Following strong cumulative performance and increased inflows to high yield bonds so far this year, investors are asking if it is time to be cautious again. In recent weeks, volatility in risk markets and signs of a slowdown in macroeconomic strength have increased against a backdrop of increasing uncertainty in Europe.
Exposure to default risk is one of the primary concerns for investors in high yield bonds. However, a significant part of total default losses tends to be concentrated within the lower quality segment of the high yield market, which is often most vulnerable to an economic slowdown and equity volatility.
We believe investors can still benefit from the income potential of the high yield asset class by focusing on the higher quality segment of the market – generally, those rated single-B and higher, secured bonds and short duration bonds. These segments can offer relatively attractive spreads over comparable U.S. Treasury bonds and may help defend investors against an expected increase in defaults.
Among the best first quarters for high yield
The U.S. high yield sector just experienced one of its best first-quarter performances since 2004 – and the best since 2010. The Bank of America Merrill Lynch (BofA/ML) High Yield Master II Index, a proxy for the asset class, returned 5% for the first quarter of 2012. While the market has experienced considerable volatility since the end of the first quarter, total return for the ML High Yield Master II Index still stands at 4.9% through May 2012.
This has come with a record inflow of approximately $20 billion to high yield mutual funds and ETFs through May 2012, which is almost twice the $11 billion of inflows during the same period in 2011, according to AMG/Lipper data.
Looking at credit losses and “excess spread”
As many investors know, high yield bonds are sometimes called “junk” bonds because of their exposure to default risk, for which investors demand a significant premium to lower risk assets, including U.S. Treasuries and investment grade-rated corporate bonds. At the end of May, the average spread for high yield bonds was 716 basis points over comparable Treasuries and 233 bps over investment grade bonds, based on BofA/Merrill Lynch’s U.S. High Yield Master II and U.S. Corporate Master Indexes.
At first glance, a spread of 716 bps may look attractive compared to the average annual credit loss in high yield of 265 bps for 1992-2011 and especially compared to the average annual credit loss of only 135 bps in 2010-2011, as reported by Moody’s Investors Service. (By credit loss, we mean loss of principal in a high yield fixed income portfolio due to default by bond issuers.) However, high yield spreads should compensate investors not just for credit losses, but also for any volatility and illiquidity associated with the asset class.
To gauge how much high yield bond investors are really compensated, we can break out the “excess spread” above the amount lost to defaults. Figure 1 plots the excess spreads embedded in high yield spreads across different components of the high yield market assuming historical default loss levels and using the average spreads as of 1 June 2012 for the BofA/Merrill Lynch U.S. High Yield BB-B Rated Constrained and BofA/Merrill Lynch U.S. High Yield CCC and Lower Rated Constrained Indexes.
The difference between the rating categories is telling, not just on an absolute basis but also on a proportional basis. While 60% of overall high yield spread in this example is “excess spread,” which compensates investors over and above historical losses, excess spread is as high as 76% for the higher quality segment of high yield, rated double-B and single-B, and as low as 33% for the lower quality segment of the market, rated triple-C. So while current average spreads on triple-C rated bonds may look appealing at 1,227 bps, most of that is compensating investors for average historical annual default losses.
All this doesn’t mean investors should necessarily avoid triple-C rated credits at all costs. For one thing, agency ratings often lag improvement or deterioration in credit quality, and therefore, they are not always commensurate with the current and future expected credit quality of an issuer. However, the point is that on a broad basis, credit selection becomes increasingly critical moving down the quality spectrum, and the excess cushion, or the room for error, tends to become proportionally smaller.
Potential for rising defaults from cyclical lows
Our analysis obviously assumes that default losses remain at or below their historical averages. Thanks to the Federal Reserve’s aggressive monetary policy, which has kept interest rates near historical lows, high yield issuers were able to refinance and extend their debt maturities in 2010 and 2011, and default rates saw a dramatic drop to 1.8% in 2011 from an unprecedented high of 13.6% in 2009, according to Moody’s.
