An Analysis of Expectation Formation in the Euro Area using SPF

As a result of current pressures on inflation in the Eurozone, policy makers and scientists increasingly follow the impact of expectations on the inflation formation process. The purpose of this paper is twofold: (i) using the European Central Bank's (ECB’s) Survey of Professional Forecasters (SPF) dataset for the period 1999: Q1 to 2014: Q3, we test for the market participants’ consistency of estimations. (ii) Additionally we test, whether volatility in short-run expectations passes through to long term inflation expectations.

Regarding the first question we control for a potential forecast bias and analyze the rationality of forecasters. Additionally the paper investigates whether the predictions made are calculated following classical theory. Thereby we compare inflation forecasts by using a typical chartist-fundamentalist framework. The model proves to be capable of explaining a significant share of the volatility of forecasts. We study forecasters’ tendency to use short- and long-run developments of inflation rates as a basis for their estimates following a New Keynesian Philips Curve and compare output gaps as a measure of aggregate demand. Thus we estimate forecasters’ performance using an extrapolative and an adaptive framework and compare it to the realized inflation creation process. We find that forecasters hold contrarian expectations, yet significant changes can be found in pre- and post-Global Financial Crisis periods. The short run inflation forecasts in particular seem to be subject to significant swings in forecast changes. For an increasing horizon of two or five years, we find forecasts to stabilize over time. Additionally, the data shows that forecasters hold contrarian expectations regarding long run swings in the inflation rate. Moreover analysts show a significant heterogeneity in the period before and after the crisis. One reason for this seems be that market experts seem to underestimate the degree of negative autocorrelation.

In a further step, we investigate the pass-through effect between short-run and long-run expectations. The paper tries to identify whether current inflation short run rate swings influence investors’ beliefs in the anchoring of the Eurozone. To see the potential influence of one variable on the other we use rolling regressions over the whole sample period to provide an overview of potential inconsistencies and changes over time. We find no major pass-through mechanism since the start of the euro area. Recently however, we see a change in forecasting behavior and forecasters do tend to adopt short-run information in their long-run forecasts.