Junk Bonds
New issuers flood the market
Junk bond issuance is again hitting record levels, both for seasoned and new issuers. Through the end of July, US corporations have raised over $300bn in the non-investment grade bond market, a 30% increase from last year’s record level (also through July). Of this year’s total, $68bn was raised by first-time issuers, an increase of 50% from the prior year. Average credit spreads have recently widened a bit, reflecting the increase in volume and lower average credit quality. Average credit spreads (option-adjusted) are now around 3.4%, up from 3% or so earlier this year. The average yield on bonds issued by first time issuers is 4.6%, a 10-year low. Refinancing activity has also been significant, with many issuers saving 100-200bp in annal interest expense.
In thinking about the junk bond market, keep in mind that bond “yields” are not the same thing as “expected returns”. If a junk bond portfolio has an average yield of 4.6%, this is the return (IRR) the investor will receive only if it holds all its bonds to maturity, the interim cash flows (interest payments) are reinvested at the same yield of 4.6%, none of the bonds are called early, and none of the bonds incur losses on default. These are all somewhat artificial assumptions for many high yield bond portfolios, of course. Portfolio managers do trade bonds, sometimes frequently, many bond issues are called (especially in a falling interest rate environment), and interim cash flows cannot always be reinvested at the same rate. And high yield bonds do default from time to time, sometimes in large number and with large losses on default. This non-trivial risk of default is what makes these bonds “high yield”, or less charitably “junk”.
Junk bond investors demand a higher yield to compensate them for the expected loss from default and for the risk (uncertainty) associated with future credit events. If a bond portfolio yields 4.6% and the expected annual loss from default is say 1%, then the expected return on the portfolio would be only 3.6% net of expected losses from default. (In the current market, high yield investors may not anticipate this level of loss from default, at least not in the near term.) The realized return may be quite different from this expected return of course, but the key point here is that “yield” does not equal “expected return”, at least not in the junk bond market. And the difference can be quite significant (and volatile), particularly in difficult economic times.
One other point worthy of note here is the relationship between risk-free rates (UST yields), junk bond yields and expected equity returns. UST notes currently yield about 0.80% for a 5-year maturity, 1.0% for a 7-year maturity, and 1.25% for a 10-year maturity. As discussed, the expected return on a portfolio of high yield bonds at an average yield of 4.6% may be only 3.6% or so. (Expected loss on default is not directly observable in the market.). But what about the expected return on equities? Presumably it must be higher than the expected return on high yield bonds, and much higher than the expected return on USTs, but how much higher?
The expected future return on equities is typically estimated using one of several methods: by extrapolating from historical (past) realized returns, by conducting surveys of CFOs and investors, or by interpolating the expected future return based on the current level of security prices and equity valuations. Historically the US stock market (S&P 500) generated average annual returns of close to 10% (depending on the time frame), but with a wide distribution of annual (calendar year) returns, ranging from negative 35% in 2008 to positive 30% in 2019. But these are historical realized returns, not historical expected returns, and these returns were generated when risk-free interest rates (UST yields) were much higher than they are currently. So even if the past expected return was in fact 10% or so, the historical “equity risk premium” (ie the difference between equity returns and risk-free UST returns) was perhaps only 5% or so. And even if we think that this estimated historical equity risk premium is still reasonable to use in today’s much lower-interest rate environment (a questionable assumption in my view), this would suggest that today’s expected equity return would be in the range of 6.25% or so. (I calculated this using CAPM, with a UST 10-year risk-free rate of 1.25% and a 5% ERP).
Now let’s put this all together. In the current market, short-term UST bills yield 0% and 10-year UST bonds yield 1.25%. High yield bonds yield 4.5% or so (on average), with expected returns of perhaps 3.5-4% net of expected default losses. And equities yield 1% or less (the dividend yield on the S&P), with expected total returns of perhaps 5-6% or so. The current expected return on US equities is much lower than historical realized equity returns, but this makes sense given the recent exceptional performance of equities. (It is reasonable to think that equity returns will revert to the mean at some point.). It also makes sense given the high current valuation of the US equity market. (For any given set of cash flows, a higher current valuation implies lower future returns.) But even at 5%, the expected return on equities reflects a healthy premium to both risk-free rates (UST bond yields) and also to the expected return on high yield (junk) bonds.
Note that these are all nominal, not real, returns. Which of course raises the question of inflation, which is currently running at something over 2% here in the US. How much over is a very interesting question, which I would like to come back to in a subsequent post.
What is my confidence level in these future expected returns? Very low. But I think it is fair to say that we are in an extremely high valuation and low-return market environment. In my view the current market is flashing warnings signs of highly asymmetric risks and returns: ie high risk (due to high valuations) with low (expected) returns. You’ve all heard of the concept of the “risk-free return”, widely used in CAPM and other security pricing models. Well, I think today we need to spend some more time thinking not about “risk-free returns” but instead about the implications of investing in a world with “return-free risk”. And today’s high yield bond market is just one good example of this.
For more on these topics, see the links below.
Links
First-Time Issuers Fuel Junk Bond Rally, WSJ 8.20.2021
Back to Junk, Banking and Beyond, 7.11.2021
High Yield Bonds Today, Howard Marks Oaktree Capital 2013
30-Years of Stock Market Returns, The Balance
Daily UST Yield Curve, US Dept of Treasury 8.20.2021
Yes, but when? I recall Henry Kaufman (Salomon Brothers in the early 80's) saying that he could forecast rates or timing, but not both. I've been predicting this correction for years now and have been consistently wrong. I'd like to say I've been "early" but of course in the financial markets "early" and "wrong" are often the same thing. No one is impressed by the accuracy of the stopped clock, which is right twice a day (or once a day in the 24-hour digital world).
Excellent analysis as usual. The conundrum remains, “that’s interesting and I agree, but I (as a fund manager / private investor / banker etc) need to invest, so I must pick the best of the bad choices”. And the question, “how long can this go on and what will be the consequences?” I think we’re beginning to see the answer to the latter in the form of planned reduced FRS stimulus, which will be followed by higher interest rates and equity markets sell off. Then, inevitably, as a consequence of sustained fiscal and monetary stimulus (the former embedding largely unfunded structural programs that will continue to expand), higher inflation. Then higher rates, further corrections especially in the debt markets, and the usual wailing and gnashing of teeth and expressions of surprise that should surprise no one. An imbalanced system will ultimately adjust in some form, and the longer the imbalance continues the more severe the adjustment. Paul Volcker II anyone?