Swedroe: Credit Risk Isn't Worth It

From 1926 through 2014, the default premium (the annual return on long-term, investment-grade corporate bonds minus the annual return on long-term Treasurys) has been just 0.22 percent. Such a small premium has led many observers, including me, to conclude that investors willing to accept higher levels of portfolio risk in exchange for higher expected returns should seek those higher returns in places other than corporate debt.

In addition to the miniscule realized premium, another problem for holders of corporate debt is that default risk doesn’t mix well with equity risk. The risk of default has a nasty tendency to appear at just the same time the risks associated with owning equities show up.

In their paper—“Explaining the Rate Spread on Corporate Bonds,” which appeared in the February 2001 issue of The Journal of Finance—Edwin Elton, Martin Gruber, Deepak Agrawal and Christopher Mann demonstrated that a substantial portion of the credit spread is in fact attributable to factors related to the equity risk premium.

They found that expected losses account for no more than 25 percent of the corporate spread. In the case of a 10-year, A-rated corporate bond, just 18 percent of the spread was explained by default risk. They also found that the Fama-French three-factor model explains as much as 85 percent of the portion of the spread that isn’t explained by taxes and expected default loss.

Finally, they found that the lower the credit rating (and the longer the maturity), the greater the explanatory power of the model. Thus, much of the expected return to high-yield debt is explained by risk premiums associated with equities, not debt. These risks are systematic and cannot be diversified away.

Evidence Supported by Theory

The case for an equity component to higher-yield bonds, turning them into hybrid securities, is supported by financial theory.

As Martin Fridson explained in his paper, “Do High-Yield Bonds Have an Equity Component?”, which appeared in the summer 1994 issue of Financial Management: “In effect, a corporate bond is a combination of a pure interest rate instrument and a short position in a put on the issuer’s equity. The put is triggered by a decline in the value of the issuer’s assets to less than the value of its liabilities, resulting in a default—putting the equity to the bondholders. For a highly-rated company, the put is well out of the money and is not likely to be exercised. The option consequently has a negligible impact on the price movement of the bonds, which is more sensitive to interest rate fluctuations. In the case of a high-yield bond, however, default is a realistic enough prospect to enable the equity put to affect the bond’s price materially. With the equity-related option exerting a greater influence on its price movement, the high-yield bond will track government bonds (pure interest rate instruments) less closely than the investment-grade bond does.”

A New Look

Attakrit Asvanunt and Scott Richardson, authors of the March 2015 paper “Credit Risk Premium: Its Existence and Implications for Asset Allocation,” took a new and interesting look at the credit premium issue.

Their innovation was to first remove the influence of interest rates. Most past research computes the credit premium as the difference between long-term corporate bond returns and long-term government bond returns. However, this creates a problem because these two types of bonds have different exposures to interest-rate sensitivities.

The higher yields on long-term corporate debt results in shorter duration than long-term term government bonds. And given the historical term premium, matching maturities (instead of duration) will result in an understating of the credit premium. Another issue raised by the authors is that, if you only look at the safest issuers (investment-grade debt), you could underestimate the credit premium.

Call Features & Taxation

I’d add two more issues not addressed by the authors. The first is that most long-term corporate debt has a call feature, which impacts their interest-rate sensitivities. And part of the premium on corporate debt is compensation for that asymmetric call risk, and thus not related to default risk. That said, in the latter part of the sample data (1988 onward), the authors do use an option-adjusted duration measure, therefore accounting for the callability.

The second issue is that while interest on U.S. government securities is not taxed at the state or local level, interest on corporate debt is. Investors must be compensated for the tax differential. That results in a tax premium on corporate debt that isn’t related to default risk.

Following is a summary of the finding from Asvanunt and Richardson’s study:

• By isolating interest-rate sensitivity, they confirm the “existence of a positive premium (referred to as the credit excess return) for bearing exposure to default risk.”
• The average monthly credit excess return on investment-grade corporate bonds over the period 1936 through 2014 was 0.11 percent (an annual premium of approximately 1.3 percent).
• The average monthly credit excess return for the aggregate high-yield corporate bond index during the period August 1988 through December 2014 was 0.21 percent (an annual premium of about 2.5 percent). Using credit default swaps, they found that over the shorter period April 1988 through December 2014, the average monthly credit excess return was 0.50 percent. They did note that the return benefited from a tightening of the spread by 0.51 percentage points over this period. That in turn explains 0.19 percent of the premium on an annual basis.
• There’s considerable time variation in the magnitude of the credit risk premium. It’s larger during periods of economic growth and during periods of lower-than-expected aggregate default rates.

Further Findings

The authors noted that the historical default premium is negatively impacted by the “bad selling practice” of divesting fallen angels (former investment-grade issuers who have been downgraded to below-investment-grade rating) as the result of portfolio rating requirements.

