How bad was the high-yield market’s pre-election sell-off?
From Oct. 25 from Nov. 4 the option-adjusted spread (OAS) on the BofA Merrill Lynch US High Yield Index widened by 60 bps, from 460 to 520. That ranks just within the first decile of eight-day widenings since the inception of OAS on Dec. 31, 2016. The 60 bps swing also ranks within the top 10% of all moves, wider or narrower.
Note, however, many eight-day widenings during the Global Financial Crisis were far bigger. In all, there were 22 eight-day widenings of 200 bps or more during 2008’s chaos, topped by a move of 438 bps. The biggest eight-day tightening was negative 336 bps, which occurred in the eight-day span ending Jan. 7, 2009.
By labeling the recent move the “pre-election” sell-off we imply—intentionally so—that the tightening presidential race produced a risk-off response. The most widespread interpretation is that Hillary Clinton, who until recently held a commanding lead in the polls, is considered the more predictable of the two main candidates. The nature of a Donald Trump administration is thought to be more uncertain.
Certainly, one could dispute that explanation. As always, other events occurred that could have been responsible for the high-yield downturn. Most prominently, crude oil prices fell by 12% between Oct. 25 and Nov. 4, as measured by the Generic 1st Crude Oil, West Texas Intermediate contract. Over the past two years, to be sure, major swings in energy prices have triggered outsized moves in the high-yield market’s Energy component.
Those industry-specific jumps have resulted in substantial moves in the high-yield index as a whole. From Oct. 25 to Nov. 4, however, the BofA Merrill Lynch US High Yield Energy Index widened almost exactly in line with the BAML High Yield Index as a whole (62 bps and 60 bps, respectively). Based on the fact that Energy essentially performed no worse than the rest of the high-yield universe, we reject the hypothesis that falling crude prices precipitated the sell-off in the last few days before the election.
There may be more to the story, though, than apprehensiveness over a Trump administration. Another possibility is that the market is attributing some probability to a hung election. In the near term, that outcome would present major uncertainty in its own right.
What if the election ends with no winner?
Prior to early November, as far as I have been able to establish, there was little or no discussion of the possibility of the winner of the presidential election remaining unresolved beyond early tomorrow morning. On Nov. 6, though, Fox News discussed the mathematically possible—if statistically improbable—outcome of an exact tie in the Electoral College, at 269–269.
Barring the deadlock being broken by a faithless elector, the decision would then be thrown to the House of Representatives. Perhaps a more easily imaginable scenario is a replay of the 2000 election, when disputed ballots in one state kept determination of the winner on hold for over a month.
Between Election Day, Nov. 7, 2000 and Dec. 12, 2000, when the Supreme Court voted 5-4 to leave intact a Florida vote count that narrowly awarded the state, and consequently the election, to George W. Bush, the BAML High Yield Index’s OAS widened by 119 bps, from 775 to 894. In-between, the spread reached a maximum of 913 bps on Dec. 7.
Was the Nov. 7–Dec. 12 widening, which corresponded to a –2.26% (–21.23% annualized) total return, a hung-election phenomenon that could serve as a guide of what to expect if today’s vote proves inconclusive for the time being? To be sure, the high-yield spread was already on the rise in November 2000. It had ended 1999 at 476 bps, meaning it was up by 299 bps in the year to date on Election Day.
The day before Election Day Bloomberg BusinessWeek reported that the default rate stood at 3.83%, up from 1% in 1997. November 2000 was also noteworthy for precipitous drops in Technology and Telecom equities, with some stocks losing about half their value.
The chart below shows, however, that November’s spread-widening represented a spike, a conspicuous discontinuity in the steady rise in the spread that reached its apogee in late 2002. In November 2000 the BAML index suffered its worst return (–3.84%) since August 1998’s –5.05%, which was a response to Russia’s sovereign default. The November 2000 sell-off was seen as excessive relative to the fundamentals, as evidenced by December reports that Warren Buffett and other value investors were snapping up depressed high-yield bonds.
To summarize, no two market circumstances are exactly identical. Measuring the impact of an event such as a hung election is inevitably complicated by other, concurrent events. Without downplaying these limitations, we submit that the high-yield widening of 119 bps in the period of uncertainty that followed the Nov. 7, 2000 presidential vote represents the best available estimate of the likely impact of a replay this year.
Investors should keep in mind, though, that the Oct. 25–Nov. 4 widening of 60 bps may indicate that the market has discounted half of the impact in advance. (By contrast, the market was probably caught off-guard by 2000’s unusual turn of events, implying that previous spread-widening was strictly a function of high-yield fundamentals.) – Martin Fridson
Research assistance by Yanzhe Yang and Jiajun Wang.
Marty Fridson, Chief Investment Officer of Lehmann Livian Fridson Advisors LLC, is a contributing analyst to S&P Global Market Intelligence. His weekly leveraged finance commentary appears exclusively on LCD, an offering of S&P Global Market Intelligence. Marty can be reached at [email protected]