Synopsis: The energy sector has widened by more than 200 bps since June and is heavily overrepresented in the distressed universe. This adversity may offer some opportunity. Meanwhile, a sharp decline in Caa issuance in October masked improvement in market-wide covenant quality.
The steep drop in oil prices is currently a leading topic of discussion in the financial markets. Since June, Brent crude futures have fallen sharply. How has the high-yield bond market been affected and what are the investment implications?
Energy and spread-widening
The BofA Merrill Lynch US High Yield Energy Index currently accounts for 15% of the market value of the BofA Merrill Lynch US High Yield Index. To the extent that Energy’s option-adjusted spread (OAS) has widened by more than the OAS of other industries, the magnitude of general weakening in high-yield pricing could be overstated. A drop in oil prices produces benefits in the non-energy-producing portion of the economy (although not as much as in previous decades). At the extreme, the observed widening in the high-yield spread since midyear might be entirely attributable to widening in Energy. In that case, the increase in the high-yield risk premium could be giving a distorted picture of the overall rise in default risk and reduction in secondary-market liquidity.
Below are the actual numbers. Our base date is June 30, this year’s low point in month-end OAS on the BofAML High Yield Index (H0A0). Furthermore, as noted above, the plunge in oil prices began in June. In the table, H0EN is the BofAML HY Energy Index. Using the weighted-spreads method, we solve for the spread on the Non-Energy portion of the high-yield index on June 30 and Nov. 21.
Energy was nearly even-yield with the Non-Energy segment on June 30. Through Nov. 21, Energy widened by 203 bps. That is a much greater widening than the rest of the index experienced. Still, the numbers show that the rest of the high-yield universe widened by 83 bps, from 354 bps to 437 bps. The actual widening, including the Energy effect, was 101 bps, from 353 bps to 454 bps. In short, high-yield spreads have widened materially since midyear, even leaving aside the weakening of Energy issues, which to some extent operate in a universe of their own.
Energy and distress
As of Nov. 21, there were 180 issues in the BofAML High Yield Index with option-adjusted spreads of 1,000 bps or more, the widely accepted threshold for defining distress. That represents 7.88% of the index’s total issues, a figure known as the distress ratio. Of the 180 total distressed bonds, 52, or 28.89%, are Energy issues. That means Energy is heavily overrepresented in the distressed segment, given its 16.32% share of the BofAML High Yield Index by number of issues (373 divided by 2,285).
As our previous research found (see “How to tell when distressed bonds are attractive,” Nov. 28, 2012), the expected 12-month default rate on distressed issues under present conditions (the expected rate is partly a function of the level of the distress ratio) is 32.86%. This suggests that the market is currently expecting a 2.58% default rate (7.88% x 32.86%) on all U.S. high-yield bonds through November 2015. The full-year 2015 forecasts of S&P and Moody’s are 2.7% and 2.8%, respectively, putting the market essentially in line with the rating agencies – at least for the time being. Between June 30 and Nov. 21, the percentage of Energy issues with option-adjusted spreads of 1,000 bps or more jumped from 2.30% to 13.94%. The market-implied default rate for Energy issues therefore is 4.58% (13.94% x 32.86%).
For the Non-Energy portion of the high-yield universe, the market-implied default rate is just 1.84% (5.60% x 32.86%). Given the idiosyncratic distress in Energy, the possibility exists that the overall default rate will surge without a downturn in the economy. Default-rate peaks in the 10%-plus range historically have been associated with recessions. Following a sharp break in oil prices in the mid-1980s, however, Moody’s U.S. percentage-of-issuers default rate jumped from 3.53% in April 1986 to 6.56% one year later. As the wave of oil patch failures subsided, the default rate declined to a low of 2.16% in May 1989 before escalating during the 1990-1991 recession, with a peak at 12.36% in June 1991.
In short, it would not be unprecedented for Energy to cause the default-rate trend to deviate from the path dictated by general economic trends. Investors therefore should monitor the default rate closely in coming months. To the extent that escalation mainly reflects defaults in the Energy sector, investors should not take the escalation as a sign that a double-digit peak is in store as early as 2015 or 2016. We currently expect the default rate to reach the historical average of around 4.5% in 2016 and to peak in double digits a year or two later.
Does distress spell opportunity?
As we pointed out in “October rebound leaves CCCs behind” (Nov. 4, 2014), the Energy subindex is currently trading very cheap to its ratings despite having net positive ratings prospects. This is a rare circumstance and may constitute an excellent opportunity. It all depends on whether the rating agencies are correct in continuing to see the industry’s credit quality as improving, on balance, despite the recent drop in oil prices.
Note that the speculative-grade universe has a natural-gas orientation that mitigates the impact of falling oil prices. A reader from Natixis reports that at that firm’s recent U.S. energy seminar, gas-biased producers were not panicking, but were even hoping eventually to pick up some cheap assets as a result of the general pressure on Energy companies. Further declines in oil prices could change the picture, but the companies indicated that their production costs are in the vicinity of $45 per barrel of oil equivalent. The companies say they are not making deep cuts in their exploration and production budgets, but are shifting emphasis somewhat to their most-likely-to-succeed projects.
Ratings mix distorted covenant quality trend in October
As detailed in “Covenant quality decline reexamined” (Oct. 1, 2013), we maintain a “shadow covenant quality index” that parallels the series published by Moody’s. In brief, Moody’s rates each major covenant of each high-yield new issue on a scale of 1 (strongest) to 5 (weakest). The agency combines the individual-covenant scores into an issue score, then reports the monthly and quarterly average scores for all issues to represent the market-wide trend in covenant quality.
As the dotted line on the chart below indicates, Moody’s reported a deterioration in covenant quality from 3.81 in August to 4.43 in September, then showed quality nearly unchanged at 4.41 in October. On closer examination, however, October’s nearly worst-ever score resulted from Caa issuance falling off a cliff. Just 4.16% of the month’s issues were rated in the bottom-quality tier, down from 15.56% in September and marking the lowest reading since September 2011. (The 2014 year-to-date range for Caa concentration, excluding October, is 11.1-39.30%.) This matters because covenant quality increases with each step down the rating scale, as investors demand greater covenant protection to offset increasing default risk.
The average covenant quality score of Caa issues actually improved sharply in October, to 2.72, from 4.05 in September. The average Moody’s score across all ratings deteriorated, however, because the Caa issues represented such a small portion of total issuance. FridsonVision’s version of the covenant-quality index eliminates such distortion by holding the ratings mix constant over time, thereby producing apples-to-apples comparisons from month to month. As the solid line in the chart shows, by our measure covenant quality improved sharply in October, to 4.00.
Martin Fridson, CFA
Chief Investment Officer
Lehmann Livian Fridson Advisors LLC
Research assistance by Kai Chen