Artificial intelligence can help predict cash, but the most significant development will be the AI’s ability to make efficient and informed decisions.
Today, the amount of data available has become too much for the human brain to be able to analyze. Moreover, the need to perform these analyzes quickly has never been greater. Artificial Intelligence (AI) may assist in cash prediction, but the most significant development will be the ability of AI to support efficient and informed decision making.
In the world of treasures, AI is used in two completely different ways: to improve cash predictions for machine learning and to optimize operations on the resulting liquidity. The first point deserves special attention and that is exactly what happened. However, in our opinion, it is the second point that treasurers will benefit the most from AI.
The forecast horizon can be short (less than a month) or long (usually one to three years ahead). For treasurers, it is essential to have a much more controlled management of their forecasts in order to optimize their financial resources.
However, currently long-term forecasting is only possible for a small number of companies. According to a survey conducted by consulting firm IDC, less than 5% of companies can reliably predict their cash flow even after three months and less than 20% are able to predict their liquidity beyond one month (1). This problem arises especially when CFOs are under pressure to ensure access to their company’s liquidity, as they reach the limits of manual cash forecasting.
In addition, these limited predictions are often wrong. According to the Association of Chartered Certified Accountants, 90% of Excel spreadsheets contain errors, but more than 90% of users are convinced that their spreadsheets are not! It is difficult (if not impossible) for his heirs to regain control.
Moreover, the world economy has entered a time of inflation. The all-item index rose 7% in the 12 months ending December 2021, the largest 12-month increase since the end of June 1982, according to U.S. Labor Statistics. United States (3). This situation forces central banks to limit access to liquidity, which would make this resource more strategic for companies. In this context, the ability of CFOs to make the best use of their data to optimize their liquidity becomes a major competitive advantage.
The only way to achieve proper performance is to rely on AI to tap the indexable volume of data accessible to enterprises. By 2025, IDC predicts that the global datasphere will reach 163 jetabytes (or one trillion gigabytes). This is ten times more than the 16.1 Zo data generated in 2016 (4). To understand how AI can contribute to cash forecasts, it is interesting to compare them with weather forecasts. Meteorology has evolved from the ability to determine whether it will rain (or not) on a particular day, whether it will rain (or not) at a certain time of day. Similarly, thanks to AI, treasurers are able to statistically predict their group’s liquidity at a given time based on the probability of different cash flows in the company. Using data from the company’s TMS and ERP systems, AI can analyze historical cash flows, train algorithms, and measure the level of statistical confidence in forecast results.
So far so good. But forecasting is not everything. Once confidence in the cash flow forecast is established, the treasurer can fix the company’s liquidity and decide how to invest the extra cash, whether in traditional money market funds or alternative products such as dynamic discounting (etc.). He can also choose the best lines of credit to optimize cost financing and payment campaigns. Optimal Treasury decisions can significantly improve a company’s earnings statement, reduce financial risk, and strengthen debt and short-term investment structures.
In terms of benefits, our internal studies show that with the help of a liquidity optimization tool, a CFO can save up to 50 basis points in financial costs without compromising their company’s liquidity access. These gains result in reduced costs of cash deposit opportunities, higher returns on financial investments and reduced charges on credit facilities. Obviously, the actual profit will vary from client to client.
But there are more general benefits. By using AI to solve problems such as liquidity forecasting, Treasury teams can spend a lot more time on other activities. Since AI supports many everyday tasks, Treasury professionals can focus on higher value and significantly more rewarding tasks!
(1) Kyriba Enterprise Liquidity Management Survey, IDC, July 2021