Leaving PIMCO.com

You are now leaving the PIMCO website.

Skip to Main Content
Economic and Market Commentary

Tom Hanks - Portfolio Manager

What I experienced on my 'island' was a lesson in financial history that could pay huge 'dividends' in future years.

H"istory doesn’t repeat, but it rhymes."
Mark Twain

There are a number of ways to skin a cat or analyze the fortunes of the bond market. Fed watching, economy and inflation forecasting, supply and demand analyzing - take your pick, mix and match ’em, above all be prepared to be humbled. If it were easy, I might not be writing Investment Outlooks for a living. (Surely a joke.) The older I get however, the more dependent I become on history. "Older, but wiser" goes the conciliatory saying with its presumption that age can lead to an understanding of sorts unavailable to the more youthful and less experienced. Those of us pressing 60 and beyond certainly hope so - it’s one of the few pegs we have left to hang our hats on. "I was here first" doesn’t cut it when you’re trying to outperform the bond market with a $350 billion portfolio. Perhaps age with its inherent appreciation for history does, however. If you can’t out analyze ’em or out Fed watch ’em then throw the history book at ‘em. Hopefully, as Mark Twain suggested, there will be a rhyme or two that leads to something that the rest of the pack has failed to pick up on.

This nouveau fascination with history actually began way back in my youth. Thomas Bailey’s The American Pageant was sort of my high school Bible - it still sits prominently on my library bookshelf. Later, Paul Johnson’s Modern Times and A History of the American People consumed hours and hours of personal reading and reflection. "They were us - we are them…we leave almost identical footprints in the sand," was the rhyme I heard more than anything else when reading them. And so it was only natural, I suppose, with such a heritage and completing my sixth decade and all, that I should turn to financial history in an attempt to outskin my feline bond market competitors. Now, there are two coffee-table sized books that sit prominently on the right side of my library desk - Triumph of the Optimists and the 2003 Yearbook of Ibbotson Associates’ Stocks, Bonds, Bills, and Inflation - where before there were none. Now, I turn to their historical statistics for bond market wisdom where before I would consume a myriad of Wall Street talking pieces. I’m placing more of my bets these days on the rhyme instead of the cacophonic noise. Let’s hope it works.

These two books are voluminous. I tell clients that one could be stranded on a desert island like Tom Hanks in "Castaway" and never finish appreciating all the information that lies inside. You want to know the long-term winner of the growth stock versus value horse race? Page 157 of the Ibbotson Yearbook will tell you. Do you want to know returns on South African bonds for the 20th century, decade by decade? Turn to page 281 of Triumph of the Optimists. Still, most of you wouldn’t go that far, even if you were stranded on a desert island. You’d start up a conversation with a volleyball named Wilson instead. So let me summarize a few of the highlights, a selective history of bonds that might make the most difference as we wind our way through the next 12 months or even the next 12 years.

First of all, as readers of my Outlooks discussing TIPS and real interest rates will remember, I find it fascinating that investors and economists believe that the real interest rate experience of the last two decades of the 20th century should be the norm for the first twenty years of the 21st. The "Outlook 2004" edition of the highly respected Bank Credit Analyst, for instance, states that the "equilibrium level for the fed funds rate is between 4 and 5%, so there is a long way to go before policy becomes restrictive." Not so, I would claim, especially given the history of real rates from Triumph of the Optimists shown on the following page.

The graph is a bar chart showing the real interest rates pre- and post-1980 (through 2000) for the United States and 15 other countries, plus the averages for all. From 1980 to 2000, average real interest rates average 3.7% for all countries, compared with negative 0.7% before 1980.  Denmark has the highest real interest rates of 7.2% per year from 1980 to 2000, and 1.7% before 1980. The United States has a rate of 2.8% post 1980, and 0.4% before that date, ranking it roughly in the middle. Italy shows the lowest real interest rates for both periods: 1.1% from 1980 to 2000, and negative 5.4% before 1980.

The fact is that 4-5% equilibrium short rates which in today’s inflationary environment equate to 2-3% real rates shown in the chart, were a product of disinflationary policies begun in 1979 and were meant by and large to be restrictive, to bring inflation down presumably at the expense of growth. But the first 80 years of the century experienced average real short rates of .4% in the U.S., .1% in the U.K., and negative in many Euroland countries. This history tells me to expect a long stretch of close to 0% real interest rates in the U.S. and most G-10 countries, especially since reflation is now the stated goal of two of three of the world’s most important central banks - the U.S. and Japan. One of the most important conclusions to be drawn from this history lesson, as outlined in last month’s Outlook, is that bonds (and stocks too) will be low return asset classes for the foreseeable future. That is so because the market’s interest rate North Star, the short-term yield which guides and steers buyers and sellers through carry and arbitrage activities, will be close to 0% real - if history rhymes .

