The guest commentary below was written by Dr. Daniel Thornton of D.L. Thornton Economics.
I have had several readers of my previous essay, Markets Needn’t Panic About Normalization, say that while they found the essay provocative and interesting, they nevertheless believe that QE was effective. To that I respond: What’s your evidence? Bernanke, Yellen, and others made it clear that the goal of QE was to reduce long-term yields.
The idea was that by reducing long-term yields, investment and, to a lesser extent, consumer spending would increase. They believed that the increased spending would increase output which would increase employment and, thereby, reduce the unemployment rate. According to the theory advanced by Bernanke, Yellen, and others, if QE didn’t reduce long-term yields it couldn’t have been effective. If you believe that QE was effective, you must believe that QE caused long-term yields to be significantly lower than they would have been otherwise. What’s your evidence?
There is Little Evidence QE Was Effective
As I noted in my previous essay, serious research finds a relatively small and statistically significant effect on the 10-year Treasury yield. What I didn’t say was all researchers—even researchers who used a flawed methodology and/or flawed data—found the effect on Treasuries to be much larger than the effect on private long-term securities, such as corporate bonds. Hence, whatever the analysis found the effect on long-term Treasuries to be, the effect of QE on long-term yields that matter for spending decisions was much smaller.
There are two other things to remember in assessing the effectiveness of QE. The first is that neither investment nor consumer spending is very sensitive to changes in interest rates. Hence, to have an economically significant effect on spending, the effect on private yields must be relatively large.
The second is that in order to affect spending, any reduction in long-term yields caused by QE must be persistent—it must last long enough for firms and consumers to respond, i.e., increase their spending. If it only lasts a few days or even a month, it could not have an effect on spending.
With all of this in mind, let’s consider some relatively simple facts about the behavior of the 10-year Treasury yield during QE1, QE2, and QE3, and whether it seems reasonable that QE was caused this behavior. Figure 1 below shows the 10-year Treasury yield from January 2007 to October 2017. The beginning and the ends of the three QE periods are denoted by vertical lines as is the period of tapering. The figure shows that during all three QE episodes, the 10-year yield was higher at the end of the Fed’s QE actions than it was when those actions were announced.
During QE1, which is from late November 2008, when the Fed announced that it would purchase $600 billion in securities, to March 2010, when the Fed announced that “the remaining transactions will be executed by the end of this month,” the 10-year yield increased by about 120 basis points. A similar thing happened during QE2 and QE3. Also, while the 10-year yield has cycled down since the end of QE2, on average it has been unchanged since the middle of 2011.
Moreover, it is clear from the figure that any announcement effects associated with the announcements of these actions were quickly offset. Indeed, the massive 50 basis point announcement effect that occurred on March 18, 2009, when the Fed announced it would purchase an additional $1.125 trillion in securities is barely noticeable using monthly data. As I noted in my research on event studies, this announcement effect was completely offset within 30 days of the announcement; hardly enough time for businesses and consumers to change their spending.
Now let’s consider the behavior of the 10-year Treasury yield over this period from another perspective. Undoubtedly many things affected the behavior of the 10-year yield over the period. However, one of these is unique to Treasury yields. Because Treasuries are default risk-free, the spread between securities that are not free of default risk and Treasury yields reflect market participants’ perception of default risk in the market. A widening spread indicates an increase in risk and a corresponding flight to safety—an increased demand for Treasuries. When market participants perceive the market is less risky, the spread narrows.
Figure 2 shows the spread between the Aaa Moody’s corporate bond yield and the 10-year Treasury from January 2000 to October 2017. Note that the spread that had been uncharacteristically low began to increase with the onset of the financial crisis and exploded after the recession. By the recession’s end, it had settled down and fluctuated in a range of about 150 to 200 basis points.
Over this period, the correlation between the 10-year Treasury yield and the rate spread was -94 percent. Hence, the flight to safety accounts for most of the behavior of the 10-year yield during the period; most of the marked decline and subsequent rise in the yield shown in Figure 1 up to the end of the recession is accounted for by changes in market participants’ assessment of risk, not by QE. The correlation between the 10-year yield and the spread after the recessions’ end is still substantial — -33 percent.
A Look At Global Bond Yields
While it is clear that market participants’ assessment of risk accounts for a lot of the behavior of the 10-year Treasury yield up to the end of the recession and some of its behavior thereafter, it is reasonable to assume that the 10-yield also responded to other factors such as the behavior of output and inflation and market participants’ expectation for these factors.
Given that financial markets are highly connected internationally, one would expect that these same factors would account for the behavior of other sovereign yields. Figure 3 shows the 10 -year sovereign yields for the U.S, German, Swiss, and New Zealand from January 2007 through October 2017. None of the other countries engaged in QE, yet they all trended down over that period.
That these yields didn’t simply mimic the behavior of the 10-year Treasury yield is evidenced by the facts that:
- None of the other yields experienced the marked decline (flight to safety) exhibited by the 10-year Treasury yield from January 2007 to the end of the recession, and
- The other three yields declined from about mid-2013 while the Treasury yield remained essentially unchanged. Hence, the downward trend in Treasury yield over the entire period cannot be due to the Fed’s QE policy.
It is likely due to global factors that were affecting all sovereign yields. Indeed, more often than not, the four yields cycled similarly suggesting that all yields were responding to similar non-idiosyncratic information. Combined, Figures 1 -- 3, suggest that the marked decline in long-term sovereign yields over that period was the consequence of factors, many of which were common, rather than to the QE policy formulated and implemented by the Fed. There is no compelling evidence that the Fed’s QE policy had a significant effect on the 10-year Treasury yield, let alone the long-term yields that matter for spending decisions.
For those who believe that QE significantly reduced long-term yields, show me your evidence. While you’re at it, tell me a convincing theoretical story of how QE reduced long-term yields. I would love to hear one. Neither Bernanke nor Yellen presented one. I doubt that chairman-elect Powell has one either. He seems to believe that these policies were effective. The press and market participants should force him to state exactly why he believes this. My advice to Powell is “listen to the Fed’s critics, they have important things to say. They’re not all wrong.”
My next essay is “How and Why the Fed Stumbled Into QE.”
This is a Hedgeye Guest Contributor piece written by Dr. Daniel Thornton. During his 33-year career at the St. Louis Fed, Thornton served as vice president and economic advisor. He currently runs D.L. Thornton Economics, an economic research consultancy. This piece does not necessarily reflect the opinion of Hedgeye.