Some investors are worried that with interest rates lower than they were in previous Fed tightenings, high-yield bonds are more sensitive to rate movements and therefore sure to fall more sharply this time around. These fears are misplaced. The record shows that the true driver of high-yield price movements is the extent to which the Fed hike at the short end of the yield curve translates into yield rises further out on the maturity spectrum.
Focus should be price return, not spread changes
The big question facing non-investment-grade investors is, “How will the high-yield market respond when the Fed raises short-term interest rates?” Sellside analysts have provided helpful input on this question in terms of the change in spreads that can be expected when the Federal Open Market Committee begins to tighten (probably sometime in 2015). The most direct concern for performance-conscious investors, however, is what will happen to returns.
In forming a judgment on that issue, we have two precedents in the modern history of the high-yield market to consider, namely, hikes centered on 1994 and launched in 2004 (see note 1). First, according to the Federal Reserve Bank of St. Louis, the Fed funds rate rose monotonically from 2.96% on Dec. 1, 1993 to 6.05% on April 1, 1995. Second, the rate rose monotonically from 1.00% on May 1, 2004 to 5.26% on July 1, 2007. The full extent of the Fed funds hike was 309 bps in the “1994” episode and 426 bps in the “2004” episode.
This report focuses specifically on an assertion regarding the present cycle that we have heard from some investors. Given that interest rates are lower than in the previous cycles, they argue, the sensitivity of high-yield bond prices to interest-rate movements has increased. Therefore, they conclude, investors must assume high-yield returns will be hit harder this time than in the past.
We concentrate particularly on the price return of the BofA Merrill Lynch US High Yield Index. Total return also depends on the current income on the index, but that will not be affected in the short run by a rate hike. For investors, the key issue is how the index’s price will respond. (Note that default rates did not greatly affect past outcomes. The Moody’s U.S. speculative-grade default rate declined in both 1994 and 2004 and remained well below historical averages in both 1995-1997 and 2005-2007.)
Factors that did not drive past high-yield price returns
The two historical precedents present a sharp contrast in price performance. At annualized rates, the BofAML High Yield Index’s price returns were as follows:
- Dec. 1, 1993 to April 1, 1995: -5.03%
- May 1, 2005 to July 1, 2007: +0.26%
If the abovementioned assertion about the impact of today’s lower rates is correct, we should find that the difference in these two periods’ price returns is a function of differences in yield on high-yield bonds. The table below addresses this question. We use yield-to-maturity due to the unavailability of the preferable yield-to-worst measure in 1993-1995. At the outset of the earlier period, in which prices fell more sharply, the index’s yield was higher, at 9.97%, than at the outset of the later period, when it stood at 8.08%. In short, the historical record gives us no reason to suppose that with the BofAML High Yield Index’s Sept. 26, 2014 YTM lower still, at 6.68%, price vulnerability is greater now than in 1994 or 2004.
We can also rule out spread-versus-Treasuries (SVT) as the differentiator between the outcomes of the two historical precedents. In lieu of the unavailable option-adjusted spread, we use the YTM difference between
high-yield and the BofA Merrill Lynch Current 5-Year US Treasury Index. Bears might suppose that the tighter the spread, the more the high-yield market is bound to fall when rates rise. In fact, though, the SVT was somewhat narrower (447 bps) in the 1993-1995 period of a smaller price drop than in the 2004-2007 period of the larger price drop (483 bps). In any event, the Sept. 26, 2014 spread of 489 bps does not differentiate it materially from the historical periods. By their own logic, the bears should predict that high-yield will have a smaller decline in 2015 and the next year or two than in 2004-2007.
To be sure, we do not know how sharply the Fed will raise rates this time around. The evidence indicates, however, that it makes no difference, at least with respect to the probable magnitude of the high-yield price drop. As noted above, the rate hike centered in 2004 was smaller (309 bps) than the one that commenced in 2004 (426 bps), which was associated with a smaller price decline.
Before we leave the topic of incorrect hypotheses, let us not fail to note that duration had nothing to do with it, either. The only measure available for both periods, Macaulay duration to maturity, shows a negligible difference between the two historical-precedent periods. On Dec. 1, 1993 duration was 5.42, and on May 1, 2004 it was 5.39. That near equivalence may seem surprising, given the drop in yields between the two dates, but it is explained by a decline in the index’s years to maturity from 9.32 to 8.20. As of Sept. 26, 2014, duration stood at 5.00, implying (according to the bears’ reasoning) that the high-yield price drop should be smaller this time around. In fact, duration in all likelihood will make no material difference. (The shortening of Macaulay duration to maturity since 2004, despite the decline in yields, is attributable to a further shortening of the index’s years to maturity to 6.52.)
The factor that did make a difference
What did make a difference for price returns between the two historical-precedent periods was the extent to which longer-term (five-year) rates rose when the Fed boosted short-term rates. The table above shows that in the earlier period in which prices fell sharply, the five-year Treasury yield soared by 195 bps. In the later period of a small price increase, the five-year Treasury yield rose by only 71 bps. The SVT tightened in both cases, but the larger Treasury yield increase in the earlier period resulted in an 89 bps jump in the BofAML High Yield Index’s yield, versus a nugatory rise of 10 bps in the later period. Therefore, investors should assume that the magnitude of decline in high-yield prices in the upcoming Fed hike in short-term rates will be primarily a function of how sharply longer-term rates rise.
It is beyond the scope of this article to predict what will happen to five-year Treasury rates when the Fed tightens at the short end of the curve. Some will argue that rates must shoot up when, well before the tightening begins, the Fed terminates its quantitative-easing program. Others will counter that the faltering European economy will keep a lid on rates in that region and by extension in the U.S., since investors will balk at giving up substantial yield to own sovereign debt of countries they view as inherently less safe.
Yet another argument is that the Fed’s unprecedented monetary actions of the past few years will fuel more severe inflation than observed in the 1990s and 2000s. In that context, a near-term uptick in inflation
could therefore boost inflationary expectations, pushing up the five-year Treasury yield more dramatically than in previous periods of Fed tightening.
The many contending interpretations of the yield curve and scenarios for change in the curve present a formidable challenge to investors who are deciding how to position themselves in high-yield. They can feel confident, however, that by focusing on the direction of longer-term rates they are sidestepping irrelevant and ultimately unproductive analysis of yield levels and duration.
1. There was also a brief tightening episode between November 1983 and July 1984, when the effective Fed funds rate rose from 9.34% to 11.64%. Data constraints prevent us from replicating for that episode the analysis shown in this study for the other two tightenings.