Chapter 12 Cash Forecasting

1
Chapter 12Cash Forecasting
Order Order Sale
Cash Placed Received
Received

Accounts Collection lt
Inventory gt lt Receivable gt lt Float
gt

Time gt Accounts Disbursement
lt Payable gt lt
Float gt Invoice
Payment Cash
Received Sent
Paid
2
Objectives of Cash Forecasting

• Liquidity Management
• Financial Control
• Strategic Objectives
• Capital Budgeting
• Cost Management
• Currency Exposure

3
Steps in the Cash Forecasting Process

• Step 1
• Forecast Horizon and Intervals
• Step 2
• Variable Identification
• Step 3
• Modeling the Cash Flow Sequence
• Step 4
• Model Estimation
• Step 5
• Model Audit

4
Short-Term Forecasting Horizon

• Daily or weekly forecast for a period of up to a
  month.
• Aids in cash concentration transfers, funding
  disbursement accounts, and making short-term
  borrowing and investment decisions.
• Helps in setting and managing balances used for
  bank compensation.

5
Medium-Term Forecasting Horizon

• Also referred to as cash budgets.
• Forecast for 1 to 12 months.
• Used to determine the need for short-term credit
  or availability of funds for short-term
  investments.
• Can be used as a benchmark of performance by
  comparing actual cash flows to forecast cash
  flows.

6
Long-Term Forecasting Horizon

• Forecast covers any period beyond one year.
• Are strategic forecasts used in long-term
  financial planning.
• Used by financial institutions and rating
  agencies for credit analysis and evaluation.

7
Forecasting Philosophy

• Number and type of forecasts
• Expenditure on forecasts
• External versus internal forecasts
• Quantitative versus judgmental forecasting

8
Degree of Certainty

• Certain Flows
• Predictable Flows
• Less Predictable Flows

9
Data Identification

• Sources
• Identification
• Account Structure
• Reporting Requirements
• Historical Data

10
Forecast Method Selection

• Establishing Data Relationships
• Selecting a Method
• Testing Relationships
• Managing the Costs of Forecast Systems and Data

11
Forecasting Methods

• Methods for Short-Term Forecasting
• Receipts and disbursement forecast
• Distribution forecast
• Modified accrual
• Methods for Medium- and Long-Term Forecasting
• Pro forma statement
• Adjusted net income

12
Statistical Tools

• Causal methods
• Regression
• Time-series methods
• Moving average
• Exponential smoothing
• Time-series regression
• Model estimation
• Model audit

13
Forecasting MethodsReceipt and Disbursement

• Receipt Schedule
• Disbursement Schedule
• Both schedules are prepared on a cash basis.
• Completed forecast indicates the projected
  deficiency or surplus of funds in relations to a
  companys minimum cash requirement.

14
Receipts Disbursements Forecast
15
Distribution Method Forecast

• Provides estimates of the cash flow effect of a
  single event, on a daily basis over a specified
  interval based on historical patterns.
• The distribution method is particularly
  appropriate for short-term forecasts.

16
Forecasting Methods Distribution Forecast

• The total estimated cash flow is spread over the
  days in the forecast horizon using proportions
  that are using actual historical patterns.
• Easy and inexpensive to prepare.
• Allows the incorporation of seasonality and
  trends.
• Large data requirements for proportion estimation.

17
Distribution Method Forecast

• Example
• A company has used regression analysis to
  estimate the proportion of dollars that will
  clear on a given business day. It has determined
  that this proportion depends on the number of
  business days since the checks were distributed.
  The estimated proportions are given below.

18
Distribution Method Forecast

• Therefore, if 100,000 in checks are distributed
  on Wednesday, May 1, the checks are estimated to
  clear according to the schedule below.

19
Forecasting Methods Pro Forma Statements

• Based on the percentage-of-sales method.
• Requires a sales forecast.
• Determine balance sheet and income statement
  items that can be assumed to be a constant
  percentage of sales.
• Assume other balance sheet and income statement
  items are constant or can be updated based on
  available information.

20
Pro Forma Statements (continued)

• Generate projected balance sheet and income
  statements.
• If projected assets are greater than projected
  liabilities and equity, there is a projected cash
  shortage.
• If projected assets are less than projected
  liabilities and equity, there is a projected cash
  surplus.

21
Statistical Forecasting Time Series Forecasting

• Simple Moving Average
• Gives equal weight to past observations
• Will always lag any trend in actual cash flow.
• Approach can be useful in identifying cycles and
  patterns of past data but does not take these
  patterns into account in the forecast

22
Simple Moving Average - Example -

• Given the information in the table below, what
  would the 4 week simple moving average be for
  week 7?

23
Simple Moving Average - Example -

• Given the information in the table below, what
  would the 4 week simple moving average be for
  week 7?

269,000
(234,000 342,000 256,000 244,000)/4
1,076,000/4 269,000
24
Time Series (continued)

• Exponential Smoothing
• Allows the cash flow forecast to be adjusted by
  the prior period forecast error.
• Allows for more weight to be placed on the most
  recent actual cash flows by choosing a smoothing
  constant close to 1.
• Forecaster must select the smoothing constant.
• Simplest version of the technique will lag trends
  in the data.

25
Exponential Smoothing - Example -

• Assuming the actual cash flow for week 5 is
  355,000, and the forecasted value for that week
  was 300,000, what is the exponential smoothing
  forecast for week 6 assuming ? .35?

Next Period Forecast ?Current Period
(1-?)Current Period Forecast Next Period
Forecast .35355,000 (1-.35)300,000 Next
Period Forecast 124,250 (.65)300,000 Next
Period Forecast 124,250 195,000 Next Period
Forecast 319,250
26
Regression Analysis

• Regression analysis is a statistical technique
  that systematically identifies the relationship
  between the variable being predicted (the
  dependent variable) and selected or appropriate
  explanatory or independent variables
• The most basic form is that of a simple linear
  regression model. This approach essentially
  determines a line that best represents the
  relationship between a single dependent variable
  and a single explanatory variable
• Regression analyses (especially the more complex
  types) are most often done using a computer
  program

27
Differentiate

• Differentiate among three basic methods of
  validating a forecast.

A method which tests how well a model works by
using the historical data from which the model
was developed. A method which tests the accuracy
of forecasting models using data not involved in
the models estimation. A method which uses
continuous feedback from projected versus actual
comparisons.
in-sample validation
out-of-sample validation
ongoing validation