /software, etc - performance / incentive fees - performance fee if performance outperforms a threshold - may have high-water mark (if you get fee for outperformance and then portfolio declines, you only earn incentive fee on future appreciation above level previously reached) - admin fees - associated with corp activities like measuring risk/return and voting on company issues - mkting and distribution - employing mkting / sales / client services teams, developing and distributing mkting materials (e.g. brochures), etc. - trading costs - explicit like broker commission, taxes; implicit like bid-ask spread, slippage costs from not completing entire trade due to illiquidity - investment and strategy costs - related to chosen investment strategy (e.g. active funds have higher fees than passive due to more analysis and transactions) - strategy costs - active strategies generally have higher cost (but passive can be hurt by predatory pricing which seeks to buy ahead of passive investors, .(i.e. knowing they are about to add it to the index) - liquidity demands - momentum strategies buy in increasing mkts and sell in decreasing mkt -> this demands liquidity (buying when everyone else is buying) and creates high market impact costs; contrarian investing supplies liquidity by buying in decreasing mkts and selling in increasing

Slippage costs - is it more or less important than commissions? Is is greater for small or large cap securities?

- more important than commissions - greater for small cap

Addressing client constraints through negative screening, positive screening, thematic / impact investing

- negative screening - exclusionary screening - exclude companies / sectors that don't meet client standards (e.g. excluding coal cos that don't meet ESG standards) - positive screening - best in class screening - uncover companies / sectors that are most favorable to client (eg finding cos that score best on ESG) - thematic investing - screen equities based on a certain theme or sector, e.g climate change or clean energy - impact investing - meet investor objectives by becoming more actively engaged with company matters or directly investing in company projects

Risk budgeting

- process by which contribution to total risk of the portfolio is allocated to constituents of the portfolio - contribution to port variance can be done on an absolute or relative basis

Equity segmentation methods (3)

- provides understanding of overall portfolio and diversification (1) size and style - size: large-cap, mid-cap, small-cap - style: growth, value, or a mix (aka blend or core) - determined by looking at P/E, P/B, div yield, earnings or book value growth - stable net income and high div yields are characteristics of value - PROS - better address client risk/return characteristics, greater ability to diversify across different sectors (can target exposure to meet specific objectives), construct relevant benchmark (beyond just S&P 500), analyze how company characteristics change over time - CONS: categories are not stable over time (this is the last pro as well) (2) geography - by stage of economic development - developed mkts (e.g US, UK, Germany, Australia, Japan) - emerging (e.g. Brazil, Russia, India, China, South Africa) -> ie BRIC + S Africa - frontier (e.g. Argentina, Estonia, Nigeria, Jordan, Vietnam) - ADV: can better understand how to diversify across mkts - DISADV: currency risk; may overestimate diversification benefit as companies themselves may already be diversified across geos (3) economic activity - companies grouped into sectors / industries - mkt-oriented approach - segments companies by markets served, how products are used by consumers, how cash is generated - production-oriented approach - segments companies by products manufactured and inputs required during the production process -> companies may land in different buckets depending on which method was used - ADV: analyze / benchmark based on specific sectors; diversification benefit across sectors - DISADV: some companies not easily assigned to one specific sector

Types of style factors - size, value, price momentum, growth, quality, unstructured data

- size: long small cap stocks, short large cap (small cap at more risk of failure so receive risk premium) - value: long cheap - stocks with high book to mkt, high CFs, low valuation mults; short expensive companies - price momentum: long companies that recently outperformed; short companies that recently underperformed - growth: long companies with high historical or expected growth; short: low growth - quality: long high quality earnings (low non-cash accrual earnings or measures relating to changes in debt, profitability, stability, or mgmt efficiency) - unstructured data: combined conventional mkt data with alternative unstructured data (eg satellite imagery, text data, credit card data, social media)

Equity investment style classification and approaches (2)

- style classification to split stock universe into subgroups by style (eg size, value, etc) - groups contain stocks that have high correlation with each other, but correlation b/w groups is low (distinct sources of risk and return) - used to classify style of a portfolio and benchmarking mgrs (1) Holdings-based approach - looks at attributes of each individual stock in portfolio and aggregates those attributes to conclude the overall style of the portfolio - chart with 9 boxes - Style on x (value, blend, growth) and size (small, mid, large) on y - given a style score, assigned a value of 0 to 100, 0 is low and 100 is high -> designed such that # of stocks in each group is the same - shows current exposure of a fund - ADV: generally more accurate b/c using actual portfolio holdings (gives current analysis of style exposure); assesses each individual holding's contribution to style - DISADV: requires availability of all portfolio constituents and style attributes of each; analysis limited if derivatives used; different systems with different Definitions of style will classify Same portfolio in different ways (2) returns-based approach - identify style by a regression of fund returns against a set of passive style indices (e.g. small-cap growth index, large-cap growth, small-cap value, large-cap value) - sum of the slope coefficients = 1, so slope coefficient can be interpreted as mgr's allocation to that style - one based on historical returns would capture average exposure over time - ADV: does not require info on holdings; can be easily and universally applied - DISADV: constraints on outputs can limit detection of extreme styles; uses historical regression so will be more of a backward looking view of mgr's historic style exposures

tracking error and causes - does it increase or decrease as sample size increases?

