Is There a Bubble in High-Yield Bonds? No!
“The media loves bubbles, but there is no universally accepted definition of a bubble, and some economists even deny bubbles exist,” said Martin Fridson, CFA, CEO of FridsonVision LLC, at the Fixed-Income Management 2012 conference in San Francisco last week. “I think if it looks like a bubble and walks like a bubble and pops like a bubble, it is probably a bubble,” Fridson said.
But that does not mean that the high-yield bond market is currently a bubble. Fridson took the audience through some typical arguments pundits have recently used to call the high-yield market a bubble. None of these arguments hold up under closer examination and scrutiny.
The Yield on High Yield Is So Low
Fridson looked at the yield-to-worst (YTW) — the most conservative measure of yield — and found that today’s YTW for the BoA Merrill Lynch US High Yield Master II Index was 6.45% the week of 8 October 2012, after reaching an all-time low of 6.20% on 18 September 2012. Yes, this is far from the long-term average of 10.09% from 1997 (when YTW measures were first taken) to 2011. But, Fridson said, “you have to look at equivalent Treasury yields during the period. It was really Treasuries that went to the extreme, not high-yield bonds.” The mean five-year Treasury was 3.89% over the period, and at today’s 0.65%, it is 2.1 standard deviations away from the average. High-yield bonds, on the other hand, were only 1.4 standard deviations from the mean at the low in September. “The yield on high-yield bonds is low because of the exceptionally low-interest-rate environment we’re experiencing,” said Fridson.
Spreads Are Too Tight
Again, Fridson examined historical statistics and found that spreads today are only 1/10 of a standard deviation narrower than the long-term average spread. That being said, Fridson continued, the average spread is the correct spread only when conditions are average. Using Fridson’s econometric model, which includes performance of the economy, current default rates, treasury yields (inversely correlated), and a measure of credit availability, he found the spread should currently be around 680 basis points — today it is 566. That 114-basis-points difference is not too far from the long-term standard deviation of 132.
The notion of spreads being too tight could have more weight in the argument if the risk premiums were defying the fundamentals — but they’re not. The default rate hit a cyclical low in 2011 (and has moved up slightly since then to 2.59% today) and is projected to rise in 2013, but we’re still well below the historical average of 4.5%. Default risk is certainly not the only thing that matters, but the overall level of risk in the markets is not implying spreads should be wider.
“I wouldn’t argue that high-yield bonds are cheap,” Fridson said, “but it’s not a dire situation.”
Record-Breaking New Issuance in High-Yield Bonds
With all the new issuance, argue the pundits, Wall Street must be making a killing and leaving clients holding the bag. “In reality,” said Fridson, “you need to look at the net increase in outstanding issues which is much less than new issuance.” (Year-to-date net growth in outstandings is currently 35% of the total new issuance in dollar terms.)
This relationship is driven in part by corporate responses to recent developments. Companies do not want to get themselves in the position they were in 2008, when they could not roll over their debt. So today, they are retiring shorter-term paper and lengthening maturities, or using spare cash to fully retire debt. In many cases, companies are coming out ahead by buying back debt more cheaply than when it was issued. “The surprising conclusion is that there is record breaking new issuance AND a shortage of supply!” said Fridson.
The Equation Often Used by Pundits to Project Total Return is Just Plain Wrong
Continuing on his theme of sound analytics, Fridson talked about the “total return” equation — a frequently used metric for expected returns and relative value that can be dangerously misleading. This misused equation starts with the beginning yield on high-yield debt, subtracts the default rate, and adds back the recovery rate.
This fallacious equation misses many moving parts including:
- Treasury rates that go up or down.
- Changes in the level of spreads vs Treasuries, which can be dramatic.
- Changes in the composition of the index. (On average, 25% of the issues in the High Yield Master II Index disappear each year!)
- Company tenders and redemptions.
- Beginning of year discount on defaulting bonds. (It’s highly unlikely they were trading at par at the beginning of the year.)
In 2009, this equation would have projected a 12.6% total return, when the actual turned out to be 57.5%! The analyst who predicted that outcome would have been fired! Of course, the Fed’s extreme monetary policy actions front-loaded much of the return in high-yield bonds that year. But over a 15-year period, the difference between the fallacious total return projection and the actual return was never less than 2.36 percentage points! Even that number is substantial in the context of managers’ success in outperforming benchmarks “Why use an equation to predict the future if you can’t even use it explain the past?” Fridson asked.
In short, “quantitative analysis can help you manage risk and identify bubble territory,” Fridson said, “but you must apply it appropriately.” Fridson also noted the high-yield bond market is driven by shocks, discontinuities, and lurches — things that are not predictable or foreseeable. Returns are not steady or smooth, and there is considerable volatility. Current yield and default losses just don’t capture it well. Investors should focus on the risk/reward tradeoff when assessing high-yield bond investments. Sound analytics, thorough fundamental analysis, and careful security selection is the only way to provide the alpha investors are seeking.
Please note that the content of this site should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute.