Yield Curve Inversions

What did yield curves look like prior to WWII when inflationary expectations in the U.S. were essentially zero? Were they as flat as we see today? Flatter perhaps? How often did they invert? I've been trying to find historical yield curves for the U.S. going back prior to the post WWII time period when inflation became a persistent feature of the economic landscape. I haven't had much success, partly because it is not possible to extend the 3 month T-Bill and the 10 year T-Bond rates back that far since the 3 month T-Bill did not exist until the 1930s and the long-term rates are spotty and unreliable. For more on this, see Michael D. Bordo Joseph G. Haubrich who use data from  Balke and Gordon to construct historical yield curves for the U.S. Unfortunately, the commercial paper and corporate bond measures they use have different risk characteristics so that the difference in the two yields will reflect this difference in risk. Here is the graph from their article:

Not as informative as I'd hoped for the questions I want to answer, but it does appear that yield curve inversions were more common prior to WWII when prices in the U.S. were stable.

Better data exist for Norway. This is from a Norges Bank publication and appears to be carefully done. These two graphs look at yields on 2-4 year and 20-60 year bonds. Construction of the two series is described in the article:

I don't know that much about historical Norwegian data, but it is evident from these graphs that prior to WWII yield curve inversions were much more common and could persist for a period of years (an inversion is when the blue line is higher than the red line). Here is 500 years of Norwegian price data to get some idea of inflation rates during the 1921-1979 period and the periods before and after:

Not sure what to make of this. In the part of the graph representing the sample period 1921-1979, prices rise, then fall, then rise again and continue rising through the end of the sample. I couldn't find any clear association between the recession periods discussed in the source for the price graph and yield curve inversions, but the comparison was not done with a careful statistical analysis. In any case, after seeing these graphs, and given the Fed's strong commitment to price stability, the possibility of an inverted yield curve does not seem as ominous.

Comments

[Hedgehog]

I wonder what kind of shocks are behind those ladder-like steps in long-term rates on Figure 8. Is their timing related to oil price shocks? Any specialists on Norwegian fixed income market on this forum??

It looks like a very interesting stuff, but it's really hard to interpret indeed. It's always important to keep in mind that the instruments that are used for reconstruction of yield curves across very long histories are likely to be time-varying.

As for the bizarre pattern of Norwegian inflation implied by the last graph, the best way to model it (if I am _really_ hard-pressed to do it) would probably be a regime-switching process with occasional jumps. I am curious if anyone has better ideas!

[anne]

Clever analysis. There appears to have been another stage in the movement to more economic flexibility that Alan Greenspan gained a sense of 15 years ago or more. The American economy has become increasingly flexible since 1980, and the flexibility has evidently increased since 1995 as international capital flows responded more smoothly to American investment opportunities. I am not worried about an inverted yield curve for much of the favorable level of long term interest rates is accounted for by international capital flows.

[anne]

http://www.nytimes.com/2005/12/06/business/06cnd-econ.html

December 6, 2005

Productivity Rise Is Fastest in Two Years
By VIKAS BAJAJ

Productivity rose at its fastest pace in two years in the third quarter, far more quickly than earlier predicted, as output rose and labor costs fell, the government reported today.

As a measure of how much the economy produced per hour of work, productivity rose by 4.7 percent in the non-farm business sector of the economy from July to September, compared with an earlier reading of 4.1 percent, the Labor Department reported. Real hourly compensation, which adjusts wages and other benefits for inflation, fell 1.4 percent, unchanged from previous estimates.

The report indicates that the productivity boom of the last several years may have more steam left in it than Alan Greenspan, the Federal Reserve chairman, and other economists had believed. Typically, productivity tends to slow in the latter parts of an economic expansion because businesses have wrung out most of the efficiencies from their operations and have to compete more aggressively for a thinning supply of employees.

For workers, however, the data shows that the rise in energy costs wiped away any advantage they received in the form of higher wages, at least for a time. Before adjusting for inflation, hourly compensation rose 3.7 percent.

Unit labor costs, which gauge how much compensation it takes to produce one unit of output, fell 1 percent in the quarter, twice as much as previously expected.

From 2000 to 2004, productivity gains averaged 3.28 percent a year, far higher than the average of 2.14 percent for the last 45 years. Those gains are one of the mains reasons cited by Mr. Greenspan and other policy makers for the ability of the United States economy to achieve long periods of growth in recent years without sparking inflation....

