CTP - Chapter 14 - Cash Flow Forecasting

Managing Liquidity

Forecasting the net cash position at different intervals to identify potential cash excesses or shortages is critical for scheduling investment decisions and anticipating borrowing requirements to meet daily obligations. Liquidity management is the most important motive for cash flow forecasting.

Budgeting Capital

Forecasts of revenue, expenditures, and funding are not only helpful for developing cash forecasts, but are also needed for the evaluation of capital investment alternatives (i.e., capital budgeting).

Known Cash Flows

The timing and amounts of certain cash flows are fixed and known in advance. Common examples include interest expense, principal repayments, dividends, royalties, rent, and tax payments.

Steps for Establishing most Important Forecasting Method

These include establishing data relationships, selecting the method, testing and validating relationships, and using technology.

Degree of Certainty

Another way to aggregate cash flows is based on a given cash flow's degree of certainty.

Maximizing Returns

A cash forecast provides the timing and amounts of anticipated cash surpluses and cash deficits. Short-term investment returns are maximized when treasury professionals have time to select optimal investment instruments and maturities. If the firm follows a matching strategy for short-term investments, the maturity of an investment will be matched with the timing of future cash deficits. Knowing that the firm does not expect a cash deficit for a certain period of time allows the treasury professional to extend investment maturities and earn higher yields.

Completed Forecast

A completed forecast combines the receipts and disbursements schedules, and compares the result to a desired minimum cash balance.

Distribution Forecast

A distribution forecast uses historical patterns to allocate proportions of total cash flow over a time period. For example, a retail store may have projected sales of $ 1,000,000 for a one-month period. Historical patterns indicate that in an average month, 20% of sales occur in week one, 40% of sales occur in week two, 30% of sales occur in week three, and the remaining 10% of sales occur in week four. This pattern may be used to forecast weekly sales of $ 200,000 in week one, $ 400,000 in week two, $ 300,000 in week three, and $ 100,000 in week four. Distribution percentages are commonly calculated with simple averages and regression analysis.

Invest the Appropriate Amount of Resources

A forecast that takes a few minutes first thing every day and provides a result that is 90% accurate most of the time may be sufficient for many organizations. Only when there is more pressure for accuracy, to comply with covenants or manage limited liquidity, for example, will it be worthwhile to invest additional resources for a more accurate forecast.

Select a Method

A forecasting method is selected after the data relationships are specified. For example, the treasury department may select a receipts and disbursements method for projecting the daily cash position, or an exponential smoothing method to forecast cash inflows from sales. (Both methods are discussed later in this chapter.) Forecasting methods should be cost effective.

Projected Cash Closing Position

A projected closing cash position statement, also known as the daily cash forecast or the cash report, is one particular type of cash forecast that is used by many treasury professionals. The projected closing position is used to determine whether there will be a cash surplus or a cash deficit in a particular bank account. A projected closing cash position is calculated by taking the day's opening available bank balance( s), plus expected settlements in the collection and concentration accounts (depository banks), and minus projected disbursement totals.

Rolling Forecasts

A rolling forecast has a constant number of periods (weeks, months, etc.) and is updated on a regular basis. That is, this type of forecast "rolls" forward. An example of a rolling forecast would be a 13-month cash forecast that is updated monthly. Assume that the initial forecast covers the 13 months from January 2015 through January 2016. Sometime during January 2015, the forecast will be updated and will now cover the 13 months from February 2015 through February 2016. If nothing has changed, then January 2015 is dropped from the forecast and the projection for February 2016 is added. In reality, the forecast data for the other months in the report are typically also updated because newer information is now available.

Simple Moving Average

A simple moving average bases a forecast on a rolling average of past values for a series. The method has the advantage of being easy to use, but since a forecast is developed using an average of past values, it will lag a trend if one exists and will tend to dampen or smooth turning points where seasonality or another irregular event is present. The larger the number of data points used to form the average, the greater the chance that random movement will be eliminated through averaging. However, the margin by which the trend will be missed will be greater, and the dampening of seasonal turning points will be stronger.

