A lively debate is taking place in the marketplace over the merits and risks of investment-grade average collateralized debt obligations (IG CDOs). Critics argue the following: investment-grade CