Last year at this time, financial analysts were puzzled by the fact that Federal Reserve Board Chairman Alan Greenspan’s tightening of monetary policy (since mid-2004) had done little to slow the economy.  By June of last year, the central bank had effectively raised the yields on one-month Treasury bills by 2 percentage points.  During this same period long term interest rate yields declined by about 1 percentage point as did mortgage rates. Chairman Greenspan termed the declines in long term interest rates a “conundrum” when he testified before the congressional Joint Economic Committee. The conundrum would have been solved for those analysts if they had just waited for the same time, next year.¹

Empirical evidence tells us that the primary determinant of long term interest rates is anticipated inflation. This was particularly true after the 1960’s when the concept of bond management came of age.²  Treasury bond yields are most closely correlated with core consumer price inflation, i.e. excluding food and energy costs. This measure of inflation was introduced in the early 1970’s by Otto Eckstein and began being reported by the Bureau of Labor Statistics in 1978.³ Chart 1 shows the relationship between the twelve month moving average of yields of the 10-year Treasury Note (T-Note) and the change of the core inflation index. Our analysis suggests that bond yields correlate best with the core inflation change in seven months. (This makes sense because bond traders try to anticipate future inflation trends.) Chart 2 depicts the ratio between the T-note yield and the inflation rate. Over the last quarter bond yields have averaged 2.2 times the core inflation trend; however, this ratio rarely is at the average. Most valuations are either at the high or low end.

Given the current core inflation trend of 2.1% one might expect yield to be 3.85% in “good” markets or 5.45% in “bad” markets.  This results in quite a spread in prices.  It’s clear that there are other forces at work here as well. U.S. Treasury bonds have been a favorite investment for foreign central banks that need to invest their massive trade surpluses into our Treasury market.  This is particularly true for those countries where interest rates are lower than in the United States such as Japan, the largest foreign holder (30%) of our Treasury debt. Chart 3 shows the yield on our 10 year Treasury note versus the level of the Nikkei stock average.⁴ As one can see there is a strong correlation between the index and our T-Note yield.

For ten years now Japanese interest rates have been near zero in an attempt to avoid a deflationary collapse.  This created what is now called the “yen carry trade”. Large investors (mostly hedge funds) could borrow yen at very low interest rates and invest those funds in higher yielding instruments such as U.S. Treasuries (usually leveraged 3:1). As the Bank of Japan raises rates for the first time in a decade and stock investors buy into Japan’s revival, the yen has been rising.  The Nikkei stock index is symptomatic of the health of the Japanese economy, and thus, in an indirect way, is linked to U.S. Treasury bonds.

Over the last three weeks the yen has risen by almost 7%. This is bad news for the yen carry trade. An unwinding will mean that hedge funds and other foreign private entities will sell Treasuries to close their positions. We believe this is occurring now, and possibly, like the Long-Term Capital Management fiasco of 1998, they will create a dangerous environment for holders all types of liquid assets.⁵  

FAIR VALUE

If one were to use core inflation as a guide for valuing what the 10-year Treasury note should presently yield, it wouldn’t have much value due to the wide range of multiples for “good” and “bad” markets.  But sentiment toward Treasuries is changing.  As recently as last September, the Market Vane index of Treasury bond “bulls” was at 74. As of last week it was 43!  So it may be that the T-Note will trade at the high multiple. To further improve our chances of being correct, we employ a regression model that presumes no increase in the current underlying inflation rate and uses the one year Treasury rate as a proxy for Fed policy. In this instance the model tells us that a fair value yield would be 6.0%. (This implies a price decline of 14.5%)  If we use the core inflation rate and the Nikkei average as variables, fair value would be 5.6 %. However, over the last six months, core inflation has been running at an annualized rate of 2.7%, not the 2.1% in the model.  If that inflation trend was to continue, our current fair value rate projections would be 6.3% and 5.9% respectively.

IMPLICATIONS FOR INVESTORS

As long term interest rates continue to migrate upward, investors holding long term bonds will suffer. (Remember, when the market yield goes up, market price goes down.)  The last six months have not been fun for long maturity bondholders (unless they were denominated in a strong foreign currency). There will be some damage as well to those equity sectors that are bond surrogates such as utilities and real estate investment trusts.  Both groups have been excellent investments over the last few years however both are overvalued at this point and are facing a rising rate headwind.  The real surprise might be the impact of long term rates on the non-financial corporate economy. Most of the press pay attention to short term interest rates and write endlessly about the “yield curve”, i.e. the relationship of short-term to long-term interest rates.

Prior to 1986, there was an interest rate ceiling on demand deposits known as Regulation Q.⁶   For most of the post-WWII years, an inverted yield curve (when short term interest rates were greater than long term rates) usually produced an economic slowdown. Since 1986, however, the yield curve has not been a good precursor of slowdowns. Chart 4 exhibits the on-again off-again relationship to non-financial corporate business output.⁷ (In this chart and in others, the yield data tends to lead economic activity by six quarters.)

A better lead indicator of economic activity is the percent change of long term interest rates.  In this instance, corporate bond yields are more highly correlated to economic activity than Treasury bonds. We use the yield of bonds designated by Moody’s as Baa (the minimum for an “investment grade” bond). Chart 5 shows this relationship.

A regression model that combines both the current yield curve and the rate of change in Baa bond yields, (see Chart 6) forecasts slowing output for our domestic economy through September of 2007.  The silver lining to this decline, if it comes out that way, is that the Federal Reserve would be inclined to begin lowering rates somewhere around year end 2006.  To do this they will need to have concrete evidence that the economy is faltering under the weight of their previous moves. This would be similar to the credit easings of June, 1984, November, 1994, and January, 2000.

The not-so-silver lining is that if bond yields rally further than expected, and output does in fact slow to a 2% growth rate from the current 7%, profit margins will most certainly decline. A review of Chart 6 shows that margins usually decline with a slowdown in output.  Mind you, we are not projecting a recession. Moreover, when interest rates rose in 1994, the output decline was very muted and profit margins actually improved. Furthermore, the decline in the dollar will benefit the earnings of multinational companies.  Since the market anticipates slowing profits by two quarters or so, it would not surprise us if the second half of 2006 was less rewarding than the last six quarters.  But just wait until the same time, next year.

CONCLUSIONS

The rise in longer term interest rates has further to go based on an analysis of the current inflation rate, Federal Reserve policy, and the rise of the yen.  This does not bode well for a number of investment categories.  The recent rise in the yen suggests that there is some possibility of a hedge-fund related financial accident that could cause some other asset categories to be liquidated in order to “close out” or unwind yen-Treasury bond pairings. The increase in bond yields over the last six months might possibly cause a slowing of domestic non-financial corporate output growth in 2007 to levels that would be associated with a decline in after-tax profit margins. This should put a damper on the stock market during the second half of 2006.

John K. Dolan                                                                                      S&P 500: 1291
Treas. Note (10 Yr): 5.14%

May 15, 2006