Here’s an investment mind bender (or mind blower if you happen to have grown up in the late `60s): A “reflating” economy is ultimately bad news for a bond investor. Still, the “ re” portion of the word is initially invigorating for bond prices and total returns since it implies a central bank on a downward interest rate path. Greenspan’smarch from 6½% at the beginning of 2001 to what is probably an ultimate low of 1% today epitomizes the feel good sizzle of “ re”-flation. Since that date, the bond market has produced a total return of over 26% — more than 8% annualized — with higher prices contributing 6% of the total. Once a central bank reaches its ultimate destination, however, (and who could argue that 1% is not far from absolute 0?), a bondholder is stuck with the worst of all worlds. Not only is she earning a low “real” interest rate (negative, as a matter of fact in today’s market), but she has to look forward to the inevitable day in which rates go back up and prices back down. She is damned with a 1% money market fund and damned with a 4.8% 30-year Treasury that almost inevitably will generate capital losses, as the re-“ flation” steers long yields higher. The 1% money market rate, as a matter of fact, along with a 1% or so risk premium for holding longer dated bonds is a pretty good approximation for what an investor in U.S. bonds should earn over the next 4-5 years . Staring 2% returns in the face, what is a bond investor to do?
My Investment Outlooks of February and March (The Last Vigilante I and II) sought to answer that question in part. The two IOs separately discussed the benefits of owning inflationary insurance via TIPS as well as employing the front end of reflating country yield curves to participate in the rolldown/carry made possible by the magnet of low short rates themselves. This month’s discussion will center on the durational position of a “reflationary” bond portfolio that might be characterized as thrice damned — damned if you shorten, damned if you lengthen, damned if you don’t do a thing. The solution to this conundrum is to be rather choosy about the country where you hold your duration, to reduce it in reflating countries (U.S., Japan) and increase it in relatively vigilant ones (Euroland, U.K.). In so doing, the potential pain from holding bonds in a reflating global environment can be reduced if not anesthetized entirely. Not only are some offshore yields currently higher than U.S. alternatives to begin with, but future price erosion should be minimized as well since reflation is not as advanced. German 10-year Bunds for instance yield 3.92% vs a 3.85% U.S. alternative. U.K. 10-year Gilts yield 4.75%, almost 100 basis points more than Treasuries. In short, I recommend getting the hell out of Dodge City, USA and reinvesting in London and Frankfurt. Construct an A B T portfolio — Anything But Treasuries, and hand those Old Maid Treasury bonds to the Japanese and the Chinese or any other country whose domestic growth concerns dominate their investment common sense.
That a U.S. Treasury note or bond resembles the Old Maid is not immediately obvious. With domestic inflation at 1½ - 2% and 10-year Treasuries close to 4%, a bond investor is not exactly making hay but things could be worse. Well maybe — but as it turns out — only in Japan where inflation is 0% and 10-year yields are 1.7%. Elsewhere, the hay makin’ is considerably more attractive as I will attempt to point out over the next few pages. My first set of rather commonsensical proofs are displayed below. Bond investors know, but oftentimes lose sight of the axiom, that a country’s interest rates should be correlated rather closely to its nominal GDP growth.
While that relationship varies based upon periods of “easy” versus “tight” money, the correlation has an economic logic and an historical validity. The two countries shown in Exhibit I’s 25-year history display 10-year government yields averaging 150 to 250 basis points over their respective nominal GDP growth rates. I would be the first to admit that the above relationship was derived during a disinflationary (high real short rate) interest rate environment and that Germany’s incorporation of East Germany in 1991 distorts their GDP numbers to the downside. Still, the comparison between the U.S. and Germany is stark. My, oh my, what a difference! Today’s German/Euroland yields remain safely above nominal GDP growth rates as measured by the difference between the red line and the blue baseline, which represents nominal GDP. Not only that, but 10-year yields presently are 32 basis points above 25-year averages as represented by the blue line. The U.S. on the other hand is a mirror image. Its 10-year Notes yield nearly 200 basis points less than nominal GDP and 356 basis points less than historical average levels. Granted, Germany’s economy is now displaying higher growth rates which will narrow the comparison over the next 12 months, but German yields currently offer 378 basis points or 3.78 % more “yield to GDP” than its U.S. bond market competition. Which country would you rather own
Here’s another mind blower. Exhibit II displays a comparison of 10-year government yields as a function of their “current account balance as a % of GDP.” The straight line is a PIMCO drawn “best-fit” regression line, which defines the statistical correlation between the two as of year-end 2003. The logic here is that a country running an annual “deficit” with its trading partners should be required to pay higher interest rates relative to those that run a surplus. While geopolitical and currency reserve status considerations ultimately affect the number, the correlation is a good one, not just for 2003 but for the past several decades as well.
If so, the attractiveness of a country’s bonds can be viewed by the “distance” from the best-fit line. Germany, as represented by “E.U.” is smack on top of where it should be given its mild current account surplus. The U.S. on the other hand trades at yields approximately 100 basis points less than where the model suggests they should. For those wanting to play the reserve status/geopolitical “get out of jail free” card, let me point out that U.S. yields as of March 2004 were at near historically negative spreads using this methodology as shown in Exhibit II’s right hand chart. There seems little doubt to me that Japanese and Chinese buying of Treasuries, combined with the Fed’s historically low interest rates which pay no respect to currency support mandates normally required of a current account debtor nation, are the key culprits. Should either of these conditions change, U.S. yields move higher. In Germany and the U.K. on the other hand, value is much more apparent and future bear markets less likely.
One additional point before I wrap up this “Bond Vigilante” trilogy, which has elaborated on our best risk/reward strategies in a reflationary world. February’s Outlook ended with a cautionary note that reflation creates asset bubbles which if popped may ultimately lead to a deflationary bust with lower not higher Treasury yields. Talk about a mind bender. Although that is not a high probability in PIMCO’s play book, the possibility is real enough to have argued for durations close to market averages until the outlook becomes a little more clear. Shifting durations offshore offers some defensive protection here without altering the actual duration target.
In addition, it’s instructive to summarize just how overvalued U.S. Treasuries are to put the proper upside or downside of this outsourcing strategy into perspective. How much do you the client stand to gain by selling Treasuries and buying Bunds or Gilts? The first two exhibits hint at anywhere from a 100 basis point to a 150 basis point overvaluation of Treasuries relative to historical parameters which if rectified is equivalent to a relative loss of 7-10% in price terms. To that should be added the following observation derived from the near century long statistics of Ibbotson Associates ( Stocks, Bonds, Bills, and Inflation) and Dimson, Marsh & Staunton ( Triumph of the Optimists). Over the past 100 years, which included periods of deflation, double-digit inflation, and government controlled interest rates, the arithmetic average long-term real interest in the U.S. approximated 2.9%. Today’s 30-year real rate offered in the TIPS market stands at 1.9%. The difference of 100 basis points complements relative analysis discussed above without delving into the complicated guessing game of inflation and inflationary expectations. Whether this overvaluation eventually leads to 10-year Treasury yields of 5% depends importantly on three events: 1) The Fed moving to a more economy-neutral Fed funds rate, 2) The Chinese unpegging their currency to the dollar, and 3) The Japanese ceasing their “dirty float” suppression of the Yen via massive Treasury purchases. As of now, only the third condition has some possibility of immediate market moving impact. Perhaps Friday’s employment report will speak to the first. Until Treasuries adjust, however, bond vigilantes and investors alike are better off following an ABT strategy by reducing Treasuries and investing on foreign shores to get their desired durational exposure.
William H. Gross