[Lord]

Isn't there some good historical British data on the subject? I believe it shows numerous short-lived inversions in the 19th century when long rates were low due to the desirability of stable returns.

[Bruce Wilder]

Inflationary expectations were zero? Expectations, smectations. During parts of the period between the Civil War and WWII, there was actual deflation. Obviously, deflation was particularly a nasty problem, 1929-32, but also 1872-1896.

During the period, 1872 - 1896, the U.S. was actually experiencing a prolonged deflation, at an average rate of close to 1% per year. This was a consequence of being on the gold standard at a time, when gold production was lagging behind a rapid economic expansion in the industrial economies of the U.K., the U.S., France and Germany, all of which were trying to tie their currencies to gold.

Some positive rate of inflation has been a certainty throughout the whole post WWII period. Increases in gold production after 1896 lubricated the economic expansion, which continued into World War I, with a modest inflation.

An inversion of the yield curve in a period of deflation is almost a redundancy, since currency, itself, has a positive rate of return. An inverted yield curve, in inflationary times, termporarily chokes off intermediation, but, in a period of deflation, intermediation is already ingesting a deadly poison, so what's the point? An inverted yield curve would be choking a dead man walking.

The general deflation of 1872-1896, predictably, resulted in short expansions and frequent or prolonged contractions. The post-WWII acceptance of a modest, positive rate of inflation as an ideal has resulted in prolonged expansions; as long as actual deflation is banned as a matter of policy, yield curve inversions and runaway (accelerating) inflation are the two monetary mistakes left on the table. Runaway inflation is not going to result in a yield curve inversion, and prolonged expansions under positive modest rates of inflation are going to make yield curve inversion infrequent. (That pretty much follows from the definition of "prolonged".)

The only long-term, non-sovereign debt on offer in a deflation is, by definition, risky. The real return on such debt is the nominal return, padded by the rate of deflation. In effect, the deflationary padding is a forced risk-premium. All borrowing carries this risk-premium and becomes, therefore, risky; lowest-risk investments cannot be made and all investments, which are made, carry the additional risk, imposed by deflation.

[calmo]

Well the 500yr Norwegian history helps me about as much as the parrot.
It does not speak to me. (Yes, time for a hearing aid and a brain scan, the works.)
Could be I'm suffering from that rare disease, EnRoachment where one discounts immediately any isolated economy that has any semblance of balance, integrity, responsibility and straight shootinness.

CR notes some bond news that I could not handle either: are lenders going to push out those long rates?
Me and the parrot are having a bad day.

[sharkbait]

The almost inverted curve does seem ominous to me in the sense that the market expects less inflation and possibly even deflation.

Kessel, who looked at business cycles between 1858 and
1961, finds a pro-cyclical pattern, with short term rates rising relative to long term rates
in expansions and falling in contractions; as the short rates often peak at the same time as
the business cycle, this suggest predictability.

http://scholar.google.com/url?sa=U&q=http://www.nber.org/~confer/2003/mef03/bordo.pdf

[Bruce Wilder]

People, I guess, will never give up the meterological view. Rational expectations was more than a front for the execrable (new?) classical view, but, sometimes, one would hardly know it. There's value in seeing the control model as the first order effect, and expecations as a second order phenomenon. The economy is not the weather; it may be maddeningly complex, but it is still almost all a product of intentional human activity, even if the intentions sometimes get lost in conflict and complexity.

A yield curve inversion is a monetary policy mistake. It is not just a leading indicator, an esoteric predictor used in a rule of thumb. It is a mistake: something, under the control of competent authority, which should not be allowed to happen. Deflation is a mistake; rapid, accelerating inflation is a mistake; an inverted yield curve is a mistake. The financial system is an intricate control mechanism, and an inverted yield curve screws up its functioning. It is the financial system tripping on a rough patch in the road; the financial system falling on its face is frequently the consequence; go figure.

Worrying about what an inverted yield curve says about the "expectations" of a mass of investors is kind of silly, when you might just as well worry about what it says about the competence of an handful of institutions, which have as much responsibility and control of the interest rates on 30 year mortgages and 10 year Treasuries as they do over the Federal Funds rate or 3-month paper.