Editor's Note: Below is a Hedgeye Guest Contributor research note 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. The ideas expressed below do not necessarily reflect the opinion of Hedgeye.
Fed Chair Janet Yellen gave a speech on Friday, August 26 at the Kansas City Fed’s annual monetary policy conference. The speech is important because Yellen discussed the motivation for and effectiveness of the Fed’s post-financial-crisis monetary policies and what the Fed’s “toolkit” will look like in the future. This essay makes a critical evaluation of a number of the speech’s claims. In particular, it shows why her analysis of past and future policy is wrong and essentially useless.
Yellen argued that the Fed’s monetary policy was effective even with its policy target near the zero lower bound on nominal interest rates, noting that the Fed provided substantially more accommodation by engaging in large-scale asset purchases and “increasingly explicit forward guidance.” By accommodation she means lower interest rates, in particular, lower long-term interest rates. Specifically, she suggested “Our purchases of Treasury and mortgagerelated securities in the open market pushed down longer-term borrowing rates for millions of American families and businesses.” This is an interesting assertion for a couple of reasons, not the least of which is the FACT that Yellen was not always that certain about their effectiveness. Figure 1 shows the 10-year Treasury yield daily from March 2 to June 30, 2009. The three vertical lines represent three interesting dates, March 18, April 29, and June 24.
March 18, of course, is when the FOMC announced that the Fed would purchase up to $1.75 trillion in mortgagebacked securities (MBS), agency debt, and long-term Treasuries. The 10-year yield declined 51 basis points that day! On April 29, Yellen said
I think our actions at the last meeting provided compelling evidence that purchases of longer-term Treasury securities worked to bring down borrowing rates and improved financial conditions more broadly. Now that we’ve tested the waters, it’s time to wade in by substantially increasing our purchases of Treasury securities. In fact, I could support increasing the amount beyond $750 billion and would consider including securities of durations down to one year (Transcript of the April 28-29 FOMC meeting, p. 121).
June 24 is important because it’s the date Yellen changed her mind. Specifically, she said
Initially I was an enthusiast for long-term Treasury purchases. I thought the purpose of it was not only to improve liquidity and market functioning, but also to influence yields to push them down…On theoretical grounds, I believe there’s a very strong case that they should have some effect, but it has been awfully hard to identify exactly what that effect is, and I think that we’re beginning to run into costs of pursuing that further (Transcript of the June 23-24 FOMC meeting, pp. 80-81).
What caused Yellen to change her mind? She didn’t say, but Figure 1 suggests that it was likely the FACT that by June 24 the 10-year Treasury yield was 70 basis points higher than it was on March 17, and 121 basis points higher than it was on March 18.
Why does Yellen now believe that QE and forward guidance have been effective in reducing long-term yields? One reason is long-term rates have declined over the entire period since the Fed introduced these policies. Figure 2, shows the 10-year Treasury yield monthly over the period from September 2008 to March 2015. The yield trended down over the entire period, suggesting that these policies have been quite effective.
Figure 2 also shows the 10-year sovereign bond yields for the U.K., Germany, and Australia, which have also trended down over the entire period. This suggests the policy not only significantly reduced long-term bond yields for all Americans, but it reduced them for the Brits, Germans, and Australians. Indeed, the other yields declined relative to the Treasury yield over the period. The Fed’s policy was more effective abroad than it was 3 in the U.S. Ok, I don’t believe this either. The four yields have not only trended down similarly, but they have very similar cycles. Hence, it seems more likely that sovereign bond yields in all four countries were affected by more or less the same events that need not include the Fed’s large-scale-asset purchases or its forward guidance. Indeed, while the Fed’s purchases of long-term securities were large, they are trivial relative to the size of the international long-term debt market; smaller still relative to the entire credit market. It seems unlikely that such a small increase in the demand for credit would have had the effect on long-term yields shown in Figure 2.
My contention is Yellen actually has no idea how much, if any, of the decline in Treasury yields was due the Fed’s aggressive actions. Indeed, it appears that Yellen and Bernanke do not even agree on about how QE reduced long-term yields. Bernanke suggested QE was effective because it reduced the term premium on longterm securities, indicating that term-premium effect was large, here, here, here, and here. In contrast, Yellen reports that the effect through the term premium “may be relatively small.” Personally I agree with Yellen. As I show elsewhere, Requiem for QE, QE should not affect the term premium because the term premium on a security is determined by two factors (investor’s relative degree of risk aversion and the duration of the specific bond) that are independent of the quantity of particular bonds in the market. QE can only affect the term premium if investors who are most risk adverse leave the security of default-risk-free Treasuries while the least risk adverse investors stay; this seems highly unlikely.
