Cash flow forecasting

Cashflows may be forecast directly, as well as by several indirect methods.

The direct method of cash flow forecasting schedules the company's cash receipts and disbursements (R&D). Receipts are primarily the collection of accounts receivable from recent sales, but also include sales of other assets, proceeds of financing, etc. Disbursements include payroll, payment of accounts payable from recent purchases, dividends and interest on debt. This direct R&D method is best suited to the short-term forecasting horizon of 30 days ("or so") because this is the period for which actual, as opposed to projected, data is available.[3]

The three indirect methods are based on the company's projected income statements and balance sheets.

• The adjusted net income (ANI) method starts with operating income (EBIT or EBITDA) and adds or subtracts changes in balance sheet accounts such as receivables, payables and inventories to project cash flow.
• The pro-forma balance sheet (PBS) method directly uses the projected book cash account; if all the other balance sheet accounts have been correctly forecast, cash will be correct, also.
• The accrual reversal method (ARM), is similar to the ANI method. Here, instead of using projected balance sheet accounts, large accruals are reversed and cash effects are calculated based upon statistical distributions and algorithms. This allows the forecasting period to be weekly or even daily. It also eliminates the cumulative errors inherent in the direct, R&D method when it is extended beyond the short-term horizon. But because the ARM allocates both accrual reversals and cash effects to weeks or days, it is more complicated than the ANI or PBS indirect methods. The ARM is best suited to the medium-term forecasting horizon.[4]

Both the ANI and PBS methods are suited to the medium-term (up to one year) and long-term (multiple years) forecasting horizons. Both are limited to the monthly or quarterly intervals of the financial plan, and need to be adjusted for the difference between accrual-accounting book cash and the often-significantly-different bank balances.[5]

The simplest method is to have a spreadsheet that shows cash coming in from all sources out to at least 90 days, and all cash going out for the same period. This requires that the quantity and timings of receipts of cash from sales are reasonably accurate, which in turn requires judgement honed by experience of the industry concerned, because it is rare for cash receipts to match sales forecasts exactly, and it is also rare for customers all to pay on time. (These principles remain constant whether the cash flow forecasting is done on a spreadsheet or on paper or on some other IT system.) A danger of using entirely theoretical methods in cash flow forecasting for managing a business is that there can be non cash items in the cashflow as reported under financial accounting standards. This goes to the heart of the difference between financial accounting and management accounting.

For short term cash flow forecasting there are a number of AI driven low cost software applications available. While AI tools can offer automation and predictive capabilities, there are limitations to their accuracy, especially in areas that involve human behaviour or subjective factors. For instance, predicting when customers will pay their bills accurately can be challenging due to various factors that influence payment behaviour. AI tools heavily rely on historical patterns and predefined rules, which may not always capture the complexities of real-world situations accurately.

Moreover, AI tools may lack the flexibility and customization options provided by Excel. They are typically designed to work within predefined frameworks and may not easily adapt to unique business requirements.

If a business is running to the limits of its overdraft and cash management is an issue, using an app is not the best way to manage it. An app, even an AI enabled application will only look at historic data and make predictions using learned patterns. The app will not be able to tell you when a customer will pay his bill, it might guess, but will invariably be wrong. A business running close to the wire, needs a human to guide it, only a human being will know who is screaming loudest for a payment, who must be paid to keep the show on the road.There is only one tool fit for purpose, Excel, with Excel you can cut and paste and move predicted payments and receipts with ease so that your balance will not exceed your facilities. It will take more work than relying on an app, but it will be more accurate and if you don’t need that accuracy then you don’t need the app.