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.
Setting the stage: Public vs. private
In previous analyses in this space we reported that, in the severe market downturn of March 2020, bonds within the ICE BofA CCC & Lower US High Yield Index that were issued by private companies outperformed bonds issued by public companies.
The 138 private-issuer bonds delivered a higher unweighted average return than the 125 public-issuer bonds, -19.70% versus -37.13%, with a lower standard deviation, 19.44% versus 26.21%. With 99% confidence, the difference in mean returns was statistically significant.
We surmised that many practitioners would have expected exactly the opposite outcome. For several reasons, they likely considered bonds issued by private companies riskier than those issued by public companies. This implied that the private-issuer bonds would be unloaded more aggressively by investors when economic or market hazards escalated:
While issuance by privately held companies is an accepted feature of high-yield investing, portfolio managers historically have perceived certain disadvantages of this feature of their market. Standard Securities and Exchange Commission filings such as Forms 10-K and 10-Q are generally available on private-issuer bonds (note 1), but companies with public shareholders are considered likely to communicate more extensively with investors, going beyond regulatory requirements. Further enhancing the information flow on public companies is the research published on them by equity analysts. An additional benefit of owning a bond of a company with publicly traded stock is that a movement in the share price may signal a material change at the issuer before the high-yield market begins to reflect it. Finally, public companies were traditionally thought to have readier access to new equity capital if needed, although growth in the private equity business in recent decades has lessened the public companies’ edge.
Based on these reasons to regard private-issuer bonds as the riskier of the two varieties, we investigated possible confounding factors. That analysis identified a substantial difference in term risk between the two classes of bonds. Excluding Energy issues, which grossly underperformed the rest of the CCC & Lower segment in March 2020, private issuers had an effective duration of 3.26 versus 2.83 for the public issuers. The private issuers consequently enjoyed a greater benefit from the halving of the ICE BofA US Treasury Index’s effective yield, from 1.11% to 0.56%, during the March sell-off.
Our study of March 2020 returns provided no evidence that private-issuer bonds offer superior downside protection that is not a function of the present difference in mean effective duration between the two categories. By the same token, we have uncovered no evidence that private-issuer bonds are worse assets to hold in a downturn than public-issuer bonds.
Response to our previous research
Our preliminary finding of superior bear-market performance by private-issuer bonds could be interpreted as a thumbs-up for the private equity industry. Leveraged buyout, or LBO, organizers might cite the raw results in seeking favorable rates on their debt financing. They could make the valid point that investors should be willing to sacrifice some upfront yield to obtain better downside protection than public-issuer bonds offer.
For our part, we would have been delighted if the preliminary finding had stood up under testing for confounding variables. We would have thus identified a strategy that high-yield managers could profitably deploy if they correctly anticipated a major market decline. If the data also helped private equity firms obtain lower capital costs, based on providing investors a bona fide benefit, so much the better.
When we threw cold water on this idea by attributing the lower March 2020 returns of public-issuer bonds to their shorter effective duration, we received some thoughtful responses from fans of private-issuer bonds. One claim we heard was that, thanks to growth and development of the private equity market, privately owned companies now actually enjoy better access to capital than their public counterparts if they experience temporary financial distress. Another assertion we heard was that CCC public issuers tend to be deteriorating credits that have been downgraded from higher rating levels since issuing their bonds. In contrast, this argument went, CCC private-issuer bonds generally began life that way and represent fundamentally sound companies rated in the lowest tier only because they opt for high leverage ratios in order to maximize return.
For the present, we will not undertake to test these claims, nor to calculate comparative default rates on private-issuer and public-issuer bonds. Our aim at this point is just to push further on the question of whether high-yield investors can expect to obtain alpha by shifting toward private-issuer bonds in advance of a major downturn. The null hypothesis is that such a strategy offers no advantage because our preliminary finding of outperformance by private-issuer bonds in March 2020 truly was an artifact of their comparatively long duration.
