Many functions of enterprise rely on cash flow forecasting. It is the life-blood of businesses, big or small. For managers, cash flow forecasting enables them to confidently fund projects or obtain new financing. In foreign exchange roles, it is important to know the volumes of incoming and outgoing currencies in order to accurately hedge foreign exchange risks. For executives, it is essential to have a clear view of cash projections to be able to act with confidence when they need to invest, acquire a company or take part in large financial operations.
How Big Data Analytics is Transforming Cash Flow Forecasting
Most finance and treasury teams rely on rudimentary tools such as spreadsheets to derive a cash flow prediction. Unfortunately, spreadsheets have volume and algorithm limitations which means cash flow forecasting derived from these tools do not provide an accurate depiction of the whole business. Why is accurate cash forecasting important, and what is “accurate” exactly?
Traditional cash flow forecasting can be likened to human weather forecasting. We’re all guilty of human forecasting; when you step outside and see a dark, cloudy sky you can make the assumption that it will rain that day. My mom, for example, used to make weather predictions based on wind flow, temperature or sun breaks. More than half the time however, her weather predictions were inaccurate or completely off. What she failed to do, despite her best intentions, was to include underlying variables required to accurately predict weather. Modern weather forecasts are computer generated; processing power enables systems to include hundreds of variables such as atmospheric pressures, humidity, proximity of depression, anticyclone depressions, etc. My mom’s weather prediction was also missing 2 centuries’ worth of weather data which includes recorded patterns, their correlations and impact. No human being has the cognitive prowess to combine historic data and current variables to deliver an accurate prediction. It’s not to say that one cannot make an assumption based on current weather conditions. It’s good, but not good enough.
Modern cash forecasting should be approached in the same manner. In order to create an accurate cash flow forecast, finance teams need to need to understand the exact nature of invoices in portfolio, not only their origin or amount, or the conditions they were booked under but also all the elements that could have an influence on the actual payments such as quality of delivery, type of goods sent, logistics, rebates and overall quality of service. These elements are important to consider because they have an impact on the way payments are made. Invoices influence both our behavior and our customer’s behavior.
It’s important to note that understanding the current situation is not enough. It is necessary to establish the patterns of all payments made and received from the years prior. Data patterns can show and predict how certain types of transactions will be paid. To be more specific, cross identifying patterns with an existing portfolio of invoices can predict when they are more likely to be paid. The theory sounds simple enough but the issue remains: current desktop or transactional treasury software cannot scale. Such tools do not have the statistical analytics capability to convey pattern recognition and projections, let alone the processing power to deliver such complex analyses in a timely manner.
Today, in-memory engines, coupled with robust statistical capabilities, can run pattern identification from years of historic transactional data and leverage every single aspect of every single invoice. When applied to current set of invoices, these finely defined guidelines can yield more reliable results compared to, say, a forecast derived from a sample or from gut-feeling.
My mom cannot compete with the machine power of the national weather service, and finance teams’ spreadsheets should not either. Personally, I tend to take my mom’s weather forecasts with a grain of salt. Ultimately, I use my phone’s weather app to see if it’s worth going outside for the day. What will you rely on for something as important as cash prediction?