Times Series Forecasting

A time series is a serial chain of values for some variable, such as sales or cash flows. A time series contains a trend, seasonal pattern, cyclical pattern, and random movement. A trend is the general direction that the values are moving (e.g., sales are increasing 5% year over year). A seasonal pattern repeats with regard to time of year. A cyclical pattern repeats without regard to time of year. It is possible that the series is stationary (i.e., the trend value is zero). The seasonal and cyclical patterns may be strong, moderate, or weak. Random movement is a change in the series not identifiable as a trend, or a seasonal or cyclical influence. The source of random movement is usually unknown, but may be triggered by impactful random events (e.g., a terrorist attack).

Historical Data

A variety of data from prior periods are useful in predicting the amount and timing of cash inflows and outflows. Credit sales, payment histories, purchases, and scheduled payments are examples. Historical data are used to analyze trends and seasonality in the firm's cash flows.

Analytical Forecast

Also referred to as simulations, are used to answer what-if questions or to predict the financial impact of a given action. They are particularly useful in strategic planning around things like capital investments or tax planning.

Accounts Receivable (A/ R) Balance Pattern Forecast

An A/ R balance pattern may be used to forecast collections from credit sales. The A/ R balance pattern specifies the percentage of credit sales during a time period (e.g., one month) that remain outstanding at the end of the current and subsequent time periods. The A/ R balance pattern is used to determine a collection pattern that is used to forecast cash inflows. An average percentage collected during a given month is calculated by subtracting the percentage outstanding at the end of the month from the percentage outstanding at the beginning of the month.

Predictive Forecasts

Attempt to predict or project what will happen in the future. They can be used to answer questions such as "How much cash will I have available to invest over the next week?" or "Will I need to borrow money against my line of credit in the next week?"

Bank Account Structure

Bank account structure is also important in that it may help group cash flows in a meaningful fashion and help identify potential changes. For example, using a concentration account to collect cash inflows and disburse cash outflows via ZBAs helps to facilitate cash flow forecasting compared to a system of multiple, stand-alone accounts.

Establishing Data Relationships

Before selecting the appropriate forecasting method, treasury professionals should ascertain the statistical relationship between the available data and the cash flow components to be forecasted. For example, cash disbursements can be related quantitatively to invoices received for payment, and cash collections can be related to prior-period credit sales.

Predictable Cash Flows

Cash collections from credit sales are an example of a cash flow component that can be predicted with reasonable accuracy. The cash flow on a given day depends on factors such as the level and pattern of past sales, age of accounts, number of customers, and payment histories. For certain firms, these factors enable reasonably accurate cash flow prediction. Disbursement for payroll and benefits tends to be very predictable because it is based on the number of employees, compensation method, and historical experience. Similarly, vendor check-clearing patterns can be predicted based on experience with clearing times.

Meeting Compliance Requirements

Cash forecasts are frequently part of internal control procedures needed to comply with loan covenants, meet minimum capital requirements, or meet requirements for imprest tax accounts. Compliance is especially important for publicly traded companies.

Meeting Strategic Objectives

Cash forecasts are used to project future funding requirements and make operating decisions to support the strategic plan.

Managing Currency Exposure

Cash forecasts attributable to foreign operations are used to assess the degree of foreign currency exposure and provide information for policies designed to control currency risk.

Managing Costs

Cash forecasts can help minimize excess bank balances, reduce short-term borrowing costs, and increase short-term investment income.

Medium-term Forcasting

Cash forecasts from three to twelve months in length are an integral part of cash budgeting and are considered medium-term forecasts. These forecasts project the inflows (i.e., collections from sales and other sources of funds) and outflows (i.e., expenses and other uses of funds) on a monthly or quarterly basis. Medium-term forecasts are used to determine the firm's need for short-term credit or the availability of excess cash for short-term investing. They also serve as a benchmark for performance by comparing actual cash flows to projected cash flows based on the cash budget.

Ongoing Validation

Continuing feedback from comparisons of projected versus actual values allows continuous evaluation and improvement of the cash forecast. This ongoing validation may be performed on a daily basis for critical forecasts. Any reporting should include an analysis of forecast errors to help foster this process.

Correlation Analysis

Correlation analysis involves identifying the degree of association between two variables. Understanding the variables that are strongly correlated with cash flow is important as these variables can be used to produce future cash forecasts. For example, there is usually a strong positive correlation between a firm's sales and its cash flows. Once a correlation or relationship between cash flow and another variable is established, then cash flow can be forecasted using the other variable. This would typically be accomplished with regression analysis. Regression analysis is a statistical method that can be used to assess the impact that a given independent variable or driver variable has on a dependent variable.

