Before the Great Recession, monetary policy was relatively simple. If inflation trends were too strong (above 2%) or the unemployment rate too low, the Federal Reserve would increase the fed funds rate as a measure to tighten monetary policy. The opposite was also true (low inflation, high unemployment -- lower fed funds rate).
By the end of 2008, the fed funds rate was lowered to such an extent that it reached the zero lower bound and could not be adjusted any lower. This would not be a problem if the economy was showing signs of improvement. Most iterations of the Taylor rule showed that the fed funds rate should, theoretically, be negative in order to spur economic growth.
Since the fed funds rate cannot be lower than zero, the Fed started using new unconventional measures (quantitative easing) to ease monetary policy further. Unfortunately, the effect of quantitative easing on the economy has been difficult to measure. Thus, there have been lots of calls by prominent economists and lawmakers to end the latest round of quantitative easing.
In a new paper by Wu and Xia (2013), the effects of quantitative easing may finally be settled. Furthermore, a mathematically viable model on when to reduce quantitative easing is now available.
Shadow Fed Funds Rates
Typically, the effects of monetary policy are evaluated using a vector autoregression and regressing changes in macroeconomic variables, such as unemployment or inflation, along with changes in the effective funds rate. The obvious problem with that method today is that the effective fed funds rate has not changed over the past few years since quantitative easing began.
Wu and Xia attempt to mitigate that problem by calculating a shadow fed funds rate.
Bond yields are a combination of expected future short-term rates -- in this case the fed funds rate -- and expected inflation. Wu and Xia essentially calculate the current fed funds rate, which they refer to as the shadow rate, by evaluating the term structure of bond yields in reverse. The known yield, from the long end to the short end, can tease out the fed funds rate. The more wonkish and technical aspects of the model can be found in the Wu and Xia paper. In layman's terms, the authors calculate the fed funds rate using the implied forward rates at any given time across the entire yield curve.
Using data going back to 1960, the effective fed funds rate and the estimated shadow rate essentially mirror each other until 2008. The average difference between the two measures during that time was only 4 basis points. These findings are statistically robust.
The shadow rate turned negative in 2009, which indicates that quantitative easing was successful in flattening the yield curve even though the effective rate was unchanged. In other words, quantitative easing was able to lower long-term bond yields in accordance with a fed funds rate that was below zero.
The Effectiveness of Quantitative Easing
The shadow fed funds rate has been basically unchanged since the start of QE2 and the extension of QE3. This suggests the expansion of the Fed's balance sheet from quantitative easing purchases has definitely produced diminishing marginal returns. Surprisingly, the talk of tapering that began this summer, which drove up long-term rates and caused an upward shift in the fed funds future curve, caused the shadow rate to fall to its lowest point (-1.68%). The reaction reinforces the Fed's theory that tapering is not a form of tighter monetary policy.
Using these shadow rates in a vector autoregression on the data from July 2009 through May 2013 (incorporating a portion of QE1, all of QE2, Operation Twist, and the beginning of QE3), Wu and Xia determined that quantitative easing was responsible for lowering the unemployment rate by 0.23 percentage points.
In other words, the Fed increased its balance sheet by about $1.4 tln, and that successfully reduced the unemployment rate to 7.6% from what would have been 7.8% in May 2013 if no quantitative easing had taken place.
Other macroeconomic gains from quantitative easing were just as weak. Quantitative easing added just 34,000 new housing starts (914,000 instead of 880,000). The industrial production index was 2.0 points higher (98.7 vs. 96.7) after quantitative easing. Capacity utilization rates were 0.6% higher as a result of quantitative easing.
In all, the findings by Wu and Xia confirm what the Curdia and Ferrero (2013) of the Federal Reserve Bank of San Francisco reported in August. The massive amount of quantitative easing did very little to stimulate the economy.
These results will only fuel the criticism of quantitative easing. The costs of increasing the balance sheet, which are currently unknown, have done little to benefit the economy.
Should the Fed Continue with Quantitative Easing?
Even if we assume that there are no costs from quantitative easing, there is a real question on whether the Fed should expand, taper, or even completely end quantitative easing. The Taylor Rule, which predicts what the fed funds rate should be given the current inflation and unemployment rates, does provide some guidance.
The original Taylor Rule that was created in 1993 assumed that the Fed's dual mandates were equally important. The Fed should react just as much to maintain low and steady inflation as it should to stimulate full employment.
However, from 1987, when Alan Greenspan became chairman of the Fed, through 2007, FOMC policy actually tilted more towards lowering the unemployment rate versus maintaining 2.0% inflation. A modified Taylor Rule, where full employment is twice as important as low inflation, thus shows a more negative fed funds rate prediction than the original Taylor Rule.
Both measures show that the Fed needed to do more to ease monetary conditions from 2009-2011 given how poorly the economy was performing. The effect of quantitative easing on the shadow fed funds rate was not enough to spur economic growth. More quantitative easing was needed at the beginning of the recession than what was actually provided.
Simply put, the packages of quantitative easing instituted by the Fed were not large enough to match the softness in economic conditions.
Interestingly, both measures also show that the Fed should not only begin tapering today, but possibly sell assets to increase the shadow fed funds rate. While the original Taylor rule calls for a positive fed funds rate, which is unlikely to happen, the modified Taylor rule predicts a fed funds rate nearly 100 bps points higher than the current shadow fed funds rate.
The data suggest the Fed is providing too much stimulus given current economic conditions.
There is an argument that elevated levels of discouraged workers has biased the unemployment rate so that it does not accurately reflect current slack in the labor market. Yet, even if the unemployment rate were half a percent higher than its current level, the modified Taylor rule would still be higher than the shadow fed funds rate. Thus, the "real" unemployment rate would need to be larger than 7.8% to justify not tapering.
New research by Wu and Xia shows the effects of quantitative easing on the fed funds rate. Unfortunately, the Fed underestimated the necessary size of the quantitative easing measures when it began enacting unconventional policies.
The Fed has been able to reduce the unemployment rate by 0.23 percentage points through quantitative easing and minimally stimulate other areas of the economy.
Both the original and modified Taylor rules suggest economic conditions do not warrant as much quantitative easing as the Fed is now providing. The economy can do well even if the Fed takes its foot off the gas pedal.