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 Great Fallen Angel Menace has been a staple of bond market commentary over the last few years. One observer warns of “a nightmare scenario in which a large swath of investment-grade companies slide into junk status and create a liquidity glut” (see note 1). As another source explains, “[T]he downgrade from investment-grade to junk pushes an immense amount of supply into the market all at once, where there isn't commensurate demand to receive it, leading to sharp price declines” (see note 2).
The pernicious cause-and-effect does not end with price declines, according to a pair of journalists who write, “When bonds become junk, many investment funds are contractually obligated to sell them. Forced sales can set off negative cycles” (see note 3). The domino effect they envision goes as follows:
1. Downgrades push many bonds into the speculative-debt category. 2. Some funds are forced to sell the downgraded bonds. 3. Heightened risk aversion makes investors unwilling to buy the new fallen angels. 4. Companies are unable to finance themselves with new bond sales. 5. Tight money leads to defaults.
In “Fair value update plus fallen angel impact detected” (LCD News, May 19, 2020) we presented evidence that fears about the impact of downgrades from BBB on the high-yield market, while not unfounded, have been overstated. Comparing recent percentage spread-widening on double-B rated deals, the sector most directly affected by downward migrations from investment-grade, with percentage spread-widening on other rating categories, we calculated that new fallen angels had amplified BB widening by all of 18 bps. “The numbers indicate that some dislocation occurred,” we wrote, “but it could hardly be termed massive. It is doubtful that any resulting spillover to the real economy from such a distortion would be very material.”
Even the more dire scenario that we imagined in “How will BBB downgrades affect the high-yield market?” (LCD News, July 11, 2018) is not really all that dire. In the conclusion of that piece we wrote:
Our judgment, based on analysis of two very big surges of the past, is that the likely spread-widening that will be directly attributable to a major downgrading of BBBs — if it happens — is in the range of the lower bound of 50 bps in the 2005 GM/GMAC downgrading and the upper bound of 60 bps in the 2016 Energy downgrading. Note that the greatly feared mass downgrading of BBBs may occur in conjunction with other events that cause additional spread widening. This piece focuses solely on the spread-widening that will result directly and exclusively from a swelling of fallen-angel supply. (Italics added.)
Amid the hullabaloo about the havoc that is supposedly about to be wreaked by downgrades from the BBB tier, which now accounts for half of the investment-grade universe, little has been said about the longer-run benefit to high-yield investors of an increased concentration of the speculative-grade universe in the fallen-angel, or FA, segment. The simple fact is that, by standard performance metrics, fallen angels are much better investments than original-issue, or OI, high-yield bonds. Over the period from January 1997–May 2020, monthly returns on the ICE BofA US High Yield Fallen Angel Index produced a Sharpe ratio of 0.21. The comparable figure for the ICE BofA US High Yield Original Issue Index was just 0.13. (Our proxy for the risk-free return in these calculations was the ICE BofA US 3-Month Treasury Bill Index. The standard deviations were 2.93% for FA and 2.55% for OI.)
Some fans of the high-yield primary market may dismiss the Sharpe ratio as an arcane calculation of interest solely to quantitative geeks who could not trade their way out of a paper bag, as a classic Wall Street slur puts it. Even those who prefer to operate by the seat of their pants, though, can grasp an even starker demonstration that fallen angels represent a superior investment category. These refugees from investment-grade territory have beaten original issues not only on a risk-adjusted basis, as indicated by the comparative Sharpe ratios, but in absolute terms, despite being higher rated on average. From 1997 to May 2020 the respective mean monthly returns were 0.80% and 0.51%.
The chart below traces the history of the average Composite Ratings on the ICE BofA fallen angel and original issue indexes. With the exception of one year (2008) in which the two sectors reached parity at B1, fallen angels have consistently had average ratings in the BB tier, one or two notches above original issues, which have never risen above the single-B tier. In short, original-issue high-yield bonds have confounded financial theory by providing more credit risk and less return.
We can hear some readers reflexively objecting, “The original issues' long-run underperformance was solely a function of the long-ago, now-irrelevant business-plan telecom bust! The predominantly BB fallen angels have outperformed only because BBs as a whole have outperformed!”
Wrong. To begin with, the table below breaks down the 1997–2020 comparative returns into five subperiods. (The last consists of a rump period 1.58 years short of a half-decade.) The FA annualized total return exceeded the OI figures not only over the full period, but in every single subperiod. In one subperiod the gap exceeded 6 percentage points per annum, and in no subperiod was it less than a full percentage point.
Insight into the cause of the original issues' inferior performance emerges from the breakdown of total return into its components. In each subperiod except the first, when the two sectors tied, OI generated higher income than FA. This is hardly surprising, given that OI had a lower Composite Rating in nearly every year, and would therefore be expected to yield more, generally speaking. Indeed, in 85.5% of the months in our observation period, OI's effective yield exceeded FA's. Despite that head start on total return, OI managed to underperform FA in each five-year period by registering a negative price return in all but one. At this point in the current subperiod, FA is on track to post a negative price return for the first time. At –0.68% per annum, however, FA is doing far better than OI, at –2.79%.
In the high-yield market's early days, then-leading underwriter Drexel Burnham Lambert used to deride investment-grade issuers as bureaucracy-bound “dinosaurs” headed for certain extinction. Investors were better advised, said Michael Milken's troops, to own the non-investment-grade market's “rising stars,” scrappy, young, innovative companies imbued with entrepreneurial zeal. The verdict of history is that many of the fallen angels maintained the financial policies of their higher-rated days, while private equity owners transformed numerous rising stars into black holes by overloading them with debt.
