Imagine Fed Governor Rip van Winkle started his nap at the beginning of 2007 and just woke up to find that inflation is close to the Fed’s objective and the unemployment rate is at its 30-year average. You could forgive him for expecting the federal funds rate to be close to its long-run norm of about 4%, and for his surprise upon learning that the funds rate is at 0.1% and Fed assets are five times where they were when his snooze began.
Is the Fed already behind the curve? Why do policymakers emphasize their expectation that rates will stay low “for a considerable time” beyond October (when asset purchases are expected to halt)? What risks are they seeking to balance?
The most common benchmark for monetary policy is the Taylor rule, which relates the central bank’s policy rate to a combination of deviations of inflation from its target and a measure of resource slack. The modified Taylor rule in the chart below shows that – even ignoring the Fed’s balance sheet expansion – the Fed’s interest rate policy is now unusually stimulative by the standard of the past three decades. [The blue line in the chart is based on the Fed’s preferred inflation measure, the price index of personal consumption expenditures, and the deviation of the unemployment rate from its equilibrium level as a measure of slack.]
U.S. Federal Funds Rate vs. Modified Taylor Rule, 1985–May 2014
So, what might warrant such large stimulus? We can think of four possible reasons: (1) the Fed’s inflation objective isn’t really 2%, it’s higher; (2) the equilibrium real interest rate is well below the 1.75% implied by FOMC members’ current long-run projections; (3) resource slack far exceeds that implied by the 6.1% unemployment rate; and (4) the Fed is purchasing insurance against a negative economic shock that would – once again – force it to rely on unconventional balance sheet policies. Let’s analyze these one-by-one.
Starting with inflation, we can explain the current level of the policy rate if we assume the Fed’s inflation objective is above 2%. Mechanically, using current data and the Taylor rule, the implied level of the inflation target is in the range of 6%. While we might think that policymakers are willing to tolerate inflation of 3% for a while – as we pointed out in an earlier post, such an outcome would be consistent with a long-run price-level target – a level much higher than that is hard to imagine.
Turning to the second possibility, it is now commonplace to argue that the neutral or equilibrium real interest rate has declined. Economists expect the real rate to reflect the marginal return on capital in equilibrium. To the extent that the trend rate of growth has slowed – as the FOMC and many observers (including ourselves) believe – the equilibrium real rate should be lower than it was prior to the crisis. One recent estimate puts the neutral real rate at -0.4% as of the first quarter of 2014. If the FOMC believed this figure, it could account for most of the gap between the modified Taylor Rule and the actual funds rate.
However, FOMC members appear to be taking this negative estimate with a barrel of salt: their June interest rate projections imply a value in the range of 1½% to 2%. Indeed, model estimates of the neutral real rate are imprecise: In an earlier study, in which the model that now points to a -0.4% neutral real rate was developed, the authors reported a large standard error consistent with a 95% confidence interval on either side of the estimated rate of 3.7 percentage points! [For those with technical interest, this interval is from the baseline model which assumes that factors other than trend growth that affect the neutral real rate follow a random walk.]
What about resource slack? Well, you can look at either the unemployment rate or GDP. For the former, we have FOMC members’ estimates of the unemployment rate in the longer run, the lowest of which is 5%. It is hard to see how it could be much lower. Given the numbers we used to construct the chart, using this lower value would knock the modified Taylor policy rate estimate down to about 2%. To get that rate down to zero, we would need to assume an absurdly low equilibrium unemployment rate in the range of 3%. [Even the broadest U.S. measure of labor market slack (which takes account of under-employment) is now only 1.4 percentage points above its historical average – down by more than 5 percentage points from the cyclical peak.]
How about using GDP to measure slack? To explain the current policy rate, the gap between real GDP and potential GDP (the “output gap”) would have to be nearly 7%, significantly bigger than the CBO’s 5% estimate of the output gap in the first quarter. Yet, the CBO’s estimate is itself increasingly viewed as high, and seems likely to be revised down if (as we suspect) the CBO continues to lower its projection for trend growth.
Okay, so what about downside risks? It is hardly surprising for a central bank that cannot lower interest rates to seek some insurance against a negative surprise to aggregate demand. However, the potential benefit of such insurance diminishes when – as now – conventional policy objectives are neared amid buoyant financial conditions and signs of sustained growth momentum.
So, is it likely that the FOMC will wait “for a considerable time” beyond this year and then raise rates as the Committee’s median projections imply (see chart below)? Compared to the current Taylor rule rate, such a gradual policy normalization would buy quite a bit of downside insurance at a time when the downside risks look to be receding and the medium-term costs of accommodation appear to be rising. Yet, based on federal funds futures prices, that’s what investors appear to expect.
Perhaps it will turn out just so, but if recent U.S. job momentum persists for a few months, market rate expectations could change rapidly, despite the FOMC’s rhetoric.
FOMC End-Year Federal Funds Rate Projections and the Federal Funds Rate Implied by CME Futures