If there were a prize for the most disparaged asset class, it would likely go to bonds. And the most disparaged bonds would almost certainly be Treasuries. Jeremy Siegel called them "a sucker's bet." Warren Buffett said: "I can't imagine anyone having bonds in their portfolio when they can own equities."

For ordinary investors, however, holding only equities makes little sense. Fixed-income investments inevitably play an important role, particularly in retiree portfolios.

Treasuries are among the most useful of all fixed-income instruments. David Swensen, who manages Yale's endowment fund, makes a strong case for them. While he seeks returns from equities, he relies on Treasuries as a hedge against equity risk.

Investing in corporate bonds involves credit risk, Swensen argues, which compromises the effectiveness of the hedge. For individual investors, he recommends splitting the fixed-income allocation evenly between fixed-coupon Treasuries and Treasury Inflation-Protected Securities (TIPS).

Treasuries and TIPS are distinct asset classes, he says. TIPS provide riskless real returns and protect against inflation, while Treasuries provide riskless nominal returns and protect against deflation.

There is much talk today about the bond bubble (and when it might burst), but I tend to place more emphasis on stock valuations in guiding my allocation recommendations. Cyclically adjusted price-to-earnings measures, popularized by Robert J. Shiller of Yale, have been good predictors of long-term stock market performance. Based on these measures, stocks appear overvalued today. At the risk of violating popular wisdom, I currently recommend some portfolio overweighting in bonds.

Of course, the Federal Reserve Board's renewed attempt at quantitative easing indicates that we could be looking ahead to an extended period of low, steady interest rates. In Early November, the Fed announced it would buy back $600 billion in Treasuries. But there's a lot of conflicting evidence when trying to predict the future course of interest rates.

 

SOME HISTORY

History helps show the effectiveness of Treasuries as an equity hedge. First, we will take a long-term view using annual data going back to 1926. Then we will turn our attention to performance during times of crisis and examine monthly returns during 2008 and early 2009. It turns out that the recent crisis provides the strongest evidence of the hedging power of Treasuries.

We can get a long view of the comparative hedging abilities of Treasuries and corporate bonds by using Ibbotson data. We can't analyze TIPS in the same way because the U.S. government only began issuing them in the late 1990s.

Sharpe ratios, which provide a measure of return in relation to volatility, allow us to measure hedging power. We calculate Sharpe ratios for hypothetical portfolios invested 50% in stocks and 50% in each of the three bond categories (see "Little Difference," above). These Sharpe ratios can be used to calculate the probability of earning an annual return in excess of the risk-free rate-3.7% in this case, based on average T-bill rates.

Note that the probabilities for the three bond categories are nearly identical (67.46% for high-quality corporate bonds, 67.34% for long-term Treasuries, 68% for intermediate-term treasuries). The higher average returns on corporate bonds almost exactly offset the Treasuries' lower standard deviation and lower correlation with stocks.

The long-term data indicate little difference in diversification power between corporate bonds and Treasuries. This may be yet another indication that it is not easy to reduce risk without reducing return.

By looking at monthly returns during the worst months of the financial crisis, we get a different view of the ability of Treasuries to hedge equity risk. Because the period is recent, we are also able to compare Treasuries to more types of bonds (see "Hedging Bets," at left).

For these examples, I have modified the Sharpe ratios so that they are based on the probabilities of 50/50 portfolios earning greater than minus 25%-i.e. the probability of diluting losses by half when equities lose roughly 50%. (Note that this 14-month period provides only a limited sample. It shows relationships but also produces some anomalies.)

The two types of corporate bonds-high yield and high quality-do a relatively poor job of diversification. They suffer in terms of both returns and correlation. The high-yield bonds perform more like equities and are least able to hedge equity risk.

In contrast, the two types of Treasuries offered significant diversification. Long-term Treasuries earned the highest returns, and intermediate-term Treasuries were negatively correlated with stocks. Both returns and correlations have an effect on diversification power.

It may come as a surprise that returns on long-term Treasuries were positively correlated with stocks during this period. An examination of the month-by-month returns indicates a lot of "noise" in the monthly returns for long-term Treasuries that smoothed out over the full 14-month period, but made the monthly correlations unreliable.

TIPS performed poorly in this instance, and their results may be an outlier produced by market anomalies in late 2008. After Lehman Brothers declared bankruptcy, TIPS were dumped in large quantities on the market, causing values to fall (and yields to rise) at the same time that stock values were falling.

Although TIPS won't always perform badly when stocks go down, Treasuries are probably better diversifiers. The demand for Treasuries increases when inflation expectations decline-this is often tied in with poor stock market performance-while the reverse applies for TIPS.

 

PRACTICAL IMPLICATIONS

Even though the long-term risk/return tradeoffs for the different types of bonds are similar, I believe that fixed-income strategy matters a great deal in times of distress. Treasuries do the best job hedging equity risk during market turmoil when investors are most tempted to make harmful decisions-such as bailing out of equities when the market is down. At such times, it can be reassuring to own investments that move in the opposite direction.

I advise clients to split their fixed-income allocations equally between Treasuries and TIPS. The simplicity of U.S. government obligations-no credit risk, no call features and good liquidity-has great appeal. I also like the protection against deflation that Treasuries offer and the inflation protection offered by TIPS.

Some might argue for a lower weighting of bonds (or investing in shorter-duration bonds) in light of today's bond bubble. Entering this term into Google produces close to 600 relevant hits; certainly there is an abundance of concern.

However, our most recent bubble busts have followed periods of exuberance, and exuberance is in short supply these days. My feeling is that risks of higher interest rates are already reflected in the yield curve, and I am comfortable recommending maturities that match the overall bond market.

 

DIG DEEPER

Historically, the correlation between stock and bond returns has varied significantly. We have developed some understanding of how correlations can vary in times of financial distress, but large questions remain unanswered. We don't know enough about why correlations between stocks and Treasuries were in the 40%-50% range during much of the 1970s, 1980s and 1990s, then dropped to minus 30% for much of the past decade before narrowing to close to zero (based on rolling 60-month data from Ibbotson).

A better understanding of the factors that influenced correlations in the past would help. In asset allocation modeling, we tend to use static correlation assumptions based on historical averages. The development of dynamic models based on a better understanding of history would improve the usefulness of financial planning models.

We also need to develop a better understanding of expected cross-correlation among TIPS, Treasuries and stocks so that we can model asset allocations involving these three asset classes. Although we do not have much TIPS history, we may be able to gain insights from models that simulate TIPS returns. We recognize, however, that such studies will involve multiple uncertainties, requiring careful judgment as well as number crunching.

 

Joseph A. Tomlinson, FSA, CFP, is a financial planner and actuary. He is managing director of Tomlinson Financial Planning in Greenville, Maine. He can be reached at joetmail@aol.com.