Monday, October 11, 2010

Why Short Duration Treasuries Are Overvalued but Long-Term Treasuries May Be a "Buy"

Just as it can be more properly argued that it is a market of stocks rather than a stock market, certain parts of the bond market can be bubbly and others be reasonable buys. An obvious example came in the wake of the Lehman-AIG market disruption, where the 30-year Treasury yield collapsed to 2.6% just as yields on sub-investment grade bonds soared. Anyone who invested bravely in the asset with the depressed price (high yield) did much better than anyone who bought Treasuries at their high prices (low yields).

This piece will present the case for investing in long-term Treasuries while at the same time believing that the 5-year and under space is significantly overpriced and therefore bubble-like in valuation.

A stock analogy to this argument came in the late 1990s. The average NYSE stock actually peaked in 1997-8, but the DJIA peaked at the end of 1999, the NASDAQ in March 2000 and the S&P 500 later in 2000. Yet many stocks bottomed in March 2000, such as homebuilders and many industrial companies. It was as if people ran from the port (tech) side of the ship to the starboard (anti-tech) side of the stock ship at the same time.

It appears that a milder version of that phenomenon could be set up to happen in bond-land.

Before reading on, you may wish to look at the nearby long-term chart of interest rates in Japan.
(Click on chart to enlarge.) I will refer to it later.

Please also consider an excerpt from a blog from 2005 by the then not-so-well-known blogger Calculated Risk, in which he commented on then-Chairman Greenspan's "conundrum" speech regarding the failure of long Treasury rates to increase as much as expected as the Fed engaged on its tightening regimen:

But I think the PRIMARY reason for Greenspan's conundrum is that the economy is weaker than it appears. Using GDP growth and unemployment, the US economy is healthy. But the level of debt (both consumer and government), the real estate "boom" that seems based on leverage and loose credit (see Volcker's recent comments), and the poor employment situation (especially the low level of participation) indicate an unhealthy economy. I believe this recovery is being built on a marshland of debt and the bond market is reflecting this weakness.

By the way, here is CR's below-consensus view nowadays:

I expected a sluggish recovery in 2010, so I thought the unemployment rate would stay elevated throughout 2010 (that was correct).

Going forward, I think the recovery will stay sluggish and choppy for some time and I'd guess the unemployment rate will tick up in the short term and still be above 9% later next year.

I more or less agree with CR though I may be a bit more bearish than he.

So the predicate for the following discussion is the view that while economic jiggles upward are to be expected, too many data points simply point to a general stagnant trend for the pace of business in this country for me to argue against accepting that stagnant trend as the New Normal.

Despite that New Normal and the fact that 0.5% interest rates on 3-year money are crisis lows and reflect a poor state of business affairs, let's consider how extreme the market's estimation of fair value for interest rates in the "out years" has become. As first Mr. Greenspan and then Dr. Bernanke led the Fed's long slow interest rate-raising campaign, the "conundrum" of a flattening yield curve caught their eye. At the interest rate peak in mid-2007, more or less all rates from 1 day to 30 years were identical at around 5.25%.

For historical purposes, please be aware that through much of the 1800s, short-term interest rates were generally higher than long-term rates. This reflected such factors as productivity gains, generally under a gold or bimetallic monetary system. As the Japanese experience the past decade reflects, though, prices may fail to rise even under a non-metal-based (i.e. "fiat") monetary system.

What is the message of Mr. Bond today?

Well, the 2-year bond is at 0.35% yearly, the 3-year at 0.52% and the 5-year at 1.10%. Using a simplified model of implied bond yields farther out the curve, one can calculate as follows.

The 30-year Treasury bond yields 3.75% each year for 30 years. Thus a $100 investment in this bond at "par" of $100 per bond provides a gross total of $112.50 in interest payments over 30 years.

