Inflation Expectations and Monetary Policy
Last month, we discussed concerns about the rising yield curve, which some people believe may be signaling an increase in longer-term inflation expectations. As we explained, the sort of increase we have seen in the yield curve would not usually garner much attention, but Fed actions to address the financial crisis had already aroused some worries about future inflation. Over the month of June, the yield curve has continued to rise. A steep yield curve implies that interest rates are expected to increase. Increases in future rates are thought to be governed largely by future increases in the real interest rate or future increases in inflation.
One way to gauge whether it is inflation or interest rates that is driving the recent increase in the yield curve is to look at information contained in inflation-adjusted treasury securities (TIPS). Since TIPS are indexed to inflation, their yields are a measure of the real (inflation-adjusted) interest rate that is expected to prevail over the maturity of the bond. When we look at TIPS-estimated real interest rates, we do not see the noticeable increase that we do in the estimates derived from nominal bonds.
Since TIPS yields give a measure of real interest rates, they are frequently used to back out a "breakeven" inflation rate, which is used as a measure of expected inflation. When we calculate this measure, we see why many are concerned with the steepening of the yield curve. TIPS-estimated expected inflation has crept up for all maturities.
However, while breakeven inflation rates have certainly increased, they are still a little below where they were a year ago. Therefore, understanding the dramatic decrease in breakeven inflation rates from June 2008 to the end of 2008 will probably shed light on their subsequent increase. Did expected inflation really decline as dramatically from mid 2008 to the end of the year as the breakeven inflation rates suggest? We argue almost certainly not.
In times of serious liquidity concerns, interpreting the TIPS breakeven inflation measure is problematic. TIPS securities are less liquid than regular nominal securities, a fact that lowers their price and increases their yield. Increases in liquidity pressures will therefore decrease TIPS breakeven inflation rates even if actual expected inflation is constant. Thus, during periods of liquidity stress, breakeven inflation rates derived from TIPS will understate actual expected inflation. This suggests that the decline we saw in 2008 is probably largely due to the intense liquidity pressures prevailing at the time.
We attempt to quantify the extent to which liquidity pressures have biased observed breakeven inflation rates. To do this we need to correct for the liquidity bias in TIPS. This correction requires us to make some assumptions. First we assume that very long-term inflation expectations (as measured by the 10- to 20-year breakeven inflation rate) would have been constant in the absence of liquidity pressures. The implied forward breakeven inflation rates are calculated by assuming the expectations hypothesis, which maintains that holding long-term real and nominal interest bonds is expected to be the same as holding a series of short-term bonds. Second, we assume that liquidity pressures affected all TIPS equally. That is, 5-, 10-, and 20-year TIPS all had an equal liquidity premium embedded in them. We use this measure of the liquidity premium to adjust breakeven inflation rates of all maturities.
Our calculations show that liquidity concerns at the height of the crisis subtracted over 1.5 percentage points from measured breakeven inflation rates. That is, breakeven inflation rates severely underestimated the deflationary pressures at the time. Since then this liquidity adjustment has come down. It especially appears to have declined after the results of the stress tests given to banks were made public.
Without correcting for liquidity, 5-year breakeven inflation rates at the height of the crisis suggested that prices would decline at an average rate of almost 1.5 percent per year over the next five years. After correcting for the liquidity premium, it looked like prices were expected to be nearly flat over the next five years.
Liquidity-adjusted longer-term breakeven inflation rates show a similar but significantly muted pattern as short-term 5-year breakeven inflation expectations. Using these two estimates, we can back out medium to long-term inflation expectations as measured by 5- to 10-year breakeven inflation rates. These medium-range estimates suggest that once liquidity is corrected for, inflation expectations have been largely unaffected by the crisis. Five- to 10-year inflation expectations are useful because they abstract from the high-frequency inflation declines that the market might have been expecting over the next five years. Currently, inflation expectations from 5- to10-years out are basically identical to what they were before the crisis.
While short-term inflation expectations have crept up recently, they have just retraced the declines they experienced during the height of the crisis. Longer-term inflation expectations as measured by liquidity-adjusted 5- to 10-year breakeven inflation rates suggest that inflation expectations did not decline during the crisis and have not crept up significantly since then. They are now right around where they stood before the crisis began. But the increases we have seen in the yield curve, coupled with a relatively flat yield curve for real TIPS, warrant an ever-watchful eye to make sure that inflation expectations do not creep up.