In this Friday, Dec. 5, 2014, file photo, specialist Charles Solomon works at his post on the floor of the New York Stock Exchange.Richard Drew/The Associated Press
The rout in junk bonds is intensifying and there's blood in the water. After claiming some high- profile casualties - notably Third Avenue Management - the turmoil is raising fears of a larger meltdown in the markets, perhaps even a recession. In other words, is high-yield debt the canary in the credit mine?
Academic work on the subject suggests that the difference in the rates for high-risk debt and rock-solid government securities - the so-called risk premium, or high-yield spread - often is a significant harbinger.
A paper published in the Oxford Review of Economic Policy in 1999 concluded that the high-yield spread "outperforms other leading financial indicators," such as the term spread and the federal funds rate. An International Monetary Fund staff paper published in 2003 offered a similar assessment, but added that "abnormally high levels of the high-yield spread have significant short-term predictive power."
The trouble with these findings is that the pool of data is focused on very recent financial history, which makes it harder to draw broad conclusions. This limitation reflects the conventional wisdom that junk bonds are a recent invention cooked up by the likes of Michael Milken and company, and as a consequence, there are no comparable data sets before the late 1980s.
But several economists at Rutgers - Peter Basile, John Landon-Lane and Hugh Rockoff - recently disputed that conclusion in an intriguing working paper that resurrected neglected data on high-yield securities from 1910 to 1955.
These forerunners of today's junk bonds initially merited Aaa or Baa ratings, but lost their appeal once they were downgraded to Ba or worse. Such speculative-grade bonds constituted, on average, approximately a quarter of the total book value of outstanding bonds before the end of World War II.
The authors of the study argue that although other kinds of spreads also have predictive power, "junk bonds may be a more sensitive indicator, perhaps a more sensitive leading indicator, of economic conditions than higher-grade bonds."
While their research opens all sorts of avenues for academic exploration - Was there a decline in lending standards in the late 1920s? Was there a liquidity trap in the late 1930s? - the most intriguing question it raises is about the predictive power of the spread between high-yield and high-quality debt.
In theory, this power to predict turning points in the business cycle could manifest in two ways. The first would be a narrowing of the spread, which would mean that investors recognized the worst was over, a trough was imminent and a rebound was in offing.
The second would be a spike in the spread, which means that investors anticipated that the economy had peaked and that a contraction was in the offing.
The authors found that a narrowing of the high-yield spread predicted a mere three of 10 troughs. But spikes were another matter: Exceptional bumps in the high-yield spread accurately predicted eight of 10 peaks (and the subsequent declines, most notably the downturn that began in August 1929 and turned into the Great Depression, as well as the recession that began in 1937 after the Fed prematurely hiked rates).
It may be too early to read too much into these findings. But when combined with the research on the predictive capacity of the high-yield spread from the 1980s onward, this recent work suggests that the spread is a leading indicator worth watching. And given the recent spike, something far worse than a junk bond meltdown may be brewing.
How bad? In the three months before August 1929, the high- yield spread spiked by 47 basis points, and in the three months before May 1937, it shot up 85 basis points. In the past six weeks of 2015, it has spiked by about 120 basis points.
That doesn't mean we're headed for disaster: There's still an apples-and-oranges quality to comparisons of the two eras, despite the best efforts to create a commensurate set of data. But if the spread continues to widen, another downturn - or worse - could be ahead.