The bank for a changing world

7 days of Economics - ECB: predictable projections create dovish bias

  • William De Vijlder Group Chief Economist BNP Paribas
  • Paris
  • 13.03.2018
  1. Core inflation projections by the ECB have followed a surprisingly predictable path
  2. Starting at close to but below 2% for long horizons, they decline as the remaining horizon shortens
  3.   The big difference between the projections at long horizons and the eventual outcome implies they provide little information to the market: they simple confirm the ECB’s objective. They can also introduce a dovish policy bias 

The ECB/Eurosystem staff economic projections may not yet have reached the status of the ‘dots’ of the FOMC members, but market participants can nevertheless use them to gauge what the ECB thinks about its ability to steer inflation towards its objective. Since March 2014, the projection tables also cover HICP excluding energy, food as well as the effect of changes in indirect taxes, in short core inflation. 

Looking at all the projections since 2014 for different horizons provides interesting insights. As shown in the chart (see Ecoweek), when the projection window is long (about 3 years), the core inflation projection tends to be high (1,7-1,8%) and stable, whatever the economic environment when the projection is made. This suggests that the ECB is of the view that in the longer run its policy will be successful in bringing inflation close to target: the inflation projections are anchored. 

Subsequent projections provide updates of the earlier ones and as the projection window shortens (in the chart we move from right to left), the inflation projection declines: the role of the ECB objective as an anchor for its inflation projections wanes whereas current core inflation, which has been stubbornly low, increasingly acts as a magnet. It becomes a pullback force for the projections when the horizon gets shorter. This pattern has been remarkably stable since 2014, simply because inflation has been stuck at a low level, despite an aggressive monetary policy and, for several quarters in a row, GDP growth above potential. It does imply that the initial projections have considerably overestimated the eventual outcome. This shouldn’t come as a surprise considering that the initial projection covers a long horizon (three years). Calling it a projection, which serves an indicative purpose, rather than a forecast makes a lot of sense. It also means that the long horizon projections provide little information to the market, except simply confirming the ECB’s objective and its confidence that its policy will be effective. On this point, the systematic nature of the projection errors, even for shorter horizons, provides food for thought: it may mean that we’re stuck in a low inflation regime (“the Phillips curve is dead”) or that monetary policy effectiveness is overestimated. Both factors may introduce a dovish bias in the reaction function of the ECB.

Read more

Give us your feedback Complete our survey