If one were instead to purchase a 3-year Treasury note, one would receive about $1.50 gross over 3 years for every $100 invested (lent to the government), or a "grand" total of $4.50 over the 3 years. It hardly seems worth bothering (especially when credit card companies that have bank subsidiaries are offering FDIC-insured yields of 1.3-1.5% even today). Subtracting interest income over 3 years from that obtainable for 30 years gives income attributable to years 4-30, In those final 27 years, one would receive $111, spread evenly over each year. This computes to about 4.1% yearly.

Turning to a comparison of the first 5 years of the yield curve vs. the final 25 of a 30 year stretch, under current rates one accepts 1.1% a year for 5 years. Similar math to the above would imply a 4.28% yield yearly for the terminal 25 years.

Let's look at matters differently. To choose 5 year paper over 30 year has one certainty. At the end of 5 years, the investor of $100 in a 5-year note will have earned $5.50. He/she will have $105.50. The investor in the 30-year bond will have earned almost $18, or about $12 more. The first investor will be more than $12 behind the proverbial eight-ball (to mix a numerical metaphor). It will not be so easy to find the right investment for the next 25 years just to come out even with the buy-and-hold investor in the 30-year bond.

The numbers get more interesting as we go farther out on the curve. If the 20-year Treasury yields 3.2% today, then it pays $64 in interest per $100 over that 20 years. Subtracting that from the total payout to the 30-year holder implies an average yield for years 21-30 (the final 10 years of the 30-year bond) of 5.1%.

Voila! We are nearing the rates extant in 2007.

A back-of-the-envelope guesstimate is that the bond market is implying a 5.5% yield is proper for year 30 on its own.

In other words, the long bond is in no bubble whatsoever. It is the low-yielding front years that are wildly overvalued (under-yielding).

Well, all this guessing about the "right" interest rate in the out years is absolute conjecture on the part of the bond market. The truth is that the bond market has little more idea than you or I what economic conditions will be like a year or two from now, much less many years out. Japan scenario? Greece (though with a printing press)? U. S. 1970s, with bear markets for both stocks and bonds? Everything good?

The truth is that we really don't know what will happen from one day to the next. One summer day several years ago, I left home and about 10 minutes later walked into the doctors' lounge to grab some coffee before seeing my hospital patients. The TV showed that an accident had occurred in downtown New York; an airplane had just hit one of the World Trade Centers. Since that time, the Fed has almost never voluntarily pushed the Fed funds rate above CPI, in contrast to its policy of most of the prior 2 decades; and the U. S. has been on a semi-war footing even despite the election of a "peace" candidate in 2008.

The world changed during my brief commute to work. Two-three years from now could be another eternity, just as the past few years has been.

The current trend is the Japanese one. Zero interest rate policy ("ZIRP") is resorted to as an emergency measure by the central bank. The acute emergency ends, but chronic illness emerges. Zero or near-zero interest rates continue to anchor longer and longer maturities toward zero. Meanwhile, market participants continue to rationally expect a reversion to the interest rate mean. This has been the Japanese experience, and to date nothing other than hope for better times has occurred in the U. S. to differentiate us from them.

Let us look again at the chart of interest rates in Japan. After the yield on the 5-year note fell to new lows in 2001-3, the 30-year actually fell more in yield than the 10-year. Whenever the yield gap, in absolute and percentage terms, was wide, it paid to buy the 30 year.

Why should it be different here and now? Because we have nukes and they don't?

In addition, there is a specific potential catalyst for outperformance of the long bond. That is the expected "QE2", or further expansion of the Fed's balance sheet, a/k/a yet more money-printing, with the now unconcealed primary purpose (per the NY Fed's executives in recent speech(es)) of keeping financial asset prices above fair value. (In other words, trickle-down economics . . .)

It is known that the public has been rebalancing its portfolio from stocks toward bonds, but the bonds it has been buying are almost entirely short-to-intermediate term securities, or so it has been reported. Thirty years ago, bonds were known as certificates of confiscation. In my humble opinion, that's likely what 5-year notes are today. Apparently the public is nervous about the long term future for interest rates and is so nervous it is moving heavily toward the short end. Where the public is nervous, it's usually a good idea to think about being a contrarian.

