Just How Difficult Is It to Time the Bond Market?

    by Mark Hulbert

    Bond market timers, though their job never has been easy, have faced even tougher challenges in recent years.

    Have they risen to the occasion?

    One of the biggest challenges they face is the structure of the bond market, and particularly the Treasury sector of the bond market. There, “value” and “price” are virtually equivalent at a given maturity, with the primary determinant being the level of interest rates.

    Relative to the bond market, the stock market is more opaque, brimming with inefficiencies that have at times led the entire market into periods of overvaluation or undervaluation. As a result, one would expect it to be relatively easier to time the stock market than to time the bond market.

    That’s a big enough challenge in itself. But recently, a couple of additional hurdles have been placed in front of bond market timers. One can be traced to the extraordinarily low interest rates in recent months, which have propelled the bond market into uncharted territory.

    The lack of recent historical parallels has made it even more difficult for bond timers.

    Another challenge facing bond timers has been a wholesale shift in the correlations between the bond and stock markets, which also makes it difficult to draw historical parallels.

    Stephen Todd, editor of the Todd Market Forecast, recently described the difficulty timers have faced: “For the past five years, analysts have gotten used to the idea that bonds and stocks tend to go in opposite directions, but this wasn’t always the case. For decades prior to 1998, the two major financial instruments marched together, frequently in lockstep.”

    These various considerations lead to two specific hypotheses about market timing:

    • A greater percentage of stock market timers is able to beat a buy-and-hold strategy than of bond market timers; and

    • The percentage of bond timers beating a buy-and-hold approach has dropped in recent years.

    The Data

    The part of the Hulbert Financial Digest database that I used to test these hypotheses measures newsletters’ abilities to time the stock or bond markets. For the purposes of this test, I ignored their abilities to pick individual stocks or mutual funds.

    When calculating newsletters’ performances on this timing-only basis, we allowed newsletter editors only two possible investments: the market and cash.

    • When they chose the market, they earned the return either of the Wilshire 5000 Total Return index (in the case of the stock market timers) or that of the Shearson Lehman Treasury index (in the case of the bond timers).

    • When they chose cash, they earned the return of 90-day Treasury bills.
    Therefore, the sole source of differences in the advisers’ returns was their market timing.

    To determine if timers were able to beat the market, I focused on their risk-adjusted returns, which compares their performances to the volatility of their returns. For every month since the beginning of 1990, I calculated timers’ risk-adjusted returns over the trailing 60 months. The 1990 starting date was chosen because that was the first point at which the Hulbert Financial Digest had five years of data for bond market timers. The end point of the study was May 31 of this year.

    The Results

    Figure 1.
    % of Market-Timing
    The results are summarized in Figure 1.

    The two data series plotted in the figure report the percentages of each group of timers that beat a buy-and-hold position.

    Overall, I found support for both hypotheses.

    First, you can see in Figure 1 that, with the exception of several years in the mid-1990s, the percentage of bond timers beating a buy-and-hold approach was below the percentage of stock timers who were able to do so. Averaging all months since January 1990, the percentage of bond timers beating a buy-and-hold approach was 19.5%, in contrast to 28.6% for stock timers.

    Second, the figure reveals the dramatic drop in recent years in the percentage of bond timers able to beat a buy-and-hold approach. Since the beginning of 2001, the proportion of bond newsletters able to beat a buy-and-hold approach with a timing strategy has not risen above 20%, and more recently only 4% have done so.

    Timing: Poor Odds

    The implications for individual investors are twofold:

    • Timing is difficult: The percentage of market timers able to beat the market was low in both markets, suggesting that timing is not a particularly successful approach in either the bond or stock markets.

    • Stock market timing offers better odds: If you were to insist on trying your hand as a market timer, the data suggest that your odds of success are higher in the stock market than in the bond market.
    What if you still wanted to try timing the bond market?

    In that event, you would probably want to focus on those few bond timers who have been able to beat the bond market’s risk-adjusted performance.

    Just one bond timer on the Hulbert Financial Digest list currently satisfies this criterion, however: Vantage Point, edited by John Harris (Vantage Point: An Independent Report for Vanguard Investors, 300 Mount Lebanon Blvd., Suite 2218-A, Pittsburg, Penn. 15234-1508, 800/237-8400 ext. 597; $149/yr.). A portfolio that switched between the Shearson Lehman Treasury index and T-bills on Harris’ bond signals produced an 8.4% annual return over the last five years, in contrast to an 8.0% return for buying and holding. Better yet, this above-market return was produced with 16% less volatility—a winning combination.

    As of mid-June, Harris’ bond model was bullish, as it had been since April 2002. Harris cautions that the Federal Reserve’s rate-cutting policy could be sowing the seeds of inflation down the road. But in the meantime, because the Fed has “officially become a deflation fighter,” Harris believes the odds of a continued bond rally remain good.

    Mark Hulbert is editor of the Hulbert Financial Digest, a newsletter that ranks the performance of investment advisory newsletters. It is published monthly and is located at 5051B Backlick Rd., Annandale, Va. 22003; 703/750-9060; www.hulbertdigest.com.

→ Mark Hulbert