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.

On June 8, 2020, the S&P 500 stock index completed its recovery to its year-end 2019 level. From 3,230.78 on Dec. 31, 2019, the equity benchmark sank to a low of 2,237.40 on March 23, 2020, and closed on June 8 at 3,233.39.

"Explosive" fairly describes the stock market’s rebound from its trough over a period of just 2.5 months. From March 23 to June 8, prices rose by 44.5%, as measured by the S&P 500, and by an even greater 53.3% over the same span based on the Russell 2000 Index. By comparison, over a comparable interval following the Great Recession low on March 9, 2009, the S&P 500 gained "just" 34.6% and the Russell 2000 rose by 45.8%.

At the individual stock level, a noteworthy feature of the recent rally was the inverse correlation between credit quality and performance, as detailed below. No such relationship characterized the downturn from year-end 2019 through the March 23 low. We see these facts as evidence that the equity rally was driven mainly by the Federal Reserve’s aggressive moves to support the credit market, rather than prospects for a normal recovery in economic activity and corporate earnings.

For high-yield investors who are tempted by prevailing yields on the lowest-rated issues, the outperformance of stocks with bottom-tier ratings should inspire a certain amount of confidence. As the market beefs up the value of the equity underlying the lowest-rated bonds, their credit risk declines by certain measures accorded attention by some high-yield analysts. These include Edward Altman’s Z-score and the market-adjusted debt (MAD) ratio pioneered by Drexel Burnham Lambert high-yield research director Larry Post in the 1980s.

Let us now describe our study and present our findings.

Sample set
The starting point of our analysis was the member listings of the ICE BofA US Non-Distressed High Yield Index. This subindex of the ICE BofA US High Yield Index excludes issues with option-adjusted spreads of +1,000 bps or more. We sought to minimize statistical noise in our study by eliminating the huge, erratic price swings frequently observed in securities of companies with high perceived risk of near-term bankruptcy. On similar grounds, we excluded companies in the Energy industry, which had recently experienced extreme volatility due to large fluctuations in oil prices.

We set about to create two test samples, consisting respectively of stocks of companies with bond ratings at the high end and the low end of the speculative-grade range. The classifications we employed were the ICE Indices' Composite Ratings, which are averages of the ratings of (as applicable) S&P Global Ratings, Moody’s, and Fitch Ratings. To prevent differing capital structure priorities from producing inconsistent classifications, we purged our sample of all issues other than senior unsecured ones. Note that an issuer’s senior unsecured rating may or may not be equivalent to its corporate rating.

READ MORE: Fridson on how the recent signals from US high yield market hint at return to 'normal,' despite increased risk.

Ideally, our lowest-rated sample would have consisted entirely of companies with senior unsecured debt rated in the CCC tier. Unfortunately, a large percentage of the companies with bonds so rated, and satisfying our other criteria, had no publicly traded stock, due to ownership by private equity investors. That constraint rendered it impossible to identify 30 qualifying companies, the number generally regarded as the minimum required to constitute a statistically valid sample. Accordingly, we broadened our definition of the low end of speculative-grade to encompass the B3 Composite Rating category.

Note that issuers with B3 corporate ratings are not just marginally riskier than others within the broad B category. For the years 1983–2019, Moody’s mean annual default rate rises only from 2.33% at the B1 level to 3.14% at the B2 level, but jumps to 7.23% at B3. Our final bottom-tier sample included stocks of 14 companies with B3 Composite Ratings. As for our highest-rated sample, we had no difficulty creating a randomly selected sample of 30 companies with BB senior unsecured bonds. This was owing to a larger number of issuers in that broad rating category, compared with the CCC universe, as well as a low incidence of private equity ownership within it.

Analysis
For each stock in our two 30-company samples, we collected the closing price at the start (Dec. 31, 2019), low point (March 23, 2020), and end (June 8, 2020) of our observation period. We then calculated mean percentage price changes and standard deviations for the down phase (labeled “Start to low” in the table below), rebound (“Low to end”), and full observation period (“Start to end”). Our objective was to determine whether stocks with higher-end ratings performed differently from those with lower-end ratings over any of these intervals.