However, we believe the cyclical bottom in default rates is behind us, and in 2012, we will likely start to see a gradual increase toward the mean in default rates and credit losses. We see indications of this in the relative tightening in lending standards compared to 2011 as reported in the Fed’s quarterly Senior Loan Officer Survey. We believe this survey is usually a good high-level/macro forward indicator of default rates, as shown in Figure 2, and it is one of the inputs into our macro default forecasting model.
Based on current lending standards, our analysis indicates a default forecast of 3% for 2012. (For context, Moody’s currently forecasts a 3% default rate overall by issuer for 2012. Various other forecasts in the market range from 1.5%-4%.) If lending standards were to tighten again given the backdrop of ongoing uncertainty in Europe, we could see defaults rise to 4%-5%. We therefore expect to see greater credit losses in high yield in the future, especially in the more vulnerable, lower credit quality segment.
Finding opportunities in high yield today
So where do we think spreads will go from here? We find it instructive, given the strong first quarter rally and subsequent volatility, to analyze where spreads stand today compared to their most recent highs, achieved last year on May 11. Figure 3 shows this comparison.
Based on the BofA/Merrill Lynch U.S. High Yield BB-B Constrained Index, the average current yield-to-worst (defined as yield under the worst case scenario for investors, including call options being exercised) for the double-B/single-B market is 6.94%, 77 bps above the low of 6.17% seen last year. However, the spread distribution tells a more nuanced story: While double-B/single-B spreads during May 2011 were rather concentrated toward the tight end of the range at 300 bps-350 bps, they are more diffuse today, reflecting more credit differentiation across the double-B and single-B high yield asset class.
Within high-quality high yield, one area that we find attractive, especially in light of available alternatives, is short duration bonds from issuers with relatively attractive credit and liquidity profiles. Given their tendency for lower price volatility than the broader high yield universe and relatively attractive yields, such bonds may appeal to investors with cash on the sidelines. Short duration implies not just bonds with short final maturity but also, given the callable nature of high yield, those with a high likelihood of being called within two years. Many of these near-call bonds were issued in 2009-2010 with high coupons compared to those today.
Another reason we like short-dated bonds is they can generally trade wider than the spread valuations implied by their credit default swaps (CDS). In other words, short-dated high yield bonds can generally trade at a better discount to the level implied by their default probabilities than longer-dated bonds. Figure 4 depicts this technical analysis for one of the largest high yield debt issuers with actively traded CDS contracts. The cash CDS basis shown in the chart represents the spread (to the matched U.S. Treasury swap rate) differential between the CDS spread and the cash bond spread. A negative number means that the cash bond offers a higher spread than the CDS contract. The more negative this number, the more attractive the cash bond looks compared to its CDS counterpart.
Why does this situation exist? Generally, the longer the maturity, the greater the compensation or spread investors demand per year for buying or selling protection on the credit, resulting in an upward sloping, or steep CDS spread curve as seen in Figure 4. Similarly, the yields on high yield cash bonds also increase with maturity. However, when it comes to Treasuries, the short end of the yield curve is anchored by the Fed’s near-zero interest rates, creating a somewhat steeper Treasury bond curve compared to the high yield curve; this generally results in higher spreads for shorter maturity high yield bonds. Consequently, investors looking for that extra spread can potentially find it in the two- to four-year part of the high yield market without having to reach out in terms of duration.
Another area of the double-B/single-B higher yield sector that we find attractive is secured bonds, especially those with first priority/lien; we believe these can offer the security and collateral benefits of a loan while retaining the liquidity and higher coupon/income characteristics of high yield bonds. Many of these bonds in high quality credits have spreads of 450 bps-550 bps over comparable Treasuries. Even without allowing the benefit of potentially higher secured recovery rates and just using historical double-B average annual cohort credit losses of 127 bps, excess spreads are as high as 75% (of total spread) for these secured bonds, compared to 60% for the overall high yield market (see Figure 2).
To summarize, after a strong performance year-to-date, we expect that high yield bonds will continue to react to overall market volatility in the months to come and corporate defaults will likely begin to rise from the recent cyclical lows. Given current spread levels in high yield, we see value in higher credit quality, short duration, and secured high yield bonds because of their potential for attractive excess spreads over and above expected credit losses from defaults.