Yellen’s rational for why the FOMC decided to maintain a large balance sheet when it decided to raise the funds rate is illogical. She notes the FOMC considered reducing the size of the balance sheet but decided against it because the Committee was less certain of the effects of changing the balance sheet on the economy. She goes on to say the Committee was concerned that “Excessive inflationary pressures could arise if assets were sold too slowly.” I guess they thought inflationary pressures wouldn’t build up if they just keep the balance sheet large indefinitely. In her discussion of policy going forward she said, “Once we stop reinvestment, it should take several years for our asset holdings--and the bank reserves used to finance them--to passively decline to a more normal level.” Concerns about “excessive inflationary pressures” due to a large balance sheet must have vanished quickly!
Apparently, the Committee did not consider experimenting with shrinking the balance sheet. This could have been done easily. For example, they could have started by announcing that they would no longer reinvest maturing assets and allow the balance sheet to contract slowly. If that caused no major disturbance, after a while they could announce that they would begin selling some of their assets very slowly by using reverse repurchase agreements (RRPs). After all, the Fed has sold about $350 billion of securities using RRPs in an effort to keep the funds rate close to the target level. If that didn’t cause any significant disruption in financial markets, the could commence outright sales; at first, slowly by converting RRPs into outright sales and more rapidly later if there were no significant disruptions, as I suggested here. Continuously rolling over temporary sales of a given amount of RRPs has the same effect on the total supply of credit as permanent sales of an equal amount. Indeed, the monetary base, a measure of the Fed’s contribution to the supply of credit, declined by a similar amount ($303 billion) with no adverse consequences.
Equally noteworthy is Yellen’s statement that long-term yields here and abroad declined over the past decade because the long-run neutral real rate of interest (“the inflation-adjusted short-term interest rate consistent with keeping output at its potential on average over time”) declined. She attributes the decline in the neutral real rate to a variety of possible factors: slower growth in the working-age populations, smaller gains in productivity, decreases in the propensity to spend in the wake of the financial crises around the world since the late1990s, and the “paucity of attractive capital projects worldwide.” However, she does not say how she is able to distinguish between the effect of these factors on long-term yields and the effect of the Fed’s QE and forward guidance policies. She concluded this discussion by saying “Although these factors may help explain why bond yields 4 have fallen to such low levels here and abroad, our understanding of the forces driving long-run trends in interest rates is nevertheless limited, and thus all predictions in this area are highly uncertain.” But apparently not so uncertain that she is unable to identify the separate effects of QE and forward guidance. It would be nice to know how much of the decade-long decline in yields is due to the long-run neutral real rate of interest and how much is due to the Fed’s policies. My guess, backed up by economic and finance theory and evidence (Thornton 2014; Kool and Thornton, 2016), is close to 100% for the former and close to 0% for the latter. Some of the Fed’s policy actions have had significant temporary effects on bond yields when they were announced, but no permanent effects (Thornton, 2016a).
Particularly interesting is Yellen’s analysis of what the Fed did with the federal funds rate over the last nine recessions. Referring to Table 1 of her speech, she says “the FOMC cut the federal funds rate by amounts ranging from about 3 percentage points to more than 10 percentage points. On average, the FOMC reduced rates by about 5-1/2 percentage points, which seems to suggest that the FOMC would face a shortfall of about 2-1/2 percentage points for dealing with an average-sized recession.” This is an incredible statement for a variety of reasons, not the least of which is the FACT that, as those of us who have been long-time students of the Federal Reserve know, the federal funds rate was capped by the Federal Reserve’s discount rate from 1957 to late 1964. This is shown in Figure 3, which shows the federal funds rate, the discount rate, and the 3-month
Treasury bill rate from July 1954 through December 1967. The figure shows that the funds rate was capped by the discount rate until late 1964. The 3-month T-bill rate was not. The reason is that during this period the federal funds market was used by banks solely to adjust their reserve positions to meet the Fed’s imposed statutory reserve requirements. When the funds rate was below the discount rate, banks would adjust their reserve positions mostly in the federal funds market. When interest rates moved up to or went above the discount rate, banks would obtain the reserves they needed at the Fed’s discount window. Hence, the funds rate never went above the discount rate.
Meulendyke (1998, p. 38) explains why the changed occurred. She notes that large banks began actively managing the liability side of their balance sheets and began borrowing in the funds market in a “sustained way” because certificates of deposits (CDs), an important source of funds for bank lending, were subject to reserve requirements and Regulation Q interest rate ceilings; funds obtained by borrowing continuously in the overnight federal funds market were not. Furthermore, she notes that funds obtained in the federal funds market “were not subject to restrictions on prolonged use that were applied to the Federal Reserve’s discount window.”