- tracking error = standard dev of difference b/w index portfolio returns and published index returns (ie vol of excess return) caused by: - mgmt fees - commissions on trades - sampling instead of full replication (ie holding smaller # of securities than index) - intraday trading (index returns are based on closing prices) - cash drag (may hold cash, which decreases returns in rising mkts but increases returns in falling mkts) - less frequent reconstitution - dividends reinvested not same day - controlling tracking error involves a tradeoff b/w higher transaction costs of full replication and increased tracking error that comes with sampling - tracking error initially declines as size of sample increases, but as more stocks are added and you approach full replication, added stocks are less liquid and transaction costs increase tracking error b/c they outweigh gains of increased sample size - one way to reduce tracking error is securities lending - can be used to generate income that can offset mgmt fees - proxy voting can improve company and thus stock returns, but can be a costly undertaking (many passive mgrs use proxy-voting services) - replication mgrs look to create a portfolio that tracks performance and vol of underlying index as closely as possible -> proper measure of skill is tracking error

Characteristics of well-constructed portfolio / risk efficiency

- well-constructed portfolio delivers promised characteristics to investors in cost-efficient and risk-efficient way - should have risk exposures that match investor expectations (eg if you say you are quality focused, does your risk factor contribution support that?) - low idiosyncratic (unexplained) risk relative to total risk When comparing portfolios with same desired characteristics: - one that achieves same desired risk exposures with fewer positions is likely to have more focus on risk mgmt and has higher risk efficiency - if they have same risk factor exposures, one with lower absolute volatility and lower active risk is preferred (assuming similar costs) -> note: active risk is more relevant than absolute volatility - if similar active and absolute risks, similar costs, similar alpha skills.. the portfolio with the highest Active Share is preferable b/w will leverage alpha skill of mgr and have higher expected return - eg achieving lower active risk with fewer securities indicates risk-efficiency

Fundamental law of active mgmt

Active return = information coefficient x SQRT(breadth) x Mgr's active risk x Transfer Coefficient IC = correlation b/w mgr forecasts and realized active returns Breadth = # of independent decision made by mgr EACH YEAR - mgr who considers a single factor defined by a single metric for all investment decisions will have low breadth because not making truly independent decisions (mgr who uses multiple factors with multiple metrics will make more independent decisions) Transfer coefficient = a number b/w 0 (mgr fully constrained) and 1 (no constraints) Mgr's active risk = SD active return; the volatility of active returns

Bottom-up / top-down approaches to active equity mgmt

BOTTOM-UP - use info about individual companies (like profitability or price momentum) to build portfolios by selecting the best individual investments - quantitative bottom-up mgrs looks for relationship b/w company level info (e.g. P/E) and expected return that will persist into the future - fundamental bottom-up mgrs incorporate quantifiable and qualitative characteristics (e.g. business model, branding, competitive advantage, mgmt) TOP-DOWN - info about variables that may affect companies (e.g. macro environment, gov policies) to select best markets or sectors (ie not using info about individual investments) - use broad market ETFs and derivatives to overweight the best mkts and underweight the least attractive (e.g. geo, sector, volatility, thematic - e.g. opps due to new technology, changes in regulation, economic cycles)

Contribution of asset i to portfolio relative/active variance (CAV)

CAVi = (w_pi - w_bi)xRC_ip; ie. (weight of asset i in portfolio - weight of asset in in benchmark) x (Cov b/w active returns of asset i and active returns of portfolio) Where RC_ip = SUM[ (w_pj - w_bj)xRC_ij ]; i.e. (Cov b/w active returns of asset i and active returns of portfolio) = SUM (weight in portfolio of j - weight in benchmark of j) x (Cov b/w active returns of i and j); - so CAVi = SUM [(Wpi - Wbi)x(Wji - Wjp)x(RC_ij)] - NOTE: This is same formula as before but uses active weights and covariance of active returns - asset contributes more to active variance if it has higher active weight and if its active return is related to overall portfolio active returns - contribution to active variance is a function of active risk, not standard deviation - eg cash can have a high active risk despite its low SD due to low correlation of cash vs benchmark - adding up the CAVs for all assets in the portfolio will give variance of the portfolio's active return (AVp) - taking the square root of total variance of portfolios active return gives SD of AR, ie active risk - like absolute risk, relative risk can be done a country, sector, or factor level (and port can be segmented into variance explained by active factor exposures and unexplained active var associated with idiosyncratic risks)