That, along with losses from rating migration and actual defaults, generates ex-post realized credit excess returns that are lower than initial credit spreads. That’s why they felt it was important to document the existence of a credit risk premium by looking directly at credit excess returns.

The authors also found there’s a consistently positive exposure between credit excess returns and equity excess returns (correlation of about 0.3). In other words, they confirmed prior research concluding that equity excess returns capture a considerable portion of credit excess returns.

There’s a very logical reason for that outcome. Corporate credit, like equities, is an asset class that will benefit from increasing expectations of economic growth as well as inflation. Companies benefit from overall economic growth because it can increase their ability to generate free cash flow that, in turn, increases their asset value, thereby making them safer.

Credit spreads naturally fall during periods of rising economic growth, giving rise to positive credit excess returns. In fact, the authors observed no reliable evidence of a credit risk premium over the 41-year period 1973 through 2014. They attributed this to the lower economic growth that persisted in the 1970s and 1980s

Diversification Benefits?

One of the arguments made in favor of including an allocation to higher-yield debt in a portfolio is that it provides investors a diversification benefit derived from the relatively low correlations of high-yield debt to both stocks and Treasury bonds. And the authors make this case. However, there’s a very logical explanation for these low correlations, and it’s not one that’s favorable to including credit risk in your portfolio.

It’s easy to understand why high-yield bonds have relatively low correlation with investment-grade bonds and equities. First, as we discussed, the higher yield of low-grade bonds results in a shorter duration, making them less sensitive than investment-grade bonds to movements in interest rates.

And as we also discussed, their effective duration is lower because low-grade bonds are often called earlier—they have weaker call protection than investment-grade bonds. In addition, their credit ratings are more likely to rise. Low-grade bonds are more sensitive to changes in stock prices than investment-grade bonds. With rare exception, investment-grade bonds respond primarily to changes in rates. Treasury bonds, of course, respond only to changes in interest rates.

Other Considerations

One of the more beneficial characteristics of Treasury securities is that, at least for U.S. investors, there’s no need to diversify credit risk because there is none. All of the risk is interest-rate risk. As you move down the spectrum from Treasurys to investment-grade bonds and then to junk bonds, you begin to take on risk that can be diversified away (the idiosyncratic risks of the individual companies).     

Investors are not compensated for taking diversifiable risks with higher expected returns. Because the lower the credit rating, the more equitylike the investment, the need for diversification increases as you move down the credit spectrum.

The implication for portfolio construction is that, once you move past Treasury securities, prudence dictates the need to diversify credit risk. And that means incurring the expenses of a mutual fund, which can be avoided if you limit your investments to Treasurys. Thus, some of the risk premium built into the prices of higher-yielding debt will be eroded by the incremental costs incurred.

For individual investors, there’s yet another important consideration. Typically, individuals can obtain higher yields than those available on Treasurys without taking any of the credit risk involved in corporate debt. For example, as of this writing, the yields on five- and 10-year Treasurys are about 1.4 and 2.0 percent, and the yield on five- and 10-year FDIC-insured CDs are about 2.2 percent and 2.8 percent, respectively.

In each case, those yields are about 0.8 percentage points higher than the yield on the equivalent-term Treasury. What’s more, that difference—0.8 percentage points—is only about 0.5 percentage points below the annual default premium found by Asvanunt and Richardson. And that doesn’t account for either the tax premium or the call premium. Add in the expenses of a mutual fund and the premium would likely disappear completely, if not turn negative.

Not Worth The Risk

While Asvanunt and Richardson make a case for the default premium being understated, I cannot find any reason to include higher-yielding corporate debt in a portfolio.

The best example of why you shouldn’t include such debt is what happened in 2008 when the S&P 500 Index fell 37.0 percent. Vanguard’s Intermediate-Term Treasury Fund (VFITX) provided much-needed shelter from the storm, returning 13.3 percent. On the other hand, the firm’s Intermediate-Term Investment Grade Corporate Fund (VFICX) lost 6.2 percent and its High-Yield Corporate Fund (VWEHX) lost 21.3 percent. Just when the safety that bonds are supposed to provide was sorely required, credit risk showed up and exacerbated equity portfolio losses instead of mitigating them.

As further evidence, over the 15-year period ending March 26, 2015, Vanguard’s high-yield fund (VWEHX) returned 6.39 percent, outperforming the Intermediate-Term Investment Grade Corporate Fund (VFICX) by just 0.05 percentage points and the Intermediate Treasury Fund (VFITX) by just 0.58 percentage points.

Assuming you were able to invest in FDIC-insured CDs with the same maturities, if today’s spreads are a good indication, you could have outperformed VWEHX without taking on any of the incremental credit risks. And CDs certainly would have produced a more positive impact on the portfolio, while also reducing the left tail risk that was experienced in 2008.

A fitting conclusion is this advice from David Swensen, the highly regarded chief investment officer of the Yale Endowment: “Well-informed investors avoid the no-win consequences of high-yield fixed-income investing.”

Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.