The fact that the funds rate was capped by the discount rate during the first two of the nine recessions (August 1957 to April 1958 and April 1960 to February 1961) means the FOMC did not cut the funds rate by 2.9 percent and 2.8 percent, respectively, during those two recessions. The Board of Governors reduced the discount rate during those recessions. Moreover, it is clear from Figure 3, the Board did this in response to changes in market rates, which had already declined substantially, not as a policy action. Indeed, there is no evidence that the FOMC used the funds rate as a policy instrument before the late 1980s (see Thornton 2001, 2006, 2016b, and Garfinkel and Thornton, 1995). Hence, the claim that the FOMC reduced the funds rate by the amounts shown in Yellen’s table any time prior to the last three recessions is doubtful, if not fanciful.
This is important because these data feed into a model whose results Yellen uses to discuss the simulated path of the effects of three different monetary policy responses. The simulated paths for the federal funds rate, the 10-year Treasury yield, the unemployment rate, and the inflation rate are shown in Figures 2 of her speech reproduced here.
The figure shows the response to three different policy responses: “the aggressive rule in the absence of the zero lower bound constraint, the constrained aggressive rule (the federal funds rate cannot go below zero), and the constrained aggressive rule combined with $2 trillion in asset purchases and guidance that the federal funds rate will depart from the rule by staying lower for longer” (parentheses added). She notes that asset purchases and forward guidance can push long-term rates lower and produce better outcomes for unemployment and inflation. She cautions, however, that “this analysis could be too optimistic” because “the FRB/US simulations may overstate the effectiveness of forward guidance and asset purchases, particularly in an environment where long-term interest rates are also likely to be unusually low,” and the real neutral rate might be close to zero which implies that “asset purchases and forward guidance might have to be pushed to extremes to compensate.”
This is true, but these are the least of the analysis’ problems. The analysis is contaminated by the FACT that it is based on fictitious data about policy actions during all but the last three recessions. There is an old saying among economists who do empirical analyses, “garbage in, garbage out.” Moreover, the figures also suggest a degree of precision that is unwarranted. I am confident that if the simulated response lines were presented with their 90 percent confidence intervals, rather than just the paths, the three lines would lie comfortably within any one of the three confidence intervals. Conclusion: the uncertainty about the response paths is so great that they tell us nothing interesting about how the responses are likely to be different; no mean path (line) is statistically distinguishable from any other. Economists tend to report data this way because, if they didn’t, they would have nothing to say.
Yellen had one final caution,
relying too heavily on these nontraditional tools could have unintended consequences. For example, if future policymakers responded to a severe recession by announcing their intention to keep the federal funds rate near zero for a very long time after the economy had substantially recovered and followed through on that guidance, then they might inadvertently encourage excessive risk-taking and so undermine financial stability.
Apparently, she is unaware of the FACT that this has already happened. Pension funds are holding more risky portfolios than they should. Retires face an uncomfortable choice—living on less or taking more risks, equity and home prices have reached what are likely to be unsustainable levels. Banks have made more risky loans than they would have made if they were not holding massive amounts of excess reserves and are poised to make even riskier loans. The M1 measure of the money supply has increased by an unprecedented amount since the end of the recession (June 2009). This may ultimately lead to a significant increase in inflation. See My Scary Graph, Monetary Policy Insanity, Normalize Now, and Unintended Consequence for a discussion of these consequences of the Fed’s policies.
- Garfinkel, M.R. and D.L. Thornton. (1995). “The Information Content of the Federal Funds Rate: Is It Unique?” Journal of Money, Credit and Banking, 27(3), 838-47,
- Kool, C.J.M. and D.L. Thornton. (2016). Federal Reserve Bank of St. Louis Review, 97(4), 303-22, here.
- Muelendyke, A-M. (1998). U.S. Monetary Policy & Financial Markets, Federal Reserve Bank of New York, N.Y.
- Thornton, D.L. (2001). “The Federal Reserve’s Operating Procedure, Nonborrowed Reserves, Borrowed Reserves and the Liquidity Effect,” Journal of Banking and Finance, 25(9), 1717-39, here.
- Thornton, D.L. (2006). “When Did the FOMC Begin Targeting the Federal Funds Rate? What the Verbatim Transcripts Tell Us,” Journal of Money, Credit and Banking, 38(8), 2039-71, here.
- Thornton, D.L. (2014). QE: Is There a Portfolio Balance Effect,” Federal Reserve Bank of St. Louis Review, 96(1), 55-72, here.
- Thornton, D.L. (2016a). “The Effectiveness of QE: An Assessment of the Event-Study Evidence,” unpublished manuscript
- Thornton, D.L. (2016b). “Greenspan’s Conundrum and the Fed’s Ability to Affect Long-Term Yields,” unpublished manuscript, here.