December 17, 2005
The real yield curve
Economist's View reported last week on a letter from Alan Greenspan that addressed some questions about monetary policy posed for the Fed Chair by Representative Jim Saxton (R-NJ). I was particularly interested in Greenspan's explanation of why he is not concerned about the seemingly bearish connotations of the slope of the yield curve.
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.
I had the impression from these remarks that Greenspan was suggesting that the recent flattening of the yield curve may be a reflection of changes in inflation expectations or risk premiums and therefore not a source of concern. I disagree, by the way, with Greenspan's assertion that changes in the risk premium are not predictive of future changes in real economic activity. Evidence detailed in an article by Dong Heon Kim and me published in the Journal of Money, Credit, and Banking showed that the term premium is indeed an important part of the historical relation. But putting this issue aside, I was curious to see what Greenspan was referring to in terms of the slope of the real yield curve.
The dashed red line in the figure at the right plots the yields on inflation-indexed securities of different maturities as they stood on June 15, 2004 (data from FRED). The real yield curve then exhibited the same upward slope as did the nominal yield curve at that time (e.g., here). As of this week, however, the real yield curve (solid blue line) is another story. The TIPS maturing 1 year from January is currently yielding 2.37%, a higher yield than that offered by TIPS of even the longest maturity. If Greenspan is right and it's the slope of the real yield curve that really matters, then we've already reached an inversion.
Perhaps Greenspan is looking not at the TIPS yields but rather at the nominal fed funds rate minus expected inflation. What fed funds rate is likely to prevail over the next year? Fed funds futures seem to reflect an assumption that we'll be at 4.75% by March. Inflation as measured by the median CPI, which I prefer to use for purposes of forecasting the future CPI, has been coming in steadily at a 2.3% annual inflation rate for quite a while, though it ticked up to 2.4% for the 12 months ended November 2005. If we take 4.75 minus 2.4, we get 2.35, almost exactly the 2.37% yield on the inflation-indexed Treasury maturing January 2007, and again, higher than the yield on inflation-indexed securities of longer maturities.
The inversion of the real yield curve seems to have already arrived.
Posted by James Hamilton at December 17, 2005 07:37 PMdigg this | reddit
Listed below are links to weblogs that reference The real yield curve:
Tracked on December 26, 2005 06:34 AM
» This Week's Carnival of The Capitalists from Financial Rounds
This week's COTC is up at Multiple Mentalities. Here are my picks of the week: they include pieces by Multiple Mentalities, Interested Participant, Econbrowser, The Other Bloke's Blog, The Real Returns, Finalcial Revolution, and Blueprint For Financ... [Read More]
Tracked on December 26, 2005 11:23 AM
Very good post. And the yield curve is not the only signal of an approacing recession.
Add the peak in housing prices, overall complacency, natural gas problems, trade deficit... Something must give up.
Posted by: TI at December 18, 2005 08:12 AM
By the way the eurodollar futures curve has been inverted as well -- since September if I remember correctly.
Posted by: sharkbait at December 18, 2005 01:55 PM
Have you seen Brian Wesbury's post about this?
Low bond yields remain a conundrum that we can only explain by pointing to the
Fed. As we saw in the 1970s, when the Fed holds real interest rates
artificially low, long-term bond yields are also suppressed. As the Fed lifts
rates, we expect bond yields to rise across the duration spectrum, even as the
yield curve remains flat.
In the end, gold prices have been a much better indicator of Fed policy than
bond yields, and gold is signaling more inflationary pressures ahead.
So, if Wesbury is correct, the near-inversion of the yield curve is an artifact of the Fed's control of real interest rates, and we'll see this correct as higher inflation kicks in.
Posted by: Tim Lundeen at December 18, 2005 03:47 PM
"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 "but would not itself be the cause of the weakening" seems debatable. If people watch it invert, and they know this typically signals a slowdown-> might people change their behavior as the yield curve flattens? Also, how many people follow the yield curve today vs. how many followed it 30-50 years ago? Times change. For example, the internet may make it much easier and widespread to follow yield curves today.
Posted by: nate at December 18, 2005 04:24 PM
"In the end, gold prices have been a much better indicator of Fed policy than
bond yields, and gold is signaling more inflationary pressures ahead."
History (e.g., "have been") does not necessarily predict the future.
Lags are real. The Fed has done a noticeable increase (1% to 4.25%?) in a pretty quick period of time. We may have yet to see the impact of the tightening that has already occurred.
Posted by: nate at December 18, 2005 04:43 PM
Maybe, just maybe, the extended run of the 90s happened because the Fed backed off in 94.
Posted by: Lord at December 18, 2005 05:32 PM
I agree with you that the risk premium is important, and I am quite confused about what Chairman Greenspan meant. But I have a question for you: Do you know any papers documenting empirical evidence that an inverted real yield curve predicts slower growth?
The reason I ask is that first, real yield curves appear to be much flatter than nominal ones generally (UK is an example with longer history of inflation indexed bonds); and second, in many theoretical models, real yield curves are flat or inverted (I have been thinking about writing a paper about this issue for a while, but haven't found time to do it yet). In other words, for all we konw, the current real yield curve could be more �normal� than the real yield curve 18 months ago.
Posted by: pat at December 18, 2005 10:21 PM
The history of the 30-year treasury bond might figure into this. Does anyone have an overview readily available? It would be nice if this overview include a timeline for the existence and relative magnitude of the 30-year treasury bond.
