Bank currency forecasts miss by more than 5% on average, Smart Currency Business analysis shows

Currency forecasts provided by bank economists are often wrong, Smart Currency Business data shows.

Analysis by Smart Currency Business has found that currency predictions made by leading banks going back to 2014 have missed by more than 5% over a 3-12 month period.

The research reviewed 1,547 individual bank and economist currency forecasts across GBP/USD, GBP/EUR and EUR/USD, comparing predictions made between three and twelve months ahead with what actually happened in the market.

The analysis found that, on average, forecasts missed the realised exchange rate by 5.3%. For 12-month forecasts, the average miss rose to 7.3%.

In practice, this discrepency can represent tens of thousands of pounds, euros or dollars on a single business transaction. At a GBP/EUR rate of 1.20, for example, a 4% forecast miss is equivalent to a difference of almost five cents. On a £1m euro purchase, that equates to around €48,000 of unexpected movement.

Smart Currency Business said the findings highlight the danger of treating forecasts as reliable budget anchors, particularly for SMEs with narrow margins or repeated foreign currency exposure.

The study found that:

Alex Bennett, Managing Director of Smart Currency Business, said:

“This stark data shows that currency forecasts are more than just occasionally wrong. For a business moving millions of pounds, not knowing where the exchange rate might end up can mean tens of thousands in unplanned cost if the risk is left unmanaged.”

The findings also showed that forecast errors increased with time. The average 3-month forecast miss was 3.7%, rising to 5.2% at six months and 7.3% at twelve months.

This deterioration is important because many companies set annual budgets based on 6-12 month planning assumptions.

The research also found that consensus forecasts were not immune. Even the median forecast across banks missed by 4.4% on average for 3-12 month horizons, and by 5.9% at the 12-month horizon.

Alex Bennett added:

“Many businesses assume that taking the average or median of bank forecasts is a safe way to build a budget. The problem is that the consensus forecast was still wrong by 4.4% on average. That is easily enough to affect margin, cashflow and pricing decisions.”

Perhaps the most striking finding was that the entire range of expert predictions failed to include the realised exchange rate in 36% of cases.

In other words, even when taking the lowest and highest bank forecasts as a range of possible outcomes, the actual market rate still fell outside that range more than one-third of the time.

At twelve months, the forecast range was more likely to capture reality, but only because the range became extremely wide. The average twelve-month expert range was equivalent to 16.8% of the realised exchange rate.

Alex Bennett said this creates a practical problem for finance teams.

“A very wide forecast range may technically be more likely to include the final rate, but it becomes far less useful for budgeting. If a forecast range is so wide that it creates six-figure uncertainty on a £1m exposure, it is not giving a finance director a reliable number to plan around.”

The analysis also suggested that forecast failure is not confined to obvious crisis periods. While Covid and Brexit-related volatility created major challenges, forecast misses were also significant in apparently more normal years.

The worst years in the sample included 2015, 2017, 2018, 2019 and 2020, showing that currency forecasting errors are not simply the result of rare “black swan” events.

Smart Currency Business said the research reinforces the need for companies to move away from chasing the best rate and towards structured currency risk management.