Alan Greenspan Answers Questions on the Neutral Real Rate, the Yield Curve, and Transparency

Today, a letter from Alan Greenspan in response to questions by Jim Saxton, Chairman of the Joint Economic Committee was released. Here's a transcription of the pdf file. I couldn't find any copies other than the pdf version, so I think I will leave the transcript uncut in the continuation frame to give easy access to the full text.

For anyone interested in the yield curve conundrum, it is discussed in some detail. Greenspan says that too much saving and too little investment is the cause of a flat yield curve. It does not appear the Fed believes that the yield curve provides information over and above what they learn from other sources and it gives unreliable signals about the future state of the economy. Therefore the yield curve does not have much impact on policy decisions. There is also a discussion at the beginning on the feasibility and desirability of using the real interest rate as a policy target, and a brief discussion of transparency at the end. Greenspan deos not believe that a strict real interest target is feasible, and he believes the current level of transparency is the correct level:

November 28, 2005

The Honorable Jim Saxton
Chairman
Joint Economic Committee
Washington, D.C. 20510

Dear Mr. Chairman:

I am pleased to enclose my responses to the additional questions you forwarded in connection with the November 3 hearing.

I also wanted to thank you, and the other members of the committee, for your kind and generous comments at the hearing and in your letter. It has been a pleasure appearing before the Joint Economic Committee over the years.

Sincerely,

Alan Greenspan

Chairman Greenspan subsequently submitted the following to written questions received from Chairman Saxton in connection with the Joint Economic Committee hearing on November 3, 2005:

Q.1. Since the "neutral" rate is not observable, how do you know when you've reached the neutral rate? What variables do you monitor to make judgments as to how close to neutral the fed funds rate is? As the fed funds rate is ratcheted up, and given the lags that exist, does the possibility of raising it above a neutral level increase?

A.1. Although the concept of a "neutral interest rate" is a useful theoretical construct, difficulties in implementing it in practice limit its usefulness as a framework for monetary policymaking. For one thing, a variety of definitions of a neutral real interest rate are possible. For another, quantitative estimates of the level of such a rate are subject to considerable uncertainty. Also, such estimates can vary widely depending on the type of measure and the prevailing and projected economic conditions. In particular, all variables that contribute to making a macroeconomic forecast relevant for estimates of neutral interest rates, greatly complicating such assessments. Thus, it is impossible to know with any certainty when the neutral rate has been reached. Moreover, the use of neutral real rates in the formulation of monetary policy in not necessarily straightforward. For instance, in some circumstances, attaining a "neutral" federal funds rate would in principle be an appropriate objective for monetary p0licy, but in others--particularly when inflation is too high or too low--aiming for a neutral funds rate in the near term would not be appropriate. These uncertainties and complications suggest that reliance on a single summary measure such as a neutral real interest rate would be unwise as a strategy for formulating monetary policy. Rather, a full consideration of current and perspective economic developments, and of the risks to the outlook, is essential for the conduct of monetary policy.

Q.2. Over the last year and a half, the Federal Reserve has raised the federal funds rate by 3.0 percentage points and indicated that further increases are likely in order to check inflation. Yet long-term rates, including mortgages, are lower now than when the FOMC began tightening. In past comments, you have termed this situation a "conundrum" without recent precedent. What explains the low level of long-term interest rates?

A.2. As I noted in my monetary policy testimony before the Congress in July, two distinct but overlapping developments appear to be at work in explaining the low level of long-term interest rates:  a longer-term trend decline in bond yields and an acceleration of that trend over the period since mid-2004. Both developments are particularly evident in the nominal interest rate applying to the one-year period ending ten years from today that can be inferred from the U.S. Treasury yield curve. In 1994, that so-called forward rate exceeded 8 percent. By mid-2004, it had declined to about 6-1/2 percent--an easing of about 15 basis points per year on average. Over the past year, that drop steepened, and the forward rate fell 130 basis points to less than 5 percent.

