The anchoring of inflation expectations has become one of the most prominent issues in monetary policy of late. While there is a clear consensus among both academics and the central banker community that this point is very important, there is a central question left unanswered: What is the appropriate measure for inflation expectations. No other person than Federal Reserve Chairman Bernanke in a recent speech summed up this point as follows: “Many measures of expected inflation exist, including expectations taken from surveys of households, forecasts by professional economists, and information extracted from markets for inflation-indexed securities. Unfortunately, only very limited information is available on expectations of price-setters themselves, namely businesses. Which of these agents’ expectations are most important for inflation dynamics, and how can central bankers best extract the relevant information from the various available measures?” In this piece we are going to address three questions? 1. Are inflation expectations – as measured by the monthly EU survey – a reliable guide to future inflation? 2. What is the interaction of the unemployment rate and inflation expectations? 3. How serious is the threat of second-round effects, i.e. the significant gap between current headline and core inflation filtering through into permanently higher inflation rates? First of all it is worth noting that for EMU the scope of measures for inflation expectations is rather more limited than for the US. For example, for consumers there is only the measure for the 12 months forward price expectations from the EU survey. And this measure is not given as a numerical expectation but in the form result of a diffusion index, i.e. the difference between the share of people expecting higher and the share of people expecting lower inflation. It is noteworthy, though, that the diffusion index for the price development correlates quite well with actual inflation so it seems possible to translate the diffusion index into a numerical equivalent for inflation rates. In order to address the first of our questions we did a correlation analysis of the future inflation expectation component of the survey with the core (ex energy, food, alcohol and tobacco) rate and the headline rate. To be precise we time-shifted the future component relative to inflation rates and indeed find relatively high rates of correlation for a 12 to 15 months lead of the survey to actual inflation rates with respect to both core and headline (see chart). So consumers indeed to seem to be having a reasonably good judgment in which direction inflation will be heading next. If we were to assume that the correlation holds over the next twelve months, then core inflation would show an upward movement from its current 1.8% year-over-year to 2.25%
However, taking this analysis into a presumption of further necessary monetary tightening for the ECB would be quite an intellectual stretch. First, a single year is hardly something a central bank would assume to be “medium-term” and no other longer-dated consumer inflation expectation measure is available for EMU. Second, we should take a look into how inflation expectations are formed or activity measures such as unemployment influence them. The following chart plots the EU Survey result for the 12 months future expectations vis-à-vis the percentage point change in the unemployment rate (with a lead of nine months on the survey data). There is a distinct negative correlation between the two time series, i.e. rising unemployment tends to dampen inflation expectations. It would almost seem as if European consumers have internalised the NAIRU concept. As can be seen from the chart, there are two striking features: a) inflation expectations are still somewhat below their 2001 levels and b) the unemployment rate might for the time being still be a bit lower than a year ago but the pace of reduction has almost come to a stand-still. Over a two year horizon, an unchanged unemployment rate should – given historical correlations – even suffice to bring the core inflation rate bang on 2 %. So right on target for the ECB if we assume that no significant rises from food and energy were to occur.
Interim conclusion: Even if inflation expectations might be “elevated” and herald somewhat rising core inflation over the coming twelve months, this is nothing that the slowdown already underway would not efficiently deal with. This brings us to the third question and the one that might be weighting most heavily on the ECB right now. The following chart plots the gap between the headline and the core rate with a lead-time of 18 months against the present core rate. We find a reasonably good and positive correlation between both variables. This can interpreted in a fashion that over time price setters make up for any loss in purchasing power induced by food and oil by raising prices on core goods. An initial, yet narrow, shock is thus likely to be transformed over time into a broad based rise in inflationary pressures. For the current situation, the sharp rise in the headline/core gap could keep core rates high for the entire year 2009.
It should be noted, however, that this correlation cuts both ways. The temporary decline in oil prices in H2 2006, for example, is most likely to act as a dampener on core rates in the middle and late part of 2008. So the ECB would not only have to reckon with continued high oil prices but ever rising oil prices in order to accurately identify a permanent threat to price stability via second round effects. Bottom line: Even if correlations suggest that 2009 inflation might remain at elevated levels due to the recent spike in oil prices and a lagged effect from a booming labour market in 2006 and 2007, neither factor will be sufficient to offset the inflationary cooling from the economic slowdown. Central to the ECB policy stance will be the interpretation “medium-term” in its mandate for price stability. Only if medium-term were to be interpreted as 2009 exclusively, additional rate hikes – ineffective as they are over that time-horizon – may be on the cards. If it is to include 2010, then in all likelihood the ECB can rest easy or even move towards easing.