A Real-time Density Forecast Evaluation of the ECB Survey of Professional Forecasters

We evaluate the real-time predictive ability of density forecasts from the European Central Bank’s Survey of Professional Forecasters (ECB SPF) using the Diebold and Mariano (1995) and West (1996) test. As the sample size for the ECB SPF is fairly small, we use fixed-b and fixed-m asymptotics to alleviate size distortions. We verify in an original Monte Carlo design that fixed-smoothing asymptotics delivers correctly sized tests in this framework. Empirical results indicate that ECB SPF density forecasts for unemployment and real GDP growth beat simple benchmarks at one-year horizon. ECB SPF density forecasts for inflation instead do not easily outperform simple benchmarks, as up to 2008 ECB SPF inflation expectations are close to the target. After 2008, we find that the predictive ability of the ECB SPF is more conspicuous for all variables, even though inflation expectations are still loosely anchored to the target.