Consistency

Due to the many uses of a cash forecast and the similarity to financial statements, non-treasury departments often develop forecasts of their own. It is important to realize that while the end purposes of the forecasts are different, the information and assumptions used are usually similar and should be consistent.

Exponential Smoothing Formula

Exponential smoothing forecasted values are calculated using the following equation: Next Period Forecast = (Alpha * Current-Period Actual) + [(1 - Alpha) * (Current-Period Forecast)] As with a simple moving average forecast, exponential smoothing will yield a forecast that lags a trend in the series. If a trend in the series is mild, the lag effect will be minimal because, unlike a moving average, the forecast is made using values from the most recent period.

Data Identification and Organization

Identifying suitable data is an important part of the cash flow forecasting process. Omitting relevant data may lead to unexpected cash shortages or surpluses.

Responsibility

In most companies, short-term forecasting is the responsibility of treasury while longer-term forecasts are often done by FP& A or the capital budgeting unit.

Available Information

Information may be gathered externally or internally. Sources include the firm's banks, field managers, sales managers, and the A/ P, A/ R, payroll, FP& A, and tax departments. In many cases much of this data is available from internal systems, such as general ledger (G/ L) systems, enterprise resource planning (ERP) systems, or treasury management systems (TMSs).

Documentation of the Process

It is important to document how the cash forecast process works, as well as documenting the sources of data, contact information for key personnel involved in the forecast, etc. This enables others to understand and work with the forecast, and allows future changes when needed.

Forecast by Currency

It is important to note that for firms that conduct business in more than one currency or country, forecasts need to be made by currency. Shortfalls in one currency are not automatically offset by excesses in another currency unless specific plans to do so have already been made and put into place.

Assumptions

It is important to understand the assumptions underlying the various projected input data to determine how accurate and useful it may be. Cash flows should be discussed with key participants to gain insight into any issues that need to be included in the forecast. For example, it may be important to know whether specific payments are made monthly, quarterly, or annually. As another example, retailers might expect that 80% of their projected annual sales occur in the last two months of the year.

Long-term Forecasting

Long-term forecasts cover any period beyond one year. These forecasts take into consideration projections of long-term sales and expenditures, as well as market factors. Long-term forecasts are of strategic importance because they help inform decisions about which type and amount of long-term financing to obtain, and the timing of obtaining funds. Financial institutions and rating agencies also use long-term forecasts for credit analysis and evaluation purposes.

Medium- and Long-term Forecasting

Medium- and long-term forecasting involves the development of pro forma financial statements. Such forecasts are typically used in strategic planning or FP& A functions. A common approach used to construct pro forma financial statements is based on the percentage-of-sales method.

Validate the Forecast

Once the forecast period is over, treasury professionals should compare the forecast to actual results and determine the reason for any significant differences. Where possible, those reasons should be incorporated into future forecasts to improve long-term accuracy.

Testing and Validating Relationships

Ongoing testing and validation of the forecast are critical steps in the cash forecasting process. Initial validation of a cash forecast and its internal assumptions is needed to ensure that the forecast provides reasonable predicted values. Ongoing feedback and testing are also important to ensure that the forecast continues to produce reasonably accurate forecasts.

Use Appropriate Detail

Perhaps the most important requirement is to make sure positions are forecasted in sufficient detail to meet the firm's requirements.

Controlling Financial Activities

Variance analysis that compares actual cash flows with projected cash flows can help identify problems such as unanticipated inventory changes, delays in A/ R collection, the mistiming of payments, and fraud or embezzlement. Early identification signals management to initiate corrective measures.

Receipts and Disbursements Forecast

Projecting receipts and disbursements is the core of short-term cash flow forecasting. This forecast begins with separate schedules for cash receipts and disbursements. The schedules are prepared on a cash basis, rather than on an accrual basis, due to the need to forecast cash rather than earnings; this is especially important for firms with high levels of A/ R and A/ P. A good place to start is with the historic information provided in bank account statements, which by their nature are cash statements.