Getting back to the potential catalyst for a rapid and large drop in long-term rates, what if Dr. Bernanke took a quick look at the interest rate curve, saw the record or near-record absolute and relative spread between the 2-year and the 30-year bond, and between the 10-year and the 30-year bond, and said the following any time in the near future?

"The yield curve is too steep. The Fed is going to target the 10-year yield at modestly below its current level but believes that a more appropriate level for the 30-year bond is at most 3%, which provides a positive return after targeted inflation of 2% with utmost security of principal and is thus a yield that is fair to both the lender and to the Treasury."

There would be a rush to the long bond, of course. Now here is where the arithmetic gets interesting. A drop in interest rates of 75 basis points on a zero coupon bond for a 30-year security from 4% to 3.25% (rates on zeros are higher than rates on par bonds that pay interest) gives a price change of the zero coupon security from $30.80 to $38.30; quite a percentage move.

Let us take the Japanese experience. If 5 years from now, a zero coupon security purchased yielding 4% turns into a bond (at that time a 25-year bond) yielding 2.5%, the price will be $54 (from $30.80, to remind you).

If at any time within those 5 years the yield drops that much, the annualized return will be greater, even if the price is less than $54.

Aside from Fed action, why might the long Treasury be a fundamentally attractive investment for individuals, especially for tax-deferred accounts such as IRAs?

Might the U. S. actually "walk the walk" of fiscal prudence? Might the authorities, after all the fruitless manipulation of monetary aggregates, get back to where they once belonged and accept the obvious principle that aiming for price increases is harmful to the people they are supposed to serve, and primarily only benefits the financial class?

Might some form of hard money serve as the chosen solution to ratify most of the unsupportable promises known as debt and social obligations the Feds have willing taken on?

Might the government actually go to a peace economy?

In any case, back to supply and demand, and to fundamentals of return adjusted for changes in the general price level ("real" return). Here is a link to a Morgan Stanley estimate of outstanding Treasuries by duration. There are not a lot of long-duration Treasuries. The yield spread is large. A buyer of a 5-year Treasury at 1.1% could get creamed simply by an average of 3.5% price inflation per year, whereas the buyer of the long Treasury would (pre-tax) be holding even and the buyer of the zero coupon bond would be accruing perhaps, after all is said and done, a small positive return yearly.

Thus, in conclusion, there are potential capital gains that might accrue soon to the purchaser of long-term Treasury bonds, with said possible gains being leveraged via purchase of zero-coupon instruments; and there is greater protection from inflation over the next few years by purchases of those types of bonds rather than the short-intermediate duration vehicles to which J. and J. Q. Public have reportedly been flocking.

Some caveats:

1. Anyone reading this who is not a regular reader of my posts would not be aware that I have spent weeks and many months criticizing the monetary authorities for unfair and inappropriate printing of mass quantities of money. I have published a series of on-line articles describing various ways in which I have been investing significant portions of my financial assets to try to protect them from the ravages of all this money-printing; principal among them is gold (I have blogged favorably about gold since winter-spring of last year). Here is a link to the first of that series, on September 8

2. For taxable accounts, well-chosen tax-exempt municipal bonds may provide the optimal risk-reward for buy-and-hold bond investors rather than Treasuries. This discussion of Treasuries unfortunately is in the context of my post titled We Are All Speculators Now, from September 20.

3. Investing is risky, speculating is risky, and purchase of long-term bonds entails a commitment to, well, the long term, even if a profitable sale of the bond is hoped for in the shorter term. The potential for gain entails meaningful potential for loss, especially adjusted for possible sustained rapid rises in living costs.

4. Nothing herein or in any of my on-line posts should be construed as investment advice to any person, as opposed to simple commentary and conjecture. Reasonable efforts have been made to present economic and numerical data as accurately as possible, but please take nothing for granted and do your own math and research as appropriate to your level of interest.

Copyright (C) Long Lake LLC 2010

No comments:

Post a Comment