If the whole organisation understands the importance of cash, it becomes easier to harvest more diverse, accurate and timely data. From here, better forecasting can drive stronger working capital performance. It’s what most stakeholders expect, so how can it be delivered?
Cash is King, maybe; it’s hard to tell as the power that this particular monarch carries has not always been appreciated. However, since the global economy tanked and then started on its long road to recovery, the value of understanding cash basics – such as how much, where and in which currencies – has been rising higher up the agenda.
An accurate picture of current and future cash needs is only possible if every function that affects the ebb and flow of cash understands its importance. Indeed, every touchpoint needs to be able to report its cash position to treasury, with equal accuracy and timeliness, if an enterprise-wide viewpoint is to be achieved.
However, rallying the troops has not always been easy. The list of key protagonists required to achieve this in an industrial context, for example, includes departments as diverse as sales, procurement, finance, production and logistics; each will harbour their own private drivers and agendas.
Regardless of how many functions commune with the King, the importance of building a sustainable cash culture cannot be overstated. It is the driver of accurate forecasting, which is used primarily as a tool to help improve accuracy in investment and borrowing decision-making by identifying in advance potential surplus cash or gaps in cash flow.
Producing today’s cash position is relatively easy compared to accurate forecasting weeks or months ahead. The further out treasury goes, the harder it becomes to offer anything beyond an educated guess. Nonetheless, the ability to deliver accurate cash predictions can be the difference between running out of money and not running out of money, says Tim Bartlett, Senior Liquidity Commercialisation Manager, HSBC. Indeed, he comments, no matter how much a company is valued on paper, without accurate cash flow forecasting, running out of day-to-day cash is a potentially fatal risk.
There is no one correct method of creating a forecast; it depends upon factors such as the nature of the business or sector, the type of forecast sought, its purpose and format. This may sound somewhat unscientific, but with so many unknowns with which to contend, not least macro-economic ‘events’, forecasting is very difficult to optimise, notes Bartlett.
Of course, there will be some regular and predictable incomings and outgoings which make forecasting a little easier: predictability is the friend of the forecaster. Nonetheless, delivering detailed forecast data with a high degree of accuracy remains a challenge.
Actuals are used to validate forecasts, giving the opportunity to adjust and improve the margins of error in light of any variances identified (and any targets set through KPIs). These validations may be made daily, weekly, monthly or on a longer timescale, according to appetite and ability to monitor, extract, analyse and respond to the data. The more frequently a forecast is updated with real numbers, the more accurate it will be but there comes a point where the effort outstrips the return.
There are many different forecasting methodologies – the distribution method, regression analysis, and time series trends and seasonal variances, for example – most relying on analysis of historic cash flow data. Because there is no guarantee that patterns identified using historical data will reoccur, these techniques will generate a forecast probability which is commonly weighting, in the form of an accepted margin of error. As the actuals are made available, the degree of ‘wrongness’ can be fine-tuned with each set of actuals.
The use of technology can take some of the hard work out of this process. Off-the-shelf solutions, from third-party providers such as Kyriba, CashAnalytics and FiREapps use analytical algorithms written with the benefit of the wider industry experience of each vendor. These can offer complex mathematical responses to common scenarios which will be tuneable to more closely represent a company’s individual circumstances and strategic approach.
Indeed, such systems can improve forecasting accuracy, for example, by incorporating a computer-based understanding of the behaviours of clients in terms of incoming revenues, and the expectations of suppliers in terms of outgoings.
Of course, many businesses use spreadsheet calculations, adding their own margins of error and iterating each set of results to reach the next waypoint. With output from bank reporting tools having become more sophisticated in recent times there is no suggestion that companies cannot produce sufficiently accurate forecasts using their own tools.
However, the nature of forecasting is such that there is always room to increase accuracy, driving stronger working capital performance.
The cutting edge
One area in which advancement is being made is in the adoption of pre-cognitive technology that borders on the realms of artificial intelligence. This is widely used in fraud detection, enabling the recognition of complex patterns of flow and the predictability of certain activities. This, argues, Bartlett, is the kind of solution from which forecasting can and will benefit.
The encouragement of open banking (largely through regulatory measures such as PSD2 in Europe), where institutions share data flows through API-led connectivity could also bring a more easily aggregated view of flows across a multi-banking environment.
Regardless of technological advances applied to any aspect of treasury or finance, Bartlett comments that “these are only ever tools, and tools by definition are something people use to help get the job done”. Of course, the treasurer must know how to use these tools but Bartlett notes a fine line between using them simply as a means of looking for or proving a preconceived notion, and accepting what these tools deliver as the absolute truth.
Is there appetite for change? Achieving 100% accuracy every time is an unrealistic goal. However, although some businesses do let cash flow drift to the point of becoming technically insolvent, many do not. This suggests that today’s forecasting measures are, by and large, adequate.
But with many treasuries having become P&L centres in their own right in recent years, Bartlett argues that there may be expectations of increased return overall for the company, itself demanding higher expectations of forecasting precision.