Using Technology

Ranging from spreadsheets to sophisticated forecasting programs, technology provides tools to make forecasting faster and more accurate. Spreadsheets provide the ability to build simple to intermediate cash flow forecasting models. The advantages of spreadsheets include availability, ease of use, graphic capabilities, and relatively low cost. Spreadsheets can also accommodate various scenarios for the cash flow forecast. There is risk associated with the use of spreadsheets. Formula and logic errors, as well as broken links, may create undetected errors in the forecast. As the data and models grow more complex, treasury professionals turn to stand-alone forecasting applications, or forecasting tools in ERP systems and treasury workstations.

Short-term Forecasting

Short-term forecasts typically range from one to ninety days in length and predict cash receipts and disbursements, as well as the resulting balances, on a daily, weekly, monthly, or quarterly basis. They aid in scheduling cash transfers, funding disbursement accounts, and making short-term investing and borrowing decisions. Short-term forecasting is also important for establishing and managing target balances for purposes of bank compensation.

Disclose Assumptions

Significant assumptions (e.g., sales will increase by 5% per year) need to be disclosed, and the source of any information should also be documented.

Less Predictable Cash Flows

Some cash flows are difficult to forecast. Examples include cash flows related to sales of a new product, unexpected repairs, or pending settlement of insurance claims. Less predictable cash flows also include the timing of marketing costs, travel, and bonuses. Collaboration with other functional areas of the organization may reduce the uncertainty associated with these cash flows. The experience and judgment of the forecaster are important in such situations.

Source of Information

Source identification is affected by the degree of centralization in the firm's cash management structure. In a decentralized firm, field managers usually have the most current financial data related to their operations. However, a centralized firm is often less dependent on numerous field sources.

Statistical Methods in Forecasting

Statistical methods are used for both short-term and long-term forecasting. There are many statistical approaches that are available for cash flow forecasting purposes. One of the most general statistical methods is that of time series, or extrapolation, where past trends are identified and used to predict the pattern of future cash flows. Simple moving averages and exponential smoothing are two of the most common applications of time series forecasting techniques. Other popular statistical techniques include correlation and regression analysis.

In-Sample Validation

The cash forecast is tested for accuracy using the historical data used to develop it. For example, if three years of monthly data are available, the first thirty months of data could be used to develop the forecast. The forecast could then be used to "predict" the cash flows for those same thirty months, and those predictions can then be compared to actual values for those months. If the forecast accurately predicts one data series from one or more other series, the relationships among the data series are deemed to be validated. When a forecast is initially developed, however, there typically is no other data available to test it.

Out-of-Sample Validation

The cash forecast is tested using data that were not used to develop it. For example, if three years of monthly data are available, the first thirty months of data could be used to develop the forecast. The forecast could then be used to "predict" the cash flows for the subsequent six months, and those cash flows can then be compared to actual values for those months.

choice of cash flow forecast method

The choice of cash flow forecast method to be used depends on its purpose, the forecasting horizon or timeframe, and the frequency with which the forecast is updated.

Desired type of Forecast

The data collected should be dictated by the desired cash forecast, not simply by the data that are available. For example, in preparing a short-term cash needs forecast, immediate A/ P data may be important, but long-term capital acquisitions and planned securities offerings are probably not needed.

Exponential Smoothing

The exponential smoothing technique produces a forecasted cash value based on the most recent actual value, the most recent forecasted value, and a number between zero and one that is used to weight these two values. The weight, designated by the Greek letter alpha (α), is referred to as the smoothing constant and is calculated using a computer program such as a statistics package. This type of function is often built into many cash flow forecasting and TMSs. The program selects the alpha value that weights past actual values and past forecasted values to produce the most accurate forecast of the cash flow for the next period. The model is developed using in-sample validation.

Simplicity

The goal of a cash forecast is to create a reasonable snapshot of the projected cash position at a point in time. The key word here is reasonable. The necessary degree of accuracy is a function of the purpose of the forecast. A short-term forecast used to manage daily cash position should be more precise than a long-term forecast used to determine the availability of capital for future expansion of a production plant. But neither forecast needs to be accurate to the $ 0.01. The ultimate goal is to develop a cash forecast with a degree of detail that is commensurate with its ultimate purpose. That is, treasury personnel should weigh the benefits of a more accurate cash forecast versus the cost of improving accuracy.

Ensure the Forecast is Usable

The investment in developing the cash forecast has to be justified by the outcome. It is no use investing lots of time and effort producing a forecast that is unusable because it is too late to act upon it. In addition, the forecast must be sufficiently robust to provide meaningful data.