Posted by: nate at December 19, 2005 07:18 AM
Pat: I didn't look at the reference in detail, but I think you might find your answers here:
Posted by: T.R. Elliott at December 19, 2005 07:57 AM
The classical perception of the risk premium is based upon the assumption that buying a long term instrument may prevent you from enjoying future interest rate increases. We have just been through a period in which many retail customers have seen their interest income fall to less than a 1% annual term. Early in this decline, many advisors suggested that the interest rate would turn up soon, so hold your funds in money markets until they do. Now, these retail investors are aware that we might have a five or ten year period with interest rates under 3% or 4%. Thus, they many me willing to lock in lower interest rates for ten years, believing that it is better to get a little income for a long time rather than none at all.
Realize that many of the retail investors with the most cash invest only in bank offered instruments, or ocassionaly a U S Treasury instrument. The inverted interest curve for this group is simply a reflection that they see risk in a manner that is inverted compared to the last half of the twentieth century. They are simply more concerned with declining rates than with increasing rates, and risk is a product of mental association.
Posted by: Bill Ellis at December 19, 2005 12:12 PM
Thanks for the reference. I am aware of the literature of nominal yield curves, but is there any research showing that real yield curves (like the ones in Jim's post) behave similarly as nominal yield curves? That is, is there evidence that inversion of the real yield curve is associated with recession or significant slowdown of the real activity? Any help is appreciated.
Posted by: pat at December 19, 2005 03:24 PM
Pat, I'm not aware of any studies using the real yield curve. Perhaps Greenspan was referring to some internal Fed research, though the relatively short sample of TIPS yields could make that difficult to sort out statistically. Or perhaps he was referring to more general evidence that real interest rates are more important than nominal rates, and he doesn't think the current short-term real rate is that high.
I think you are probably correct that the real yield slope of a year ago is more anomalous than the current real slope.
In any case, my point was not to endorse Greenspan's theory, but rather to point out that, even if he is correct, his argument still suggests a reason for concern about the current slope of the yield curve.
Posted by: JDH at December 19, 2005 03:51 PM
The yield curve for real (vs. nominal) rates was fascinating. Great work!
I went shopping today in a big city. Stores are not crowded at all. Watchout is all I have to say. It is hard for me to believe inventory is moving adequately for retailers.
Anyone who tells me that a flattening yield curve is not the cause has explaining to do. A lot of purchasing managers may decide to hold back on acquiring inventory given a flattening curve. A business mgr may delay spending, or cut-back on extra nicey niceys on purchases. Thus, the flattening may cause a change in a rational business person's behavior that induces or worsens a business slowdown.
Posted by: anonymous at December 19, 2005 09:02 PM
Professor DeLong gave Greenspan an A+ a while ago.
I give Greenspan an A (not an A+). The fed during Greenspan did not anticipate or adequately regulate the impact of the internet in the late 1990s.
Greenspan did a great job. I could not do better. We could have had a lot worse. He presided over an amazing run. However, we are not out of the woods yet, and not everything was perfect enough for an A+.
I think the existing tightening is going to have a more adverse impact than anticipated by many. I also do not see how further tightening will dampen inflation at this point. Inflation is driven by other things than interest rates.
Wesbury gets a lot of his information and analysis from the Heritage Foundation. It is not surprising to see his glasses become rose-colored during a Republican-straight.
Posted by: anonymous at December 19, 2005 09:10 PM
One more: the fed govt has not been optimal in its historical spending increases. The govt is probably never optimal, but it may have been a little worse recently. So even supply-sider, Keynesian economists should worry about increased risk due to the skewed asset allocation of the fed govt.
Posted by: anon at December 19, 2005 09:15 PM
I think the yield curve is not only made of expectations but also of investment plans. There is also supply and demand of money. The demand side is affected by longer term demand for credit. If investments are slowing down this may show as inverting yield curve by lowering the longer term rates. Here we have a clear real-economy link between recessions and the yield curve.
Posted by: TI at December 20, 2005 01:51 AM
Excellent post, and I do think the TIPS yield curve is the place to look for the inversion warning. One caveat, due to the extreme negative inflation compensation in January the carry on the 1 year TIP is horrible. The forward yield (to the end of January) of that TIP is currently 1.65% or about 35bps less then 10 year TIPS.
Posted by: Jacob at December 20, 2005 04:59 AM
Economic biology indicates that money is necessary like water in the body. We don't need economic hypertension or anemia.
How does the body regulate the amount of blood produced? Thirst.
What are the indicators of thirst in the economy?
1) lower gold prices.
2) higher commercial interest rates.
We have not seen either one and, as much as I want my home price to appreciate, I know a lot of waste is occurring with so much extra money in the economy. People are buying boats, expensive cars and plasma TV's with the perceived appreciation in their home equity.
Is that why long-term rates remain so low?
Posted by: bob at December 20, 2005 07:53 AM
Thanks for answering my question. And I agree with you that the current real yield curve doesn't support Greenspan's conclusion. It seems that the conundrum persists ...
Posted by: pat at December 20, 2005 04:29 PM
Very Interesting topic on the Yield curve....
Consider this.... With so many people owning homes without the traditional equity perecntage, the Fed cannot consider raising interest rates too far, less they send the country into total turmoil. I think that the market sees how much the average person is carrying as far as debt is concerned and therefore is holding the "Long Rates" down.
I do think we'll see this very flat curve for 3-5 years then once people start having "Real Equity" no "Paper Equity" the Yield Curve wil get its slope back.
my 2 cents....
Posted by: theShaft at December 23, 2005 04:33 PM