As for the notion of fallen-angel outperformance being a function of superior returns by BB issues in general, the table above indicates otherwise. For the full 1997–2020 period, and in all subperiods, FA total return exceeded BB total return. That is to say, FA pulled up BB total returns rather than vice versa. Notably, FA's income advantage was just 31 bps for the full period, yet its total return advantage was a whopping 246 bps. The FA annualized price return of 1.81% contrasted sharply with BB's –0.34%. Even the high-yield market's strongest credits recorded net principal losses over our nearly two-dozen-year observation period, but not so the fallen angels.
Other objections to fallen angels
In years past, managers who regarded OI as the real high-yield market voiced disdain for the fallen angels' lack of meaningful covenant protection. It is true that, because of their comparatively low default risk, IG companies are able to issue bonds without offering the safeguards HY investors consider their due. When the highly rated issuers fall to speculative-grade, their previously floated bonds remain as thinly covenanted as before.
OI's edge in this area has weakened over the years, however, especially when one compares BB fallen angels, which account for 81% of the ICE BofA US High Yield Fallen Angel Index's constituents, with comparably rated high-yield new issues. In May 2020, 52% of new issues rated Ba by Moody’s (and 39% of all issues) were classified as high-yield lite, meaning they lacked restrictions on dividend payments and debt incurrence. Such issues automatically receive Moody’s lowest Covenant Quality score (5.0 on a scale of 1.0 = strongest to 5.0 = weakest). Five issues rated in the B tier received the bottom-of-the-barrel 5.0 CQ score in May.
A final complaint lodged against fallen angels is that their comparatively long maturities expose investors to greater interest rate risk than original issues. It is true that the average maturity in ICE BofA's FA index, at 10.51 years, far exceeds that of its OI index, at 5.50 years. That comparison is not the most relevant consideration for high-yield investors, however.
For one thing, the FA index's average maturity is skewed by a number of 40-year-plus issues. As of May 31, maturities were inside of five years for 39.3% of the bonds and inside of 10 years for 62.6%. (For OI the inside-five-years figure was 47.6%.) At the individual issue level, in short, there are plenty of shorter-dated fallen angels available to investors who are particularly concerned about interest rate risk.
At the asset class level, the average maturity of present fallen angels is not what really matters to high-yield investors. Much more important is the impact on average maturities of adding more of them to the overall high-yield index. By our calculations, the probable impact is far from devastating.
On May 31, 2020, the ICE BofA US High Yield Index’s average maturity stood at 6.28 years, well below the 1997–2020 monthly mean of 7.23 years. That difference is partly explained by the fact that, on a face amount basis, fallen angels represented just 16.2% of the ICE BofA US High Yield Index on May 31, 2020. The 1997–2020 monthly mean FA share is 18.0%.
Since Dec. 31, 2019, though, the face amount of fallen-angel debt has essentially doubled (+98.8%), representing an increase of $109 billion. That last number suggests that a non-trivial portion of the downgrading of BBB bonds that will occur in the current recession has already taken place, with the likes of Ford Motor Co., Kraft Heinz, and Occidental Petroleum Corp. dropping to the BB tier. Having said that, on May 31 a total of $271 billion face amount of debt within the ICE BofA US Corporate BBB Index was priced at an option-adjusted spread, or OAS, of +402 bps or greater. That spread represented the May 31 midpoint between the median spreads on BBB3 and BB1 issues, respectively, +339 bps and +464 bps. In other words, the market had already downgraded $271 billion of BBB bond debt by that date.
The rating agencies are under no obligation to follow the market's lead. Let us suppose for sake of illustration, however, that the face amount of BBB debt remaining to be downgraded in the current recession is $271 billion. Based on May 31 numbers, the new fallen angels would then represent exactly one-sixth (16.7%) of the ICE BofA US High Yield Index, which had a May 31 face value of $1.352 trillion, with an average maturity of 6.28 years.
Let us further assume that the new fallen angels' average maturity will match the 11.52-year average of the pool from which they will be drawn, i.e., the BBB index. The average maturity of the high-yield index, weighted by face amount, would then be 7.16 years. This figure is not perfectly comparable to the high-yield index’s 7.23-year historical average, cited above. Our calculations suggest, however, that under reasonable assumptions the impact of new fallen angels in the remainder of the current recession will be to increase the high-yield index's average maturity from below normal to about normal. That scenario does not qualify as an apocalypse, in our judgment.
Note that high-yield volatility derives from both interest rate fluctuations and changes in the spread-versus-Treasuries. Over the December 1997–May 2020 period the ICE BofA US High Yield Fallen Angel Index’s monthly standard deviation of total returns was 2.93%, versus 2.59% for the ICE BofA US High Yield Index. Accordingly, if $271 billion of fallen angel debt is added to the present $1.352 trillion face amount of the present high-yield universe, a very minor increase in the high-yield standard deviation can be expected, likely to a little more than 2.60%. This outcome, too, constitutes something less than a calamity.
Over the past few years, investors have been conditioned to expect the worst from the coming wave of fallen angels. It does appear likely that the new supply will exert some near-term pressure on high-yield spreads. Over the slightly longer term, however, the evidence indicates that the high-yield universe will be improved by the influx of former BBBs, both in terms of credit quality, and in terms of absolute and risk-adjusted total return.
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. Jeff Cox, “‘Fallen angels’ are posing the biggest threat to the bond market this year,” cnbc.com (Jan. 22, 1999).
2. Evan Kurinsky, “Fallen angel risk shifts back into focus,” wilmingtontrust.com (April 2, 2020).
3. Sarah Slobin and Feilding Cage, with illustrations by Eric Palma, “Falling angels?” graphics.reuters.com (April 3, 2020).