Some, but not all, of the decade-long trend decline in that forward yield can be ascribed to expectations of lower inflation, a reduced risk premium resulting from less inflation volatility, and a smaller real term premium that seems due to a moderation of the business cycle over the past few decades. As I noted in my testimony before the Joint Economic Committee in February, the effective productive capacity of the global economy has substantially increased, in part because of the breakup of the Soviet Union and the integration of China and India into the global marketplace. And this increase in capacity, in turn, has doubtless contributed to expectations of lower inflation and lower inflation-risk premiums.

In addition to these factors, the trend reduction worldwide in long-term yields surely reflects an excess of intended saving over intended investment. This configuration is equivalent to an excess of the supply of funds relative to the demand for investment. Because the intended capital investment is to some extent driven by forces independent of those governing intended saving, the gap between intended saving and investment can be quite wide and variable. It is real interest rates that bring actual capital investment worldwide and its means of financing, global saving, into equality. We can directly observe only the actual flows, not the saving and investment tendencies. As best we can judge, both high levels of intended saving and low levels of intended investment have combined to lower long-term interest rates over the past decade.

Q.3. I was intrigued by your response to my question relating to the yield curve and associated yield spread between the fed funds rate and the 10-year bond yield. In particular, your response to the spread question was as follows:

"...that used to be one of the...most accurate measures we used to have to indicate when a recession was about to occur and when a recovery was about to occur. It has lost its capability of doing so in recent years...it has significant financial impacts, it's no longer useful as a leading indicator to the extent that it was."

In pondering this comment, three considerations appear to be especially relevant: (1) First, the importance of a yield spread for monetary policy has long been recognized by classical economists. Both Henry Thornton and Knut Wicksell recognized that when the central-bank-controlled short-rate moves relative to a long-term market rate, relative prices, incentives, and behaviors change. (2) Second, the recent (2005) extensive review and summary of the literature pertaining to research on the yield spread (published by the Federal Reserve Bank of New York) concludes that the weight of the evidence supports the potency of the yield spread.  (See Estrella, October 2005).  (3) Third, the Conference Board includes a yield spread variable in its index of leading economic indicators. The Conference Board conducts an ongoing evaluation of these indicators and an especially thorough, major reevaluation of the composite was made last July. The bottom line is that the yield spread remains a key component of this composite.

In light of these considerations, what available evidence or other factors support the view that the yield spread is no longer especially useful?  Has the Board staff assessed this relationship recently?

A.3. Although the slope of the yield curve remains an important financial indicator, it needs to be interpreted carefully. In particular, a flattening of the yield curve is a not a foolproof indicator of future weakness. For example, the yield curve narrowed sharply over the period 1992-1994 even as the economy was entering the longest sustained expansion of the postwar period.

Three basic factors affect the slope of the yield-curve--the current level of the real federal funds rate relative to the long-run level, the level of near-term inflation expectations relative to expected inflation at longer horizons, and the level of the near-term risk premiums relative to risk premiums at longer horizons.

Statistical analysis indicates that the first factor--the gap between the current and long-run levels of the real federal funds rate--is a key component from which the yield curve slope derives much of its predictive power for future GDP growth. When the level of the real federal funds rate is pushed well below its long-run level, economic stimulus is imparted and the yield curve steepens. The economic stimulus influences output growth with a lag; as a result, the steepening of the yield curve in this scenario is a predictor, albeit not the cause of, stronger economic activity ahead. Conversely, when the level of the real federal funds rate is pushed above its long-run level, economic restraint is imparted and the yield curve flattens. Once again, the economic restraint influences output growth with a lag, so the flattening (inversion) of the yield curve in this scenario would signal weaker economic growth ahead, but would not itself be the cause of the weakening.

The connection between future output growth and the other two factors affecting the slope of the yield curve--the gap between near-term and long-term inflation expectations and the difference between near-term and long-term risk premiums--is far less certain and likely to depend on economic circumstances. For example, a rise in near-term inflation expectations above long-term inflation expectations would tend to flatten the yield curve and might also signal a prospective weakening in aggregate demand.  This configuration in inflation expectations might reflect adverse supply factors that have pushed up inflation in the near term but that are expected to dissipate over time. In this case, the flattening of the yield curve might well be a signal of an improving inflation picture that could also be accompanied by a favorite outlook for economic growth.

