ECB riddle on inflation expectations
by Peter Warburton
Over the past two years, a seemingly innocuous difference in terminology has assumed great significance in the conduct of monetary policy. Whereas the US Federal Reserve is careful to describe forward inflation rates derived from the interest rate swap curve or from Treasury inflation-protected bonds as ‘inflation compensation’, the European Central Bank has opted for the stronger term ‘inflation expectations’.
This subtle distinction can be summarised as follows. ‘Inflation compensation’ refers to the pay-off from selling inflation protection or the cost of buying inflation protection. The accompanying narrative is bland and non-committal: the market is pricing an unknown outcome relating to future inflation expressed as a specific consumer price index.
‘Inflation expectations’, by contrast, has a deeper meaning. It infers the market has formed a collective view of the central expectation of inflation over a specific time horizon, taking account of all known information. This definition carries connotations of statistical efficiency. It is a small step from here to the elevation of the inflation swap or break-even rate, to the status of an independent barometer of central bank performance. The result is a riddle over the ECB's chosen benchmarks.
Under Mario Draghi’s leadership, the ECB appears to have fallen victim to this illusion. In numerous speeches and policy statements, the five-year, five-year inflation swap has been described as a market inflation expectation. When this swap rate plunged in 2014 as the latest instalment of the Greek crisis unfolded, the ECB invited the interpretation that the financial markets were losing confidence in its ability to achieve its declared objective of inflation ‘below, but close to 2%’. In other words, the weakening of this market indicator represented a call to action. By implication, to have ignored such a clear market signal would have brought risks for the ECB’s credibility. The central bank did not disappoint, launching its quantitative easing programme in January 2015.
The ECB’s strong interpretation of the five-year, five-year forward swap rate (see chart, below) ensured a repeat performance in early 2016. Indeed, this measure of future inflation dipped beneath its previous lows, and despite an initial rally after the eventful 10 March meeting has headed lower once again. The problem is that the policy actions open to the ECB – interest rate cuts, long-term repos and augmented QE – appear to have little traction with its policy objective. ECB council members questioned the wisdom of extending QE for this very reason.
Meanwhile, the ECB’s inflation forecasting has been woeful. In March and June last year, it projected a rebound to 1.5% inflation for 2016. In March, this forecast was reduced to 0.1% in recognition of oil price weakness. The ECB is in danger of appearing impotent in the face of these inflation setbacks.
A misplaced emphasis on the inflation swap rate as a representation of medium-term inflation expectations is all the more remarkable in view of the good work the ECB staff carried out as early as 2007. In ‘Working Paper 734’, the authors concluded that ‘the break-even inflation rate is a noisy measure of expected inflation, because it can include an inflation risk premium component.’ Moreover, ‘our results suggest that fluctuations in the raw break-even rate have mostly reflected variations in the inflation risk premium, while long-term inflation expectations have remained remarkably anchored since 1999'.
In 2015, Bank of England researchers comprehensively evaluated the informational content of market-based measures of inflation expectations. The authors found that liquidity premia in gilt break-even inflation rates explained a large part of the total risk premium during certain periods, especially in the post-2008 crisis period. ‘The results suggest that the negative sign of the risk premium in break-even inflation rates during these periods was, to a large extent, the result of negative liquidity premia, which we conclude were driven by bouts of illiquidity in the market for index-linked gilts.’
Many commentators have highlighted the extraordinarily strong direct correlation between inflation break-evens, or inflation swap rates, and the crude oil price during the past two years. It is important to note that this correlation is relatively recent and used to be entirely absent. This is a correlation that has its roots in financial risk management, not economics.
Research by Economic Perspectives confirms that oil prices from five to 10 years ago have no explanatory power for today’s inflation rate. The monetary and real forces that will determine CPI inflation rates in five years have themselves not yet been settled.
The ECB has allowed itself to be chastised by a noisy market indicator with threadbare credentials as a predictor of inflation. Either the ECB must change its inflation language or it will need to change its inflation target to rescue a semblance of credibility. People have expectations. Markets have only prices.
Peter Warburton is Chief Economist at Economic Perspectives. Back