Cash Forecasting: A Standard Process?

“We’re just a bog standard treasury…” is an apparently innocuous phrase that should in fact set off an alarm bell or two in the minds of technology vendors and consultants who are analysing a prospective client treasury’s operations and requirements. In practice, there are several business areas – notably the accounting treatment and rules, the financial risk management policy and the cash forecasting framework – where detailed analysis usually reveals unique or specific requirements for each corporation.

Cash forecasting, as the forward-looking component within the broad area of cash visibility, is a topic that has increased in importance at senior management levels. The value of minimising working capital demands and credit facility utilisation and other borrowings, and of assuring liquidity and funding continuity, is clearly perceived as many major economies navigate a fragile economic recovery. The relative importance of cash forecasting to treasury managers is emphasised in Deloitte’s 2009 Global Treasury Management Systems Survey: in their analysis of treasury concerns Deloitte finds the leading primary concern to be cash flow forecasting (27%), with liquidity management in second place with 24%.

In the general field of cash forecasting, there are a number of factors that in practice combine to make each company’s cash forecasting requirement unique. These include matters such as company organisation, business cycle patterns and treasury cash management policy and the detailed corporate information management infrastructure that delivers the forecast information.

Company organisation

The geographic and commercial diversification of the company, and especially the degree of centralisation, very much define the complexity of structuring the forecast, and of collecting the data.

Business cycle

The longer-term business cycle of the company or industry impacts the strategic pattern of cash flows, and therefore the requirements of cash forecasting. The practical effect of this is that the forecasting system needs to be sufficiently flexible to accommodate secular changes such as the effects on revenue of a declining product line, or on expenditure for the development and launch of a new product line.

Commercial patterns

For example, the airline industry has a relatively predictable income cycle, based on fluctuations linked to holiday seasons, and relatively infrequently disrupted by truly unpredictable events such as volcanic eruptions. Fuel and staff cost patterns are also relatively predictable. However, airlines differ quite sharply in the ways in which they finance aircraft acquisitions, with the method used ranging from cash purchases to complex leasing arrangements; accordingly, different airlines will require specific cash forecasting solutions that properly reflect the demands of their aircraft acquisition method.

Treasury cash management policy

The specific methodology defined in an organisation’s treasury policy determines the way the treasury constructs its cash forecast. As a generality, European treasuries tend to build the forecast from the bottom up, basing it on commercial forecasts and committed receipts and payments, whereas American treasuries often derive their forecasts using statistical and mathematical models. So the forecast derivation (and perhaps results) may be very different, depending on the mandated methodology.

Cash Forecast Components

The base components of generalised cash forecast are:

  • Bank account balances.
  • Maturing and opening treasury transaction cash flows.
  • Forward value accounts payable (A/P) and accounts receivable (A/R) flows.

Treasuries using bottom-up forecast construction collect and apply commercial forecasts to the netted base components. The commercial forecasts are collected from around the organisation, typically via spreadsheets and e-mail, for consolidation with the central forecast spreadsheet system. The frequency of collection depends on the treasury cash management policy, and the complexity of the operation depends on the organisational and other factors described above.

Forecast Issues – a Problem Area?

Behind this simple account there lie the real-life complexities of collecting the data, and of integrating it into the forecast. And then there are the additional issues of the accuracy and dependability of the forecast. These issues pose a range of questions about the suitability of spreadsheet-based cash forecasts, such as:

  • How much central treasury effort is required to chase finance subsidiary personnel to provide current forecast data?
  • How reliable/accurate is the forecast information received?
  • How much central treasury effort is required to integrate the forecast information?
  • How much central treasury effort is required to validate the forecast data submitted, and to correct any errors?
  • How is the forecast measured against actual results?
  • Is the forecast sufficiently accurate to support optimised cash and working capital management?

There is a surprising number of large and medium-sized companies who still depend on spreadsheets and other manual methods to construct their cash forecasts. This observation is supported by the 2009 AFP Benchmarking Program result on the solution used by responding US treasuries for various treasury and finance activities. The AFP finding is that 82% of the sample use spreadsheets to develop cash flow forecast – with 7% using treasury workstations, 6% using in-house/other systems, and 4% ERP or similar packages.

We have indicated that some US treasuries use advanced statistical methods to construct and maintain cash forecasts. There are a number of alternative models that are encountered, and some more sophisticated operations may use more than one method, for benchmarking purposes.

Among the methods in use are:

  • Deriving the forecast based on a pre-defined growth pattern.
  • Apply moving averages.
  • Using linear regression, typically with the facility to eliminate extreme cases.

As these methods can involve a relatively heavy data processing and computational load, treasuries using them would seem to demand strong technology support for their cash forecasting operations.

How Automation can Help the Forecast Process

So cash forecasting is clearly an activity that would seem to cry out for an automated solution: it is mechanical, it is repetitious, it can involve the processing of relatively large volumes of data, it can require significant calculation power – and it is mission-critical, given the importance of cash visibility to today’s post financial crisis treasury department. So why are integrated, properly-automated solutions still encountered relatively rarely?

The AFP hints at an answer: after suggesting that lack of clear cost/benefit justification might be a factor, it also mentions the issue of a lack of the necessary financial or IT resources. The PricewaterhouseCoopers (PwC) Global Treasury Survey 2010 (‘Treasury in the Crisis; Put to the test – can the crisis make treasury stronger?’ by Sebastian di PaoIa and Damien McMahon, (PWC), identifies ‘faster access to data via integrated platforms, [which] improves accuracy and timeliness of cash flow forecasts’ as the first area of process improvement that could be achieved through properly-applied automation.

Our own view is that the vast majority of corporate treasurers would welcome the availability of integrated, automated cash forecast for the entire enterprise, to help their staff to remain focussed on their professional cash management duties, rather than on chasing, processing and correcting forecast information. The key requirement of integration using a treasury management system (TMS) is to automate the construction of the forecast. The solution will combine the necessary external and internal data, including bank account balances, treasury cash flows and A/P and A/R information. It is worth pointing out that the payables and receivables could be handled in detail or in summary, depending on volume and performance considerations.

The key issue relates to the collection of the commercial forecasts, perhaps from a disparate, global network of subsidiaries. Web-based solutions are current best practice in this area, offering a 24/7 user friendly mechanism for submitting forecasts to central treasury, quickly and easily. Forecast templates are often customised, using different local languages and business terminology: this adds to the solution’s user-friendliness, and can therefore contribute to the accuracy and timeliness of forecast entry.

One of the frequently reported issues in practical cash forecasting is the matter of encouraging finance subsidiaries to submit their forecast on an accurate and timely basis. If forecasts are submitted perhaps monthly, and then disappear as if into a black hole, the submitter has little or no interest in the process – and little or no reason to forecast accurately and punctually. At a minimum, all forecasters (and their finance managers) should receive regular forecast versus actual reporting, so that forecasting performance can be accurately monitored. If an automated in-house bank is in place, it is perfectly feasible to increase the debit interest rates applied to un-forecasted overdrafts or loan draw-downs, so that a tangible performance measurement is applied. However introducing a bonus for generated un-forecasted surpluses is probably not advisable!


We encourage treasurers to take a hard look at the value of implementing an integrated, company-wide cash forecasting solution under the general principles that we have outlined. One multinational client who went through this exercise recently surpassed expectations by achieving (for their European treasury) a 40% reduction in working capital requirement. Savings of this type may not be feasible for all companies – but with the probable medium-term financial future involving ongoing strains in liquidity and the availability of funding, investments to help optimise working capital clearly merit serious consideration.


Related reading