5 Reasons why financial forecasts fail - and what to do about it in 2023

For the fourth year in a row, the reliability of cash-flow forecasts comes ahead as treasurers’ No. 1 priority in EACT’s Treasury Survey 2023. Given the upheavals in the world economy in recent years, this result is no surprise. Yet, the sobering fact is that most financial forecasts fail. Cash flow forecasts, currency forecasts, economic and political forecasts — all of them are vulnerable to a common set of problems. Below we examine these vulnerabilities. We also assess some of the tools deployed by treasurers to improve the reliability of their financial forecasts. Finally, we take a slightly contrarian view from the perspective of FX risk management: when it comes to currency hedging, the degree of forecasting accuracy is less important than what is generally assumed. So, let us start by asking ourselves: what is wrong with financial forecasts these days?

1 – It is impossible to predict the future

Whether financial forecasts are formulated by economists or gurus such as ex-finance ministers, central banks officials, financial journalists or university professors, it does not matter. Nobody knows what the future will bring. Forecasts tend to be mere projections of the current situation, extrapolated in time.Most financial forecasts assume that the future takes the shape of a linear variation of the present. That methodology is good enough in stable times, but it proves entirely useless when discontinuity sets in. Such methods cannot predict “black swans” — unexpected, non-linear events that occur with surprising frequency.Earthquakes are impossible to predict. We know that they are going to happen, but not when or where they will strike. The best strategy is to be prepared, and that involves implementing efficient emergency procedures enabling managers to minimise their impact.

2 – Forecasters suffer from behavioral biases

CFOs and treasurers need to pay attention to the all-too-human tendency to behave in an overconfident manner. When managing a firm’s currency exposure, being unrealistically sure of ourselves can prove a risky proposition.This overconfidence bias manifests itself mostly in currency forecasts. Such financial forecasts may well turn out to be accurate during a year or two. Then disaster inevitably strikes — and one single miss can take an entire company down. The best antidote against overconfidence in currency management is to deploy automated hedging programs that allow CFOs to systematically achieve their goals, whatever the outcome in FX markets (see: “Conquering behavioural biases in currency management”).

3 – Financial forecasts are based on incomplete research

Too often experts and forecasters are affected by the event they are trying to analyse, making it more difficult for them to draw objective conclusions. When it comes to financial institutions analysing economic scenarios, forecasters tend to have a personal interest regarding the outcome.A good example of this is what Nobel laureate Daniel Kahneman calls the “inside view.” Managers tend to build a detailed case for what is going to happen based on the specifics of the case at hand, rather than looking at analogous cases and other external sources of information.The critical step is to identify the right reference alternatives, a process that is part art and part science. There are usually more of them than executives care to admit. According to a recent study, taking this “outside view” —that is, statistically assessing projected outcomes based on a reference class of similar projects— can reduce forecast errors by 70%

4 – Financial forecasts misses are soon forgotten

Meteorologists say that weather forecast misses are quickly forgotten unless they fail to predict a massive disaster like a tsunami or a hurricane. The same rule applies to economic and currency forecasts.Many experts publish their predictions, secure in the knowledge that miscalculations will soon be left behind except in those cases in which they got it completely wrong on a major issue like Brexit in 2016. Most of the time, failures are forgotten, while the few accurate predictions can be proudly recalled to highlight their expert’s reputation.However, let’s be honest: any analyst that would be able to anticipate half of the economic developments would be living on the wealth produced by his/her investment decisions rather than making quarterly financial forecasts.

5 – Many experts forecast one thing and the contrary

If missed financial forecasts are readily forgotten, one can indulge in saying one thing and then the contrary. The prediction will turn out to be accurate in one case or the other. Some days before the 2016 U.S. elections, currency forecasts put forward by a Japanese investment bank suggested that the dollar index would drop 1.3%, while the EUR-JPY exchange rate would sink 4.7% in case of a Donald Trump victory.The yen, euro and the Swiss Franc would subsequently appreciate, with investors looking for safe havens. Yet, three days after Mr. Trump's electoral win, the investment bank happily changed its views, arguing that the dollar would rally against the euro and the yen.

What should you do, then?

Despite the relentless pressure they face in terms of forecasting accuracy, managers must recognise the inherent limitations of all financial forecasts. A good practice is to anticipate different scenarios, analyse their impact in terms of business performance and establish procedures to react in each case. But there are some positive developments as well. Managers can take solace from developments in budgeting techniques and in business process automation in the following areas:

• Better tools to assess scenarios. While they will never solve the eternal problem of forecasting accuracy, developments in Artificial Intelligence (AI) in pricing, forecasting and simulation are making it possible for business managers to quickly assess an array of potential scenarios with more precision than ever before.
• Emphasis on rolling forecasts. As pandemics, inflation and war become the norm, there is a discernible move towards rolling forecasts, where estimations are continuously adapted to recent trends. Such a move is also bound to have implications in terms of currency hedging (see: “The digitalisation of treasury operations”).
• Automated FX hedging programs. Finally, the availability of automated currency hedging programs is good news for beleaguered currency risk managers relying on forecasts. Whether they address cash flow or balance sheet hedging, these programs are configured in a way that lessens the need for super-accurate forecasts (see: “The myth of forecasting accuracy).