Investors desiring something more than 0% after inflation from their bond investments will be comforted by what Ibbotson labels the "horizon premium" and what others might call the "yield curve risk premium." The chart on the following page displays a 77-year history of the annual returns of long-term Treasury bonds versus 30-day Treasury Bills.

The figure is a bar chart showing the history of the annual returns of long-term U.S. Treasury bonds versus 30-day Treasury bills, from 1925 to 2002. The chart indicates the arithmetic average over the period is 1.6%. In 2002, returns for bonds outpaced T-bills by about 14%. Since the early 1980s, bond returns have outpaced those of T-bills in all years but six of them, and the bonds’ gains are greater than the gains of T-bills during the years they have better returns.

Although the yearly numbers are obviously volatile due to the direction and price change of the long bond, the historical annual outperformance of 1.6% for the long bond has to be instructive. First of all it alerts a bond investor to the risks and rewards of "horizon" or maturity extension. It says in any given year you should expect 1.6% more from owning 30-year bonds than 30-day bills, but to expect a Wild Toad’s ride for the advantage. Secondly, it almost screams that today’s 30-year TIPS, which provides a real yield of 2.4%, is still a bargain by historical standards. Since 30-day bills have averaged approximately .5% real and long Treasuries carry a premium of 1.6% to that, then a long maturity TIPS should yield 2.1%, all else equal. Since it still yields more, and because today’s reflationary environment should afford an insurance yield discount to the TIPS as opposed to the nominal 30-year bond, history says tilt your Treasury duration in the direction of inflation protected securities. We have been.

The figure is line graph and bar chart showing default premia, the U.S. corporate bonds index, and the government bond index, from 1900 to 2000. The corporate bond and government indices, scaled on the right-hand side, are both near or at a peak on the chart in 2000. The corporate bond index is around 8.2, up from about 2 in the early 1980s, using an index value of 1.0 in 1900. Government bonds, reach about 5.0 by 2000, up from a low of about 1.0 in the early 1980s, also using a value of 1.0 in 1900. The bars show a downward trend in default premia since the early 1980s, with it going negative in 2000, to negative 7% or so, its lowest value on the chart.

Financial history’s next lesson concerns corporate bonds and the risk versus return of owning them over time. Triumph of the Optimists offers a chart displayed above that details the "default premium" and the cumulative total return of U.S. Aaa/Aa corporate versus Treasury bonds. The annual default premium includes not only losses from defaults, downgrades, and early calls, but spread widening and spread narrowing.

This is history’s total package of risk versus reward when it comes to corporate bonds, with a standard deviation by the way of 3.0% over the past 100 years. The message it sends is that yes, Aaa/Aa corporates do outperform Treasuries over time - by an average of 53 basis points a year. Since the annual default premium is 48 basis a year however, it says that in order to get that 53 you need to start off with a spread of (53+48) or 101 basis points. Today’s spreads of 30-35 are far shy of that and indicate that the odds of successfully outperforming Treasuries are substantially reduced. Holders of lower investment grade and junk bonds should heed this warning light as well.

Finally, if only to keep this Outlook reasonably brief, let me acquaint you with two charts from Ibbotson that absolutely fascinate me - and hopefully will do the same for you. The first is a table of long-term versus intermediate-term U.S. government bond returns over the past 75 years. Based on the horizon premium example mentioned on previous pages, an investor might reasonably expect to earn a total return advantage from long bonds, especially during a 75-year environment which offered a mild bull market as the table below indicates via the "capital appreciation" row. The returns however are almost identical (a 100-year " Optimists" study of the U.K. shows 5-year intermediate Gilts outperforming long Gilts by .2% annually). The secret to this conundrum comes from the simplistic phenomena of yield curve roll down, which allows for a 5-year Treasury to morph into a 4-year Treasury at a lower yield and a higher price over a 12-month period of time.