The connection between output growth and risk premiums is also quite uncertain. A fall in distant horizon risk premiums would flatten the yield curve and might signal a weakening in economic activity if, for example, the drop in risk premiums in fixed-income markets was associated with a "flight to safety" on the part of global investors seeking a safe haven from turbulence in equity markets and other risky assets. But it is also possible that a decline in distant horizon risk premiums could be a sign that investors are generally more willing to bear risk. In this case, a flattening of the yield curve stemming from this factor could be an indicator of an easing in financial conditions that would stimulate future economic activity.

In summary, many factors can affect the slope of the yield curve, and these factors do not all have the same implications for future output growth. In judging the indicator value of any particular change in the slope of the yield curve, it is critical to understand the underlying forces that may be affecting the yield curve at the moment. As the 1992-1994 episode attests, simply relying on an average statistical relationship estimated over a very long sample can be quite misleading.

Q.4. One of the strategies or institutional changes that you have supported in recent years relates to the increased transparency of the Federal reserve. This increase Federal Reserve transparency has, for the most part, been associated with more benefits than costs. Doesn't this increased transparency work to the benefit of both the Federal Reserve and the public?

A.4. Greater transparency with regard to Federal Reserve actions encourages public discussion and informed scrutiny, important aspects of accountability in a democratic society. Transparency also enables financial markets to better predict monetary policy decisions, which can contribute to improved policy outcomes. However, providing more complete information about policy decisions is not without cost. Transparency requires careful attention by policymakers, and so constrains the time they have for actually making decisions. More importantly, excessive transparency could inhibit policymakers, making them less spontaneous in their remarks and less willing to explore new ideas. Such an outcome would have adverse effects on policy decisions. The Federal Reserve's current practices strike a reasonable balance between transparency and the degree of confidentiality appropriate to support the policy process.

Comments

[sharkbait]

Apparently Greenspan does seem to think that the yield curve has some predictive powers. Although he does make it clear that the slope of the yield curve is not determined by expectations of future output alone. From his analysis it follows that if the yield is sufficiently inverted, and its negative slope cannot be explained by dimished inflation expectations nor by diminished risk premia alone, such a yield curve would in fact be predicting a slowdown in output.

[Mark Thoma]

True. My point is that by the time you sort that all out, you already know most, if not all of what the yield curve will tell you...

[Bruce Wilder]

I woke up this morning thinking about this, believe it or not. Greenspan's remarks really bother me. Am I the only one, who thinks that the yield curve is instrumental as well as informational?

Long-term rates are puzzlingly low, and there is a housing boom/bubble, driven by low long-term rates. When the housing bubble bursts, which may be now, the Fed will want to greet it with plenty of liquidity. Instead, the Fed is pushing up short-term rates, because, I guess, that's what the Fed does (-- the only it can do, legitimately, in some people's limited imaginations). If the Fed pushes up short-term rates high enough to invert the yield curve, intermediation is choked and a credit crunch will curtail long-term lending ("economic restraint is imparted" in Greenspan's opaque and passive phrase), bursting the housing bubble, probably causing a recession, which will induce the Fed to lower short-term rates. Tell me again, why do we want to go on that roller coaster?

What's wrong with just pushing up long-term rates and keeping the yield curve upright? Push up long-term rates, burst the housing bubble, leave short-term rates and liquidity alone to help the economy survive the inevitable housing slowdown.

The Treasury's debt maturity structure is such that there's plenty of room to push long-term debt onto the market, if that's what it takes to push up long-term rates. (And, if it comes to that, the Federal government also dominates the mortgage securities market, doesn't it?)

If the national debt had a longer maturity structure, the market would worry more that the Fed gov't would be tempted by inflation; long-term inflationary expectations would rise, and that would be a good thing, because it would also increase the political pressure for deficit reduction.

All the talk about global savings gluts strikes me as somewhat disingenuous. Like all markets, there's supply AND demand -- a geometry where two lines (over)determine a single point. Those prodigious savers in China and Japan are not saving, originally, in dollars. There's a critical institutional intervention there, before that supposed savings glut gets dumped into dollar world.

