Interest rates, which have sat at or near historical lows during the past seven years, have led many investors to seek additional yield in the form of credit risk. The recent trend, and its popularity, gives us an opportunity to determine if this risk historically has been rewarded by examining the credit risk premium in investment-grade as well as high-yield corporate bonds.
To measure the credit risk premium, my colleague Brian Haywood first looked at how investment-grade and high-yield corporate bonds have performed relative to Treasurys:
|Standard Deviation (%)||6.69||17.19|
|Negative Years (%)||42||38|
The table above shows the investment-grade credit premium and high-yield credit premium as measured by the difference in returns between the Barclays U.S. Corporate Investment Grade Index (duration of about 7.2 years), the Barclays U.S. Corporate High Yield Index (duration of about 4.2 years) and their respective Treasury counterparts.
To make sure we are measuring securities with similar duration risk, we've also used data from Barclays that computes a comparable Treasury index that closely matches the duration and convexity of each bond index. In doing so, the credit premium can be measured accurately.
Keeping durations as close as possible, we find that over the past 26 years, the investment-grade credit premium has averaged 0.51% per year, with a Sharpe ratio of only 0.08 and a t-statistic (a measure of statistical significance) of 0.4. A t-stat of 2.0 signifies statistical significance at the 5% confidence level. Thus, the investment-grade credit premium hasn't been robust.
However, this result does not consider implementation costs. Because there is no credit risk in Treasury securities, there isn't any need to diversify them. Thus, there is no need to own them through a mutual fund (with the exception of the possible benefit of convenience).
On the other hand, because of the need to diversify the credit risk inherent in corporate bonds, investors are generally better served by owning such bonds indirectly through a mutual fund (or ETF).
That involves expenses in the form of expense ratios and possibly commissions (and in the case of ETFs, bid/offer spreads will be incurred as well). It's also important to note that the investment-grade credit premium has been negative about 42% of the time.
At 3.25% per year, the high-yield credit premium might look enticing (it certainly is economically significant). But when you examine its Sharpe ratio of 0.19, it becomes much less appealing thanks to the significant increase in volatility as measured by standard deviation.
Much like the investment-grade credit premium, the t-stat for the high-yield credit premium is also statistically insignificant at 1.0. In addition, the high-yield credit premium was negative 38% of the time. Also consider that the high-yield credit premium was actually negative over the full 20-year period from 1989 through 2008.
With that said, one of the most common mistakes investors make is to consider an investment in isolation instead of how its addition impacts the risk and return of the entire portfolio.
Keeping that in mind, we'll examine how the credit premiums have performed within the context of a standard 60% equity/40% fixed-income portfolio. The table below demonstrates how investment-grade and high-yield securities have fared in a balanced portfolio versus a balanced portfolio using comparable duration-matched Treasurys.
The equity portion in each of these four diversified portfolios is allocated 60% to the S&P 500 Index, 30% to the MSCI EAFE Index and 10% to the MSCI Emerging Markets Index. The fixed-income portion of Portfolio A is allocated to the Barclays U.S. Corporate Investment Grade Index (duration of about 7.2 years). The fixed-income portion of Portfolio B is allocated to the Barclays U.S. Corporate High Yield Index (duration of about 4.2 years).
We'll then use data from Barclays to compare the performance of Portfolios A and B to the returns of portfolios that use duration-matched Treasurys for their fixed-income allocation. In this way, we avoid mixing two issues (the term premium and the credit premium) and isolate the impact of credit risk:
|60/40 Treasury||Portfolio B
|Average Return (%)||9.54||9.33||10.31||9.01|
|Standard Deviation (%)||12.19||10.46||16.33||11.01|
As you can see from the table above, the portfolio with investment-grade corporates posted 0.21 percentage points in additional return when measured against a portfolio that employed Treasurys. That additional return, however, came with higher volatility as measured by standard deviation.
This trade-off is reflected in the two portfolios' risk-adjusted returns as measured by the Sharpe ratio. On a risk-adjusted basis, the portfolio using Treasurys produced superior risk-adjusted returns. With a Sharpe ratio of 0.58 (versus 0.51), the portfolio with Treasurys was about 14% more efficient, relatively, than the portfolio with investment-grade bonds.
The high-yield portfolio paints a similar picture. The addition of high-yield corporate bonds resulted in a return 1.3 percentage points higher. However, it also came with significantly more volatility. With a Sharpe ratio of 0.52 (versus 0.43), the portfolio with Treasurys was about 21% more efficient, relatively, than the one with high-yield bonds.
In conclusion, over the period we examined, both the investment-grade and high-yield credit premiums have not only been insignificant from a statistical perspective, but the additional returns produced by the credit premiums have been "drowned out" by high volatility and their higher correlations (they are less effective diversifiers) with the equity holdings in the portfolio.
The result was that by including them in a diversified portfolio, instead of safer and lower-yielding Treasurys, investors received lower risk-adjusted returns.
Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.