Contribution of factor to absolute portfolio variance

CVi = SUM [ (Beta aka sensitivity of portfolio to factor i) x (Beta j) x (Cov of factor i & j) Or, = (Beta i) x (Cov of factor i and portfolio) Where (Cov of factor i and portfolio) = SUM [ (Beta j) x (Cov factor I & j) ] - NOTE: this is same as asset contribution to portfolio variance but weights replaced by beta and assets replaced by factors - portfolio variance is the sum of each factor's contribution to portfolio variance + unexplained variance - relative contribution / proportion of total portfolio variance explained by factor = absolute variance of factor / total portfolio variance

Contribution of asset i to absolute portfolio variance (CVi), incl relative contribution of portfolio variance

CVi = SUM [ (weight i) x (weight j) x (Cov returns between assets I & j) Or, = (weight i) x (Cov returns b/w asset i and portfolio) Where (Cov returns b/w asset i and portfolio) = SUM [ (weight j) x (Cov returns between assets I & j) ] - portfolio variance is the sum of the absolute contributions to variance of each asset - relative contribution to portfolio variance = absolute contribution to variance of asset A / portfolio variance - this calc can be done with sector -> (weight of sector in portfolio) x (covariance of the sector with the portfolio)

Attribution analysis - (2) parts, & % of excess returns arising from each

Examines excess return - 2 components: (1) active effect due to factor weight: SUM(port weight - bench weight)(port factor return) (2) active effect due to security selection: SUM(port return - benchmark return)(port weight) % arising from active factor weightings = active effect due to factor (or security) / total excess returns

Stock concentration of an index and how to measure

HHI = sum of (weight of stock)^2 - ranges from 1/n (equal weighted portfolio) to 1 (single stock) -> as HHI increases, concentration risk increases - effective # of stocks = 1/HHI

What does active risk say about dispersion vs. benchmark?

Higher active risk = higher dispersion

Types of long/short funds - long extension and market-neutral

LONG EXTENSION - constrained to have net exposure of 100% (e.g. 130% long and 30% short) - this is a constrained from of long/short fund and mgr has no real discretion over gross/net exposure - would be used by investors that want 100% net mkt exposure but want to allow mgr some ability to short to benefit from negative views MARKET NEUTRAL - aim to remove mkt exposure through long/short exposures -> has net market beta of 0 (e.g. long $200M of assets with beta of 0.9 and short $150M of assets with beta of 1.2) - if long and short positions are equal size, gross exposure will be 2x long position value and net exposure is 0 - these funds show lower volatitly than long-only strategies and low correlation with other strategies -> used for diversification purposes rather than seeking high returns - difficult to maintain zero beta as correlations between exposures continually change - pairs trading is a common technique in mkt neutral portfolios - risk of pairs trading is that observed price divergence is not temporary and could be due to structural reasons -> use stop-loss to limit this risk

Can you determine which fund is likely to outperform other based on active share?

No - merely shows how much different weights are than benchmark

Active vs. passive investing; rationale for shifting to active and costs of active

PASSIVE: try to replicate index or benchmark ACTIVE: seek to outperform benchmark and add value - riskier as you may underperform benchmark; also has higher turnover, which can lead to higher tax burden - rationale: (1) confidence mgr has expert knowledge and skill (2) client preferences - need enough investors to give you money to cover costs of investing (but too much capital can make it hard to find opps to add value) - benchmark should contain a broad range of underlying equities with sufficient liquidity to support active mgmt -> narrow limited benchmarks don't give the active mgr much room to deviate while keeping trading costs reasonable and passive approach would be better -> country and sector specific equity funds (e.g. US consumer defensive companies) tend to be passively managed (3) mandates to invest in certain companies (e.g. ESG) may require a more active approach - results are less certain and costs are higher - other risks: (1) reputation risk - from violations to rules, regulations, client agreements, or moral principles (2) key man risk - individuals who are essential leave the fund (3) higher turnover - leads to higher transaction costs

Process to create a fundamental active investment strategy and pitfalls of fundamental investing

