Swedroe: Fixed Income’s Low Risk Anomaly

As investors search for yield, low risk doesn’t necessarily mean lower returns.

Modern financial theory now includes the existence of many anomalies that shouldn’t exist if investors were perfectly rational and markets were perfectly efficient.

Perhaps the most important anomaly is the persistent and pervasive momentum premium. Among the others are the low-volatility anomaly (low-volatility stocks have outperformed high-volatility stocks) and the poor performance of extreme small growth stocks, IPOs, penny stocks and stocks in bankruptcy. And while there is no debate about the equity or the small-cap premium being anomalies, there are many behavioral finance people who believe the value premium is an anomaly.

Interestingly, there hasn’t been much debate about the default and term premiums in bonds. Yet investors have historically received very little in the way of compensation for taking default risk, and not much of a premium for taking term risk once you get beyond the intermediate term.

The default premium from 1927 through 2013 has been just 0.3 percent. And while 20-year Treasurys produced higher returns than did five-year Treasurys (5.48 percent versus 5.29 percent), the longer bonds had much higher volatility (9.85 percent annual standard deviation versus 5.66).

The result is that the Sharpe ratio on five-year Treasurys was 0.33, while the Sharpe ratio on 20-year Treasurys was 0.24. It appears that just as with stocks, there is a low-volatility anomaly in bonds.

Historical Risk Premiums

From 1927 through 2013, the annual average risk premiums and annual standard deviations of those premiums have been:

• Equity Premium: 8.2 percent/20.3 percent
• Size Premium: 3.1 percent/12.7 percent
• Value Premium: 4.9 percent/12.8 percent
• Term Premium: 2.5 percent/9.4 percent
• Default Premium: 0.3 percent/4.2 percent

Not only has the default premium been negligible, its volatility has been 13 times the premium itself. In the cases of the other premiums, the ratio is in the range of about 2.5 to 4 times. In addition, while the term premium is uncorrelated with the equity premium, the default premium is correlated, with an annual correlation of 0.3 percent.

While investors have been well rewarded for taking the risks of investing in stocks in general, and specifically small stocks and value stocks, as well as for taking term risk, they have received almost no reward for accepting corporate credit risks.

While corporate bonds yields have been much higher than the 0.30 percent premium, the realized return has been very small because of defaults and the risks of bonds being called prior to maturity.

We’ll take a look at more of the historical evidence, evidence that’s likely to surprise most investors as well as provide a warning to those investors who are now stretching for yield by taking more credit or term risk.

Over the 88-year period of 1926-2013, the annualized return on long-term (20-year) Treasury bonds was 5.48 percent, and the return on long-term corporate bonds was 5.95 percent.

However, since assets shouldn’t be viewed in isolation (but in how their addition impacts the risk and return of the portfolio), we’ll see how the two investments performed within the context of a typical 60 percent stock/40 percent bond portfolio. Portfolio A holds long-term Treasury bonds and Portfolio B holds long-term corporate bonds.

As you can see, despite its 0.47 percent return advantage (5.95 versus 4.48) adding default risk resulted in only a very small improvement in the portfolio’s return (just 0.10 percent). And the higher return was offset by higher portfolio volatility. The result was no improvement in risk-adjusted returns (the Sharpe ratio). It’s important to understand that the above results are the returns of indexes, not live portfolios.

With Treasury bonds, the most liquid investment in the world, trading costs are extremely low. And since there is no credit risk, you don’t need to use a mutual fund to implement the plan. Instead, you can buy individual Treasurys on your own, directly from the Treasury, saving the expense ratio of a mutual fund.

If you’re going to own corporate bonds, you should use a mutual fund to diversify the credit risks. Thus, not only will you incur the fund’s expenses, trading costs in corporate bonds are higher than for Treasurys. The bottom line is that after considering implementation costs, the returns of the portfolio with corporate bonds would have been lower than the portfolio with Treasurys, despite the higher risk.

The following example, using high-yield (junk) bonds makes an even more compelling case. For the 30-year period of 1984-2013, the Barclays U.S. High Yield Index returned 9.5 percent, outperforming five-year Treasurys’ return of 7.4 percent by 2.1 percent. The table below shows the returns and Sharpe ratio for two 60/40 portfolios.

While the addition of high-yield bonds added to returns, the volatility of the portfolio increased far more than did the returns, resulting in a much less efficient portfolio—the Sharpe ratio of the portfolio with Treasurys is 20 percent higher than it is for the portfolio using high-yield bonds.

Over this period, the annual correlation of returns of the high-yield index to the S&P 500 was +0.6. On the other hand, the annual correlation of five-year Treasurys to the S&P 500 was -0.1. This demonstrates that not only are Treasurys better diversifiers of the risks of stocks, high-yield bonds are really hybrid securities, containing equitylike risks.

Another way to look at the issue is to consider that if you wanted to have a portfolio with higher expected returns than a 60/40 portfolio using Treasurys, the more efficient way to increase expected returns would have been to increase the equity allocation (not substitute high-yield bonds), as the table below demonstrates.

*measure of risk - adjusted returns

International Evidence

The authors of the paper “Low-Risk Anomalies in Global Fixed Income: Evidence from Major Broad Markets,” examined global data to see if the same phenomena existed outside the U.S. The study covered the 16-year period 1997-2012. They found that the same low-volatility (risk) anomaly that exists in the stock market existed in the bond market. Following is a summary of their findings:

• Portfolios with the lowest term and default risk would have delivered the largest positive alpha and highest Sharpe ratios.
• Portfolios invested in riskier (longer-term and lower credit quality) bonds would have delivered the most negative alpha and lowest Sharpe ratios.
• The results were extremely robust and confirmed for government bonds, quasi-government and foreign government bonds, securitized and collateralized bonds, corporate investment-grade bonds, corporate high-yield bonds, emerging market bonds and aggregates of some of these universes.
• The robustness of the results was confirmed by using different measures of risk.

While the above study covers a relatively short period, it does provide out-of-sample evidence that the same low-risk anomaly that is well documented in stocks exists in bonds as well. This has implications for investors considering taking credit risk and/or term risk (beyond the intermediate term) in a search for yield in the low-rate regime we have been living with since 2008.

Larry Swedroe is director of Research for the BAM Alliance, a community of more than 130 independent registered investment advisors throughout the country.