Credit Premium: Fact(or) Fiction
Are corporate bonds additive to portfolios that already own stocks and government bonds? Jared Kizer revisits this question with a look at the historical justification for, and some facts and fiction regarding, the investment-grade credit premium.
The above title may well end up being my most significant marketing achievement. Yet, it is a good descriptor of a research topic I’ve tackled, and I continue to be surprised by how little attention has been paid to whether there’s any historical justification that the credit premium — the difference in return between corporate bonds and comparable maturity government bonds — is additive to portfolios that already own stocks and government bonds. Based upon work I’ve previously done and the working paper I posted in March, I’m convinced that it isn’t, at least in the form it takes in most investment-grade corporate bond indexes and many investment-grade corporate bond funds. Here, in the spirit of AQR’s pieces on various premia, I’ll detail what I believe are some facts and fiction related to the investment-grade credit premium that investors should understand.
Fiction: Early 20th century data on the credit premium is reliable and good quality.
I’m as much of a fan as anyone of using the longest-run data possible for research purposes. However, a close inspection of the primary series typically used for long-run analysis of the credit premium — Ibbotson and Sinquefield’s default premium series — reveals that its early history is highly suspect. In the piece I first linked to above, I find that the series had a Sharpe Ratio in excess of 1.30 (!?!?!?) during the 1930s and that the credit premium’s annual returns exhibited positive correlation with Moody’s annual measure of corporate bond defaults. This directional correlation relationship, of course, makes no sense since one would expect to see years with larger numbers of corporate bond defaults to correspond with below average, and commonly negative, returns for the credit premium. This expected relationship is indeed what you do see when examining more recent historical measures of the credit premium (e.g., Barclay’s measure of the credit premium that it began calculating in 1988). In my mind, the fact pattern of impossibly high risk-adjusted returns within the context of the 1930s and a nonsensical relationship with measures of corporate bond defaults renders the early history of the Ibbotson and Sinquefield data useless.
Fact: After controlling for equity market risk, the credit premium hasn’t been statistically significant.
Once you focus on the reliable history of the credit premium, you find that the credit premium has been meager in absolute terms and not statistically significant either before or after adjusting for embedded equity market risk. Using Barclay’s data, the credit premium was just 54 basis points (bps) per year with a volatility of 351 bps over the period of 12/1974–10/2017 (the period used in the research paper linked to above). Even before adjusting for equity market risk, this is not statistically reliable.
After adjusting for equity market risk via regression, the series has alpha that is no different from zero. Practically, this indicates that corporate bonds have not been additive to portfolios that include equities. In other words, corporate bonds provide exposure to two risks — interest rate risk and equity market risk — that are already present in most all investment portfolios.
Fact: The credit premium isn’t really a credit premium.
If you’re familiar with how most investment-grade corporate bond indexes and many corporate bond funds are managed, it’s interesting to note they aren’t generally exposing investors to credit risk in the true sense of the word. When I think about credit risk, I think of it as the risk that an asset I own won’t be able to pay back some portion of principal and interest while I own it. Since most investment-grade strategies own bonds rated Baa or higher and sell securities if they are downgraded to Ba or lower, it’s rare that bonds actually default while still in the index or fund. Instead, the risk an investor is exposed to is more adequately characterized as downgrade risk or repricing of credit risk risk (J).
The perhaps more insightful way to think about this is that when an investor purchases an investment-grade corporate bond fund, the portfolio management approach is similar to buying a portfolio of investment-grade corporate bonds and simultaneously buying puts on each of the holdings that will be in the money if the bond is downgraded to Ba or lower. Thought about in this way, it’s not surprising that what is characterized as the credit premium has been extremely weak historically because investors are effectively buying protection against the dominant risk of holding the bond (i.e., the risk of it actually defaulting).
Fact: Investment-grade corporate bond funds would be better served by continuing to hold bonds that are downgraded to junk.
In the Financial Analysts Journal paper “Capturing Credit Spread Premium,” Bruce Phelps and Kwok-Yuen Ng’s document that reconstructing Barclay’s investment-grade corporate index to allow bonds that get downgraded to remain in the index improved the realized annual credit premium from 48 to 80 bps over the period of 1990–2009. This result was statistically significant, meaning the improvement in return doesn’t appear to be a random result, and indicates a substantial portion of the true credit premium is lost as a result of the sell-when-downgraded to junk methodology. That said, my guess is this change in methodology doesn’t change the result that the premium is not significant after accounting for equity market risk, but investors who still want to own corporate bonds (or must own them due to regulatory constraints) would be well-served by looking for solutions that adopt the methodological changes outlined by Phelps and Ng.
By clicking on any of the links above, you acknowledge that they are solely for your convenience, and do not necessarily imply any affiliations, sponsorships, endorsements or representations whatsoever by us regarding third-party Web sites. We are not responsible for the content, availability or privacy policies of these sites, and shall not be responsible or liable for any information, opinions, advice, products or services available on or through them.
The opinions expressed by featured authors are their own and may not accurately reflect those of the BAM ALLIANCE®. This article is for general information only and is not intended to serve as specific financial, accounting or tax advice.
© 2018, The BAM ALLIANCE®