The figure is a table showing long-term U.S. Treasury bond and five-year Treasury note returns. For each security, the table includes total returns, income and capital appreciation, and gives the geometric mean, arithmetic mean, standard deviation, and serial correlation for each of the metrics. Data are detailed within.

The 5-year Treasury’s nearly identical performance, however, comes with the benefit of sharply reduced volatility as seen in the chart on the next page.

The figure is a line graph showing the volatility of long-term U.S. Treasury (government) bonds, intermediate-term government bonds, and Treasury bills from 1930 to 2002. Volatility is expressed on the vertical axis as monthly standard deviation. For the long-term and intermediate bonds, volatility peaks in the early to mid-1980s, then declines by 2002 to about 2.3% for long-term bonds, down from 4.8%, and 1% for intermediates, down from about 2.8%. For T-bills, volatility is fairly low and flat over the time span, at about 10 basis points in 2002, compared with about 25 basis points at its peak in the early 1980s. The volatility for long-term and intermediate government bonds trends upward since the mid-1960s, from around 0.75% for the long-term, and 0.40% for the intermediate. Volatility in 1930 is 1% for the long-term bond and 0.5% for the intermediate, with both moving up and down in a range until the late 1960s.

Such combinations are a bond investor’s dream. Identical returns - half the volatility. This history leads to a myriad of possible portfolio structures, all emphasizing the short to intermediate portion of the curve. Last month’s Outlook detailed some of them. For those investors who value higher returns as opposed to volatility but want to match liability durations of 10+ years, a double or triple barreled portfolio of intermediate bonds should be a viable solution. For those investors who treasure a stable net asset value and a good night’s sleep, yet want returns close to the yields offered by long-term bonds, a simple intermediate-term portfolio might be the answer. Long bonds are the loser in this historical and presumed future bond market environment.

And for those of you already conversing with Wilson the volleyball, I offer my humble apologies. Desert islands inhabited au solitaire can lead to strange behavior from even seemingly normal types. What I experienced on my "island" was a lesson in financial history that could pay huge "dividends" in future years. That history points towards an environment of lower than expected real rates of interest, low total returns for bonds (a 4% total return future world), an apparently overvalued corporate sector, and intermediate maturity bonds that should perform equally with long bonds at half the volatility. The one bond investment that fits into each of these boxes? Intermediate maturity TIPS. You’ll likely only earn 2-3% annually after adjusting for inflation, but hey - 25 years on that island, you get rescued, come back, cash in that 401k and you’ve got twice as much as you had before in inflation-adjusted terms. For those of you who prefer to avoid islands altogether while managing a bond portfolio of countless millions or billions of dollars, I suggest you bone up on your financial history anyway. It may not repeat, but it surely rhymes and what a sweet sound that outperformance can make. Mark Twain, Wilson - and Tom Hanks - would be envious.

William H. Gross

Managing Director


Past performance is no guarantee of future results. The graphs portrayed are not indicative of the past or future performance of any PIMCO product. This article contains the current opinions of the author and such opinions are subject to change without notice. This article has been distributed for educational purposes only and is not a recommendation or offer of any particular security, strategy or investment product. Information contained herein has been obtained from sources believed to be reliable, but not guaranteed.

Each sector of the bond market entails risk. Municipals may realize gains and may incur a tax liability from time to time. The guarantee on Treasuries and Government Bonds is to the timely repayment of principal and interest, shares of a portfolio are not guaranteed. Mortgage-backed securities and Corporate Bonds may be sensitive to interest rates. When interest rates rise, the value of fixed income securities generally declines and there is no assurance that private guarantors or insurers will meet their obligations. An investment in high-yield securities generally involves greater risk to principal than an investment in higher-rated bonds. Investing in non-U.S. securities may entail risk due to non-U.S. economic and political developments, which may be enhanced when investing in emerging markets. Inflation-indexed bonds issued by the U.S. Government, also known as TIPS, are fixed-income securities whose principal value is periodically adjusted according to the rate of inflation. Repayment upon maturity of the original principal as adjusted for inflation is guaranteed by the U.S. Government. Neither the current market value of inflation-indexed bonds nor the value of shares of a fund that invests in inflation-indexed bonds is guaranteed, and either or both may fluctuate. Duration is a measure of price sensitivity expressed in years.

No part of this article may be reproduced in any form, or referred to in any other publication, without express written permission. ©2004, Pacific Investment Management Company LLC.

Select Your Location


  • The flag of Canada Canada

Europe, Middle East & Africa