Process: define investment universe, prescreen (e.g. value mgr will cut out high P/E multiples), analyze each industry player (eg financials and competitive positioning), forecast / run valuations, construct portfolio with desired risk profile, rebalance as needed Pitfalls: - behavioral biases: fundamental strategies rely on human judgment, which can be affected by biases - value trap: stock that appears to be attractive because of price fall, may in fact be overvalued and decline further -> quality of business may have deteriorated - you need to determine the stock is trading below intrinsic value given the company's future prospects and be able to identify the catalyst that will lead to upward revaluation - growth trap: favorable future growth prospects are already reflected in price -> growth stocks already trade at high multiples, so any drop in growth can lead to significant multiple and price compression

Pros/cons of long/short strategies

Pros - greater ability to express negative views - the most negative a long-only mgr can take it to not hold a security (so max underweighting is the weight of the security in the benchmark) -> short seller not constrained in this way, leading to higher information ratio b/c lower constrained will increase transfer coefficient - ability to use leverage generated by short positions to gear into high-conviction long ideas - ability to remove mkt risk -> provides diversification value vs. other strategies - ability to control risk factors - most rewarded factors (size, value, momentum, etc) are obtained via long/short portfolio Cons - unlimited potential losses - negative exposures to risk premiums - leverage magnifies losses as well as gains - losses can be greater because you can be incorrect on both the long and short position; long only can only be wrong on long position - cost of borrowing securities can become too high, particularly for securities that are difficult to borrow - losses on short position will increase collateral demands from stock lenders, which may force mgr to liquidate at unfavorable price. Also vulnerable to "short squeeze" where the sudden rise in price of a heavily shorted security forces short-sellers to cover positions, buy back shares and potentially force share price even higher - lenders can recall shares at an inopportune time, causing disruption to mgr's strategy

Rebalancing (and when it is required for each type of index) and reconstruction

Rebalancing - adjusting portfolio weights for index weightings change - equal weighted - weights no longer equal once prices change - price-weighted - stock splits and stock dividends - market cap weighted - issue new shares / repurchase shares Reconstitution - removing and replacing stocks that don't fit market exposure of index reduce trading costs via: - buffering: threshold level of change in firm's mkt cap before moving from one index to another, eg from small to mid cap - as long as stocks remain in buffer, they remain in index, which can cause holdings to exceed holdings of the index - packeting (when mid cap moves to large cap, half of portfolio is moved to large cap on reconstitution date; if still meets large cap criteria at next reconstitution date, the rest of the position is moved) - rebalancing / reconstitution incur trading costs and decreases returns - derivatives / futures preset a cost efficient rebalancing method -> can get exposure to desired ranges without impacting mgrs

How to calc portfolio turnover Eg: $120M BoP, $145M EoP, $130M avg assets. $18M of purchases, $25M of sales

The lower of purchases or sales divided by avg monthly net asset 18/130 = 14%

Activist investing

goes a step beyond SH engagement: - propose shareholder resolutions and launch media campaigns to influence the vote - seek representation on the board - launch proxy fights (seek to persuade other shareholders to support their proposals) - take small stakes in companies (less than 10%)

Are small or large cap stocks better for active mgmt?

small cap stocks - more public info is available on large cap stocks, making efficient pricing more likely and generating excess returns less likely - less research on small cap stocks

Characteristics of 4 investment approaches (systematic vs. discretionary and bottom-up vs top-down)

systematic vs. discretionary - the degree to which the mgr follows systematic rules, rather than using discretionary judgment - discretionary mgrs more likely to engage in factor timing, hold concentrated portfolios, and less likely to use formal portfolio optimization techniques - systematic approach is designed to extract a premium through balanced exposure to known, rewarded factors - incorporates research-based rules across a broad universe of securities - discretionary looks for active returns from firm specific factors, such as pricing power and competitive landscape Bottom-up vs. top-down - bottom up stock-specific info vs. macroeconomic info Systematic, top-down: emphasize macro rewarded factors, factor timing possible but rare, diversified, portfolio optimization used, few mgrs in this category Systematic, bottom-up: emphasize security specific, no factor timing, diversified across, formal optimization used Discretionary, top-down: emphasizes macro rewarded factors, most likely to use factor timing, diversified or concentrated, less formal portfolio construction Discretionary, bottom-up: emphasize security specific, potential factor timing, diversified or concentrated, less formal portfolio construction - discretionary is more likely to use factor timing, make concentrated bets (and so more exposed to idiosyncratic risk), and less likely to do portfolio optimization - systematic design portfolio to extract risk premiums from a balanced exposure to known, rewarded risk factors - achieved using broadly diversified portfolios (so low idiosyncratic risk) -> targeting low idiosyncratic risk with low concentrations indicates a systematic approach - note: factor timing is difficult to use in a systematic, rules-driven way (so more likely with discretionary). Signals used to generate timing ideas are usually top-down vs. bottom-up -> so discretionary, top-down most likely to use factor timing techniques