GDP Forecasts Often Fail Basic Rationality Checks

A new study from the Reserve Bank of Australia (RBA) has found that major global GDP forecasts often fail basic 'forecast rationality' checks, suggesting they may be overly optimistic or biased.

The RBA team compared real GDP growth forecasts from the International Monetary Fund (IMF), World Bank, European Commission, and Organisation for Economic Co-operation and Development (OECD) with other projections. Across most economies, each forecaster showed patterns that standard tests flag as 'irrational', including a mild optimistic tilt, revisions larger than expected, and calls that were too extreme.

However, when researchers lined the groups up country by country, the accuracy gaps were usually not statistically significant, meaning any apparent 'best forecaster' could be due to luck rather than skill. The RBA paper argues part of the mismatch is structural: these organizations often publish forecasts conditional on assumed paths for other inputs, such as commodity prices or fiscal policy.

The study's findings have implications for markets, which often translate GDP forecasts into expected growth paths for rates, currencies, and earnings. The RBA suggests that instead of reacting to a single organization's upgrade or downgrade, investors should be wary of overreacting and wait for multiple organizations to revise in the same direction around the same time.