Results
In the sell-off phase, from the end of 2019 to the March 23, 2020 trough, stocks of companies with B3 to CCC3 Composite Ratings suffered a larger mean price decline than those with BB1 to BB3 ratings. The respective price changes were –51.1% and –46.2%. That difference was not statistically significant, however, as indicated by the statistical data (t-statistic and p-value) shown at the bottom of the first of the table’s three major columns.

The results were much different in the recovery phase, March 23 to June 8, 2020. In this period, the B3 to CCC3 sample’s superior mean price gain of +108.5%, versus +74.6% for the BB1 to BB3 sample, was significant at the 95% confidence level. After displaying indifference to credit quality while the equity market plummeted in the earlier part of the year, stock investors exhibited a documentable preference for companies with the riskiest credit profiles when the market came roaring back.

For the full period from start to end, investors’ behavior in the recovery phase did not completely offset their behavior during the selloff phase. True, over the Dec. 31, 2019, to June 8, 2020, interval, the B3 to CCC3 sample performed better in terms of mean price change, at –9.3%, versus –10.4% for the BB1 to BB3 sample. The bottom-tier sample’s edge in that interval was not statistically significant, however.

Through early June, at least, emphasizing the shares of companies with the lowest credit ratings was not a winning strategy overall. It paid off only in the period in which a rally sparked by the Federal Reserve’s declaration of determination to support the credit markets drove investors to the biggest beneficiaries of that policy. Those were shares that suddenly became better risk-reward propositions as the danger receded that they would succumb to financial difficulties during the prevailing recession.

The table below offers corroboration that the results reported in the previous table were not a function of one or two outliers within a modest-sized sample. With 99% confidence, the B3 to CCC3 sample contained a statistically larger number of “extra-base hits.” We define these as stocks that either doubled or tripled during the March 23 to June 8 market rebound.

Note a possibly surprising aspect of the fact that companies with Composite Ratings of B3 and lower benefited the most, as evidenced by their relative stock performance. This occurred even though, as far as speculative-grade credits were concerned, the direct corporate bond purchasing component of the Federal Reserve’s initiative to support the credit market extended only to recent fallen angels, which are generally rated in the BB tier. (The central bank did, however, undertake purchases of high-yield exchange-traded funds containing some issues rated well below BB.)

Impact on equity cushion
The Federal Reserve’s boost to companies with the lowest credit ratings matters to high-yield investors because of its favorable impact on the equity cushions beneath bondholders’ claims. In the 1980s, Drexel Burnham Lambert’s high-yield research department promoted the idea of calculating leverage ratios on the basis of equity market capitalization, rather than accounting-based measures of shareholders’ equity. A related concept, developed by economist Robert Merton, posits that a default occurs when the value of a company’s assets declines to a value no greater than that of its liabilities, that is, when the economic (not book) value equity drops to zero. By this reasoning, an increase in a debtor’s equity value renders the possibility of default more remote.

Drexel’s market-adjusted debt (MAD) ratio was sometimes criticized on the grounds that the market values of creditors’ outstanding shares were likely to shrink at precisely the time that the cushion against bankruptcy was needed most. Support for incorporating market-based equity valuations into credit analysis, though, comes from the long record of empirical validation of Edward Altman’s Z-score model. One of that bankruptcy prediction model’s explanatory variables is Market Value of Equity/Total Liabilities.

Postscript
Shortly after we completed the preceding analysis, stocks sold off sharply on June 11, 2020. Using the same 30-stock samples employed for the periods discussed above, we tested whether bond ratings continued to be a determinant of variance in stock returns during the brief selloff. The table below confirms that they did, once again demonstrating the central role in current stock performance of Federal Reserve support for the credit markets.

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.