And, it is weird to hear Greenspan narrate a story about "low levels of intended investment" at a time when stupendous advances in computing and communication technology, is driving rapid productivity improvement. Call me a naive Schumpeterian if you like, but it looks to me like we are on the beginning of one of his long waves, and ought to be investing furiously in computers and biotechnology; the rapid improvement in factor productivity ought to be a clue that potential returns on investment are actually high, not depressed.

Do you think, just maybe, that depressed long-term industrial investment in the U.S. might have something to do with the tilt given by exchange rates to international trade conditions? It looks to me like that manipulation of exchange rates has made manufacturing investment in the U.S. artificially unprofitable, with the result that the manufacturing base is being eroded here, while China furiously builds a new (and not incidentally, post-computer) base there. Shouldn't we be holding Bush and Greenspan responsible for the enormous trade deficit and the decline of manufacturing employment at a time when the technology frontier is wide-open?

I am sorry if this just seems like a useless rant. It has been 25 years since I worked as an economist, and the depth of my expertise was nothing to write home about, even then. I think that what I have said has a firm foundation in economic reasoning, even if my rhetoric is a bit over the top. I feel frustrated, because it seems like the expression of an important point of view is being excluded by the language used by many economists, including Greenspan.

The whole "debate" over whether an inverted yield curve is a "still" a reliable leading indicator seems to me a clever way of giving the impression that the issues are technical and esoteric, when they are not. The maturity structure of federal debt and erosion of manufacturing in the U.S. should not be excluded from the policy debate. The wisdom of responding to an economic problem (the housing boom/bubble) relating to long-term rates, with a squeeze on short-term rates, ought to seem questionable on its face.

[Anon]

Bruce, How does the fed push rates (short or long term) up? I understand that they can push rates down by creating money and purchasing debt in the open market. To push long rates down, don't they have to have a large supply of long term debt instruments to push into the market?

[Max]

Anon,

it's exactly how they do it.

What Bruce is suggesting is to stop FOMC purchases of long-term Treasuries.

[knzn]

As far as I know, the Fed purchases few (if any) long-term Treasuries to begin with, so there wouldn’t be much to stop. They would need the cooperation of the Treasury in order to push up long-term rates. But it raises the question of why the Treasury doesn’t finance longer-term. Inflation expectations should not be a major issue, because they could finance in part with TIPS. It seems irrational for the Treasury not to take advantage of the opportunity to lock in historically low interest rates, especially now that the yield curve has flattened, and I can’t think of any public policy reason to keep the yield curve deliberately flat. Perhaps the objective is to “starve the beast” in the future by following bad debt management policy intentionally.

[Bruce Wilder]

Anon, yes, they would have to sell long-term securities.

Conveniently, the Treasury is in a position to make an ample supply available.

My larger point is that inverting the yield curve is a policy choice. The yield curve is a mechanism, central to the banking system's functioning. Banks borrow at short-term rates to loan money at long-term rates; banks depend on long-term rates being higher than short-term rates. Inversion is an extreme measure, and no one should be surprised that inversions cause recessions -- and "cause", not "predict", is the right word; the yield curve is a mechanism, not a statistical curiosity. As a highly risky policy choice, there really ought to be a lot more attention paid to alternative interventions; having the Treasury aggressively shift its maturity structure would be an obvious one, better suited to the case at hand.

[Anon]

Lengthening the maturity structure would worsen current budget deficits. That's why treasury shortened it in the first place: to save money in the short run so the tax cuts wouldn't look like such a bad idea.

[Bruce Wilder]

Anon: the maturity structure shortened in the Clinton administration as a means to reduce the debt service. By redeeming a lot of long-term bonds, which paid very high rates of interest, and taking advantage of low, short-term rates, they were able to accelerate the reduction in debt service costs, which allowed Clinton to claim a surplus.

With a nearly flat yield curve, long-term debt issued now would bear close to the same interest cost as short-term debt. With an inverted curve, long-term issues would actually be cheaper.