Use the Appropriate Platform

The most appropriate platform should be used to develop the forecast. When the cash flows are relatively stable and known in advance, there is very little need for sophisticated specialty tools to analyze the input data. On the other hand, where there are major fluctuations in data and significant variables involved, it may be necessary to use a system or software with greater processing power to generate the forecasted positions.

Communication

The need for collaboration with other units in the company, such as accounts payable, accounts receivable, tax, and financial planning and analysis (FP& A), creates a need for effective communication, clear expectations, and common terminology.

Percentage of Sales Method

The percentage-of-sales method involves projecting the income statement items as a percentage of sales. Likewise, certain items on the balance sheet, such as inventory, A/ R, and A/ P, are also projected as a percentage of sales. These percentages are then used in conjunction with the forecasted sales level to produce a forecasted income statement, balance sheet, and statement of cash flows.

Percentage of Sales Method (Three Steps)

The percentage-of-sales method requires three steps: 1. Forecast the income statement and balance sheet based on the relationships mentioned previously. 2. Calculate the projected ending cash balance by determining how the forecasted income statement and balance sheet values impact cash (i.e., utilize a cash flow statement given the impact of changes to investments, changes to capital/ depreciation, and potential slowing of cash collections or payments). 3. Compare the projected ending cash balance with the firm's target cash balance, and adjust the pro forma statement to show the source of funding for a cash shortfall or the investment of a cash surplus.

The Purpose of Time Series Model

The purpose of a time series model is to identify repetitive patterns in a historical series, so that this pattern can be used to predict future values of the series. There are many kinds of time series models.

Receipts Schedule

This schedule consists of a projection of collections from customers (e.g., cash sales or payments on A/ R) and other cash inflows (e.g., interest or dividends received from investments). It should also include expected, nonrecurring cash inflows, such as the proceeds from asset sales and external financing activities.

Disbursement Schedule

This schedule involves forecasting the cash disbursements for purchases and other cash outflows, such as payroll, taxes, interest, dividends, rent, and debt repayments. A/ P systems are typically a good source for this information, but A/ P data must be adjusted to recognize that payments typically do not clear immediately after the actual disbursement.

Short-term Cash Flow Forecasting Methods

Three of the more important methods include an A/ R balance pattern forecast, a receipts and disbursements forecast, and a distribution forecast.

Reporting Requirements

To ensure the usefulness of the data selected, it is essential that the data are defined precisely and reported accurately and in a timely manner. Reporting should include an analysis of variances between the forecast and actual results to help track accuracy and improve future forecasts.

Cash Flow Components

To produce an accurate cash forecast, it is helpful to separate cash flow into its key components. One approach is to aggregate cash flows by type of activity: operating, investing, or financing. These categories can be segregated further into inflows and outflows. As a general rule, it is better to begin with broad categories then refine the categories as more information about the inflows and outflows becomes available.

Cooperate and Communicate

Treasury depends on other business units for accurate and timely information, and in turn the business units are often dependent on the reliability of cash forecasts provided by treasury. Clear expectations and common terminology should be established from the start of any forecasting effort.

Distribution Percentage (Regression Analysis)

While most distribution forecasts can be performed using a simple average approach, when there are multiple factors involved (e.g., day-of-the-week or -month effects, or holidays) a more sophisticated approach, such as regression analysis, may be needed.

Distribution Percentage (Simple Averages)

With a simple average, the distribution percentages can be estimated by averaging past cash flows related to a particular category of disbursements. For example, a treasury professional may sample past clearings of A/ P disbursements and discover that, on average, 40% of a certain category of disbursements cleared the day the payment was issued, 50% cleared the day after that, and 10% cleared after two or more days. The problem with taking a simple average is that actual payment posting dates may be influenced by more factors than the number of days after issuance.

Static Forecasts

a static forecast is one in which the time period of the forecast does not change but remains static even though the actual numbers or data in the forecast change over time. The period of time used in the forecast remains constant, and the forecast is updated at the end of the forecast period to show the next period. A monthly (or quarterly) cash forecast that is done each month (or quarter) is an example of a static forecast. An annual budget is another common example of a static forecast in that the budget, which is a type of forecast, is normally always for the same 12-month period (e.g., January through December), even if the numbers are updated during the year.