This commentary is written by Martin Fridson, a high-yield market veteran who is Chief Investment Officer of Lehmann Livian Fridson Advisors LLC, as well as a contributing analyst to S&P Global Market Intelligence.
The worst pandemic in a century rages on and unemployment remains in double digits. According to the U.S. high-yield market, however, things might have returned to normal. That, at least, is how one would usually interpret the fact that, recently, the high-yield spread-versus-Treasuries narrowed to slightly less than its historical average.
Admittedly, these are not normal times.
Specifically, the option-adjusted spread (OAS) on the ICE BofA US High Yield Index tightened to +550 bps on June 5, 2020, from +671 bps one week earlier. The June 5 OAS compares with a weekly mean of +553 bps since the beginning of 1997 (see table below).
The recent dramatic spread-narrowing was not solely a function of risk premiums contracting on distressed issues as the near-term threat of default diminished for some of the high-yield universe’s most troubled companies. As the table also shows, the ICE BofA US Non-Distressed High Yield Index’s OAS ended the first week of June at +414 bps, a touch below its historical average of +422 bps. A week earlier, the non-distressed spread was +477 bps.
In short, the high-yield market risk premium was almost exactly at its historical average.
According to the factors that determine the high-yield spread in normal times, however, risk is much higher than average. The article “Fair Value update and methodology review” (LCD News, Jan. 24, 2018) shows that 80% of the historical variance in the ICE BofA US High Yield Index’s OAS is explained by the five factors listed, along with a dummy variable discussed below.
As of last week, four of the five indicators indicated higher-than-average risk, with the fifth being of little consequence.
Net Credit Tighteners: Credit availability exerts a powerful influence on the high-yield spread. We measure it by the percentage of banks tightening credit standards minus the percentage easing standards, as reported by the Federal Reserve Board. (A positive number thus indicates tight credit.) The latest quarterly figure for this series is exceptionally high, at 41.6%, versus a 5.2% historical mean.
Economic Indicators: Both capacity utilization and the month-over-month change in industrial production are well below their historical means, as detailed in the table. Some readers may wonder why our Fair Value model uses these particular indicators, rather than the more closely followed unemployment rate and quarter-over-quarter change in gross domestic product. The reason is simply that our empirical investigation over the past 25 years has consistently found that a multivariate model incorporating capacity utilization and industrial production better explains variance in the high-yield spread than one that instead incorporates unemployment and GDP.
Default Rate: The speculative-grade, percentage-of-issuers default rate is right at the series’ historical mean of 3.9%. That is consistent with the high-yield risk premium, which is sometimes misconstrued as purely a default risk premium, being almost exactly at its historical mean. The default rate, however, is a backward-looking variable with little explanatory power regarding the high-yield spread. We include it in our model mainly to avoid the question, “How can you leave out the most important factor, the default rate?”
Five-Year Treasury Rate: The record clearly shows that all else being equal, the lower the Treasury yield, the wider the spread on high-yield bonds. That is to say, the popular notion that the spread is proportional to the underlying Treasury yield is flat-out wrong. Currently, the Treasury yield is far below its historical average, supplying one more reason that, were these normal times, we would expect to observe a spread well above its historical average.
The dummy variable mentioned above has a value of 1 when quantitative easing by the Fed is in place and a value of 0 when it is not. The presence of QE, all else being equal, reduces the ICE BofA US High Yield Index’s OAS by a substantial 148 bps. Even with the dummy variable currently set at 1, the Fair Value spread far exceeds the June 5 actual spread of +550 bps. In fact, it is more than double that level, at a whopping +1,302 bps.
Clearly, the market has not somehow overlooked the present, scarce availability of bank credit, weak economic indicators, and low Treasury rates. The existence of a historically average high-yield risk premium is testimony to the power of current Fed intervention in the market, which goes well beyond the quantitative easing instituted during the Great Recession. Most strikingly, the Fed has broken new ground by buying corporate debt, including, to a limited extent, the non-investment grade variety.
Shorting high-yield at this point would constitute fighting the Fed. That is a course that the late, illustrious money manager Martin Zweig cautioned against. At the same time, investors are well advised to keep in mind what the preceding analysis demonstrates. By process of elimination, Fed intervention, over and above QE, is the force keeping the high-yield spread at no more than its historical average. If that force were ever to be removed, prior to the factors displayed in the table returning to their historical averages, a huge widening of the high-yield spread would almost surely result.
Leveraged loans recommendation: Neutral
We are revising our previous Overweight recommendation to Neutral on leveraged loans in a portfolio that also invests in high-yield bonds. The Overweight recommendation was instituted in March, as the index value in the chart below swung from –1.64, indicating that bonds were extremely cheap versus loans, to 2.58, indicating exactly the opposite. That swing of 4.22 points on our index was the largest in at least a decade. The index value dropped to 1.25 in April but still represented an extreme relative undervaluation for loans.
In May, on the other hand, the index value dipped to 0.03, indicating almost perfectly fair relative valuation of the two asset classes. A Neutral stance is now clearly warranted. May’s 0.03 index level is the second closest to perfectly fair relative valuation (0.00) in the history captured by our analysis, behind April 2012’s reading of 0.01.
The present relative valuation of loans and bonds represents something of a quandary for high-yield corporate treasurers. Companies that finance in both markets, favoring the one that offers the cheapest capital at a given time, ordinarily choose the option that offers investors the least value. To a considerable extent, therefore, investors are able to buy new issues, whether bonds or loans, only when they are relatively overpriced. With our index almost exactly at the 0.00 Fair Value level, issuers are temporarily unable to get the better of either group of investors.
Details of our relative valuation methodology for the two categories of debt appeared in “Loans vs. bonds – Determining relative value” (LCD News, April 26, 2017). In brief, we compare the three-year discounted spread on the S&P/LSTA Leveraged Loan Index and the option-adjusted spread on the ICE BofA US High Yield Index, after first adjusting for differences in ratings mix between the two asset classes. We convert the difference in these spreads into an index geared to one standard deviation from the mean in either direction. A reading of plus 1.0 indicates that bonds are extremely rich versus loans, and a reading of minus 1.0 indicates that loans are extremely rich versus bonds. Inside those bands, we recommend Neutral weightings of loans and bonds.
Research assistance by Lu Jiang and Zhiyuan Mei.
ICE BofA Index System data is used by permission. Copyright © 2020 ICE Data Services. The use of the above in no way implies that ICE Data Services or any of its affiliates endorses the views or interpretation or the use of such information or acts as any endorsement of Lehmann Livian Fridson Advisors LLC’s use of such information. The information is provided "as is" and none of ICE Data Services or any of its affiliates warrants the accuracy or completeness of the information.