Our research objective was to determine conclusively whether private-company bonds outperformed in the March 2020 sell-off specifically as a function of some inherent advantage conferred upon bondholders by the issuers’ private ownership form. We approached the problem through “match play.” This consisted of creating a matched sample, pairing public-company and private-company bonds and eliminating as fully as possible performance differentiators other than issuer’s ownership form and idiosyncratic returns. A finding that the private-issuer bond in most pairs delivered a higher return in March 2020 would uphold a strategy of emphasizing private-company issues to reduce short-run, downside risk.
Our initial sample consisted of the 263 bonds in the ICE BofA CCC & Lower US High Yield Index as of March 2020. (We previously found a high concentration of private-issuer bonds in this rating category. As detailed earlier, private-company bonds constituted a slight majority of the bottom-tier index’s issues during our observation period.) For our final sample, we minimized confounding factors by selecting only bonds of U.S. and Canadian companies and eliminating upstream Energy (exploration/production and equipment & services) issues, which drastically underperformed the CCC & Lower index in March 2020 due to a plunge in oil prices (see note 2). Our initial sample was further reduced by bonds for which no match was available on our three criteria: Identical Bloomberg Composite Rating, identical capital structure priority, and close-to-equivalent effective duration.
Our final sample consisted of 36 matched pairs, displayed in the table later in this piece. The median, absolute value difference in effective duration was 0.0515, in a range of 0.001 to 0.999 (mean = 0.088). Given the large total return differences in most pairs of issues, as well as the negligible differences in effective duration in the two cases of total return near-ties, we do not believe the absence of 0.000-level effective duration matches cast any doubt on our conclusions. Although only two of the 36 pairs were matched on industry, we believe that excluding upstream Energy issues from the analysis eliminated the possibility of our conclusions being compromised by industry factors.
The competition between public-issuer and private-issuer bonds ended in a perfect tie. Among the 36 matched pairs, 18 public-company issues produced higher returns in March 2020 than their private-company peers and 18 private-company issues beat their public-company counterparts. The absence of a detectable advantage for public-issuer or private-issuer bonds cannot be explained away by the presence in the sample of multiple issues of certain issuers.
As much as private equity firms might wish the results to have come out differently, and as much as a different outcome might have pointed to a useful strategy for limiting downside protection, our analysis found no short-term performance advantage in down markets for private-issuer bonds. Fans of private-company bonds can take some solace from the fact that we found no performance disadvantage, either. When we used “match play” to normalize the two categories’ comparative returns for differences in effective duration, no bias in either direction emerged from the data.
Our findings do not preclude the possibility that one category’s long-run credit performance, measured by default losses and rating changes, is better than the other’s. Such an edge could arise from superior access to capital under duress, with plausible conjectures on that point existing for both forms of ownership. If such a bias should turn out to exist, measured on a rating-for-rating basis, then investors who overweight the category with superior credit characteristics may realize a long-run return advantage. Reallocating from one category to the other in anticipation of a major market downturn, however, will likely contribute nothing except added transaction costs.
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.
1. This was not always true. When LBOs first began to figure prominently in high-yield bond issuance, some issuers took advantage of a loophole in SEC rules whereby they were exempted from filing 10-K and 10-Q reports if they had fewer than 200 holders of their securities. After issuing registration statements and thereby creating the impression that they intended to provide ongoing disclosure, the companies stopped filing statements after the first one or two. Outraged investors prevailed on the investment banks to underwrite only deals in which the issuer covenanted to provide 10-Ks and 10-Qs or the equivalent.
2. The Generic 1st Crude Oil, West Texas Intermediate contract plunged from $44.76 on Feb. 28, 2020, to $20.48 on March 31, 2020. In March 2020 the ICE BofA US High Yield Energy Index returned -33.774% versus -11.759% for the all-industry ICE BofA US High Yield Index.