- Long credit spreads are extremely high at present.
- Corporate finances still look very strong.
- We believe current long credit market pricing offers compelling opportunities.
Introduction and Summary
While 14% declines in stock prices since the summer of 2015 have been the headline story in the financial media, conditions in credit markets can be said to have been even more troubled, and the pain there has lasted much longer than that in the equity markets. In fact, credit spreads have risen to levels indicative of impending financial stress. This is especially true for long credit, for which option-adjusted spreads (OAS) are at levels surpassed only during the depths of the 2008–09 financial meltdown.
While we acknowledge the problems that have driven these market movements, we believe that reaction has been overdone, so much so that corporate credits—especially long-duration credits—now provide extremely attractive value. We present the reasoning behind this position in this paper.
First, we work through the details of pricing in the credit markets, specifically how much of the widening in spreads remains even when allowing for various extenuating circumstances. We then summarize the history of spread widenings, to put the present experience into context. Finally, we briefly analyze current corporate financial conditions.
Long Spreads Are Extremely High
Exhibit 1 shows OAS for the Barclays Intermediate and Long Investment-Grade Indices.1 As seen there, both spreads have increased sharply in the past 18 months, but long credit spreads have risen to especially high levels.2 Even with a 14 basis point (bp) tightening in recent weeks, February 29 long corporate spreads are still at levels surpassed only between September 2008 and June 2009, the depths of the 2008 crisis.3
Granted, markets have been buffeted by global growth worries, but these concerns do not approach in intensity the reality of what credit markets were experiencing over 2008–09. Yet, market pricing has long-term corporate bonds trading at spreads eclipsed only during that calamity. Actual credit crises in 1990 (Junk Bond Crisis), 2000–02 (Enron/Sarbanes-Oxley) and 2011 (peripherals’ crisis and European bank fears) failed to produce levels of long spreads matching what we are currently experiencing.
Now, a number of extraneous details need to be considered. Oil markets are battered. Financial system worries linger. Average ratings for the credit indices have declined over the years. Allowing for these, however, does not substantively alter our conclusions.
Exhibit 2 deals with the trauma in the oil patch and banking sectors by splitting the Barclays Long Corporate Credit Index into major components, with and without energy. Yes, there is variation in credit market stress across sectors. Long utilities’ spreads are “only” 194 bps and are way below the levels of 2000–02 as well as 2008–09, but this is not surprising given the relatively sound nature of the utilities industry.
Long financial spreads were indeed higher during the 2011 crisis, before European Central Bank President Mario Draghi backstopped the European banks, but let’s put this in perspective. US banks have reduced leverage and recapitalized in recent years, regularly passing stress tests conducted by the Federal Reserve (Fed). Still, OAS levels for long financials are higher than those for industrials ex-energy. So, spreads are high for financials as well as industrials. The currently high spreads for long credit in general are not just a matter of sky-high spreads in non-financial sectors.
As for industrial credits, the left panel in Exhibit 2 tells the story. Even upon abstracting from energy, long industrials ex-energy are trading at a 246 bps OAS, matching or exceeding previous highs outside the 2008 crisis. This is not a minor finding. Industrials ex-energy account for 59% of the long corporate universe.
Average ratings on the Long Corporate Credit Index have deteriorated some in the last 10 years, from A2/A3 in 2000 and 2008 to A3/Baa1 presently. However, the effect on long spreads is minimal. The fact is that within each quality strata the same findings hold as were obtained for long corporates in general.
Exhibit 3 shows historical Barclays Long Credit OAS by credit quality. As seen there, within each letter grade, current spreads still show up as being extremely high, generally higher than at any point outside the epicenter of the 2008–09 crisis, with only a few minor exceptions.
Long Aaa spreads spiked for 1 month in April 2000, when two Aaa issuers suffered severe trauma, but Aaa spreads quickly went back to more normal levels in May 2000, when those issuers were downgraded. Long Aa and long A spreads were a bit higher in 2011 than now, reflecting the importance of financials for the A and Aa universes. Exhibit 3 makes clear that even upon abstracting for the effects of ratings drift within the Long Credit Index overall, current long spread levels are extremely high by historical comparisons: at financial crisis levels despite the absence of a financial crisis.
It might also be argued that higher OAS for the Long Credit Index merely reflects higher duration for the index, 13.6 years presently, compared with 11.4 years in 2008 and 9.3 years in 2000—higher duration connoting higher risk. However, the fact is that the extension of long credit duration has occurred even while the average maturity of long credit has steadily declined, as Exhibit 4 makes clear.
If longer credit duration is not due to longer maturity, then it must be due to lower yield levels. Should lower yields mean wider spreads? We would say no. Historical evidence recounted in the next section indicates that spread levels have generally moved in line with yield levels, not inversely to them. If anything, lower yields should mean lower spreads.
More directly, we can look at risk premia for long Treasuries (USTs) to see whether there is any indication that lower yield levels have affected that market similarly to the way they have allegedly affected long credit. Term premia are the commonly accepted measure of the risk premia attached to long UST yields over those for shorter-maturity USTs. Commonly-referenced models of term premia show very low term premia at present. Similarly and more simply, the present slope of the UST yield curve from UST bills to UST bonds is relatively steep, at 225 bps, but it is a far cry from the 300 bps to 400 bps yield curve slopes holding at comparable points in the Fed hiking cycle in 1990 or 2004. In sum, we find nothing in history or in long UST pricing to support a contention that higher long credit spreads are reflective of higher duration now in long corporate bonds.
A History of Spread Tightening
Without question, it can be said that long credit spreads have widened to extremely high levels over the last 18 months. It is tempting to attribute that widening to the declining equity prices and rising recession risks of recent months, but history provides reasons to be skeptical of such conclusions.
Exhibit 5 extends previous analysis by plotting “raw spreads” for long corporate bonds back to 1919, the beginning of Fed data on interest rates. (Raw spreads from the Barclays Long Credit Index are also shown there to 1973, for comparison purposes.) Grey-shaded areas mark recessions.4
We have already discussed spread widening episodes in 2000, 2008 and 2011. The three episodes of spread widening in 1969–1970, 1974–75 and 1980–82 were all periods of aggressive Fed tightening and sharp increases in both UST and corporate yields. In each of these episodes, the Fed tightened to fight inflation that was already entrenched.
Earlier episodes of spread widening can also be characterized as occurring during periods of aggressive Fed tightening and/or financial systemic stress. Most of those times, spreads widened alongside rising UST bond yields. Recessions induced by systemic stress or aggressive Fed tightening saw wider spreads, but other recessions typically saw spreads unchanged or even narrowing. Purely counter-cyclical (non-aggressive) Fed tightening in the 1950s, early-1960s, 1987, 1993–94 and 2004–06 did not generate spread widening.
The notable exception to these rules was the spread widening that took place in 1985–86. That was the only widening that occurred alongside falling UST bond yields. That episode coincided with plunging oil prices. It was also short and relatively mild. Exhibit 5 shows long credit spreads rising more than long Baa spreads then, suggesting that spread widening was more severe in higher-quality sectors, the opposite of what we have seen recently.
Notice that sharp stock market declines did not typically drive spreads wider. Spreads did not widen alongside plunging stock prices in 1929–30 or the incipient bear market of the late-1960s (until aggressive Fed tightening began), and there was only a very brief, minor widening opposite the October 1987 stock crash.
The State of Corporate Finances
Corporate financial conditions are clearly relevant for market pricing. Exhibit 6 shows measures of corporate leverage. Debt-equity ratios have improved either substantially or dramatically, depending on how you slice them (left panel). The same is true for debt/asset ratios (right panel). Exhibit 7 shows debt service burdens as a ratio of earnings before interest, taxes, depreciation and amortization (EBITDA) to be at their lowest levels in 45 years. It could well be the case that individual companies are facing elevated financial risk. Aggregate finances for the corporate sector suggest, however, that overall default rates will remain low for the foreseeable future.
Yes, profit growth has stalled for the last 2 years. However, profits remain at extremely high shares of corporate output. So profit margins are declining, but from extremely high levels. Lack of profit growth is a problem for the equity market. Stocks are merely a perpetual bond with a coupon rate that rises with profit growth. Bonds, however, have a fixed coupon, and with interest burdens very low, there is no reason why a modest rate of decline in profits from extremely elevated levels should be a problem for corporate financial stability or for the corporate bonds market.
No matter how you reckon them, credit valuations look very attractive, especially so for long credit. In order to find higher spreads than now, you must look either to the dark days of the 2008 crisis, to the bad old days of high inflation and high yield levels, or to the calamitous days of widespread bank failures during the Great Depression.
In the knee-jerk risk-on/risk-off recent market environment, we have gotten accustomed to spreads and equity prices moving together, in the opposite direction of UST bond yields. However, this coherence is counter to the bulk of history. Should financial system crisis fail to emerge and should Fed tightening prove relatively non-aggressive—with inflation not rampant—history suggests that spreads should narrow sharply, whatever happens in the equity market.
Whatever ills onerous bank regulation might impose on the economy, it does minimize the chances of systemic crisis. As the old adage says, “Generals fight the last war,” and when it comes to the banking system, that “last war” is indeed still being fought. And recent developments certainly suggest that upcoming Fed rates hikes will be distinctly non-aggressive. Against that backdrop, there is reason to believe that credit spreads are due to tighten whether or not stocks and oil prices rebound. Finally, corporate finances remain in favorable condition, even with the heavy debt issuance of the last year. While investing in corporate bonds at present is not without risk, we believe those risks are manageable, and that current market pricing should more than adequately compensate investors for bearing those risks.
- What we here call “raw spreads” are the simple differences between yields-to-maturity on credit bonds and on UST bonds of equal maturity. OAS attempt to adjust these raw spreads for the effect of optionalities embedded in the prices of credit and UST bonds. The idea behind OAS is that these spreads better reflect the yield premium due solely to the default risk or credit risk presented by the credit bond in question. OAS data from Barclays are generally available as far back as 1989. Our analysis here focuses on OAS where those data are available. For long historical analyses, we analyze raw spreads in the expectation that the results gleaned from that analysis are not substantively different from those that would have arisen from an analysis of OAS, were such data available prior to 1989.
- The difference between the Barclays Long Credit Index and the Barclays Long Corporate Credit Index is that the former includes issues from non-federal-government agencies, specifically taxable municipals bonds—issued mostly under the auspices of the Build America Bonds (BABs) program of 2009—as well as USD-denominated debt issued by investment-grade-rated foreign governments. While our focus in this paper is on corporate credit, for some purposes, data on corporate credit are either not available (Exhibit 3) or not sufficiently long-lived (Exhibit 5).
- September 2008 marked the collapse of Lehman Brothers, Wachovia Bank and AIG, the federal government’s conservancy of Fannie Mae (FNMA) and Freddie Mac (FHLMC), and collapse of the commercial paper market. The following 9 months were the panicked aftermath of those events.
- Recession beginning- and end-dates are as determined by the National Bureau of Economic Research, reported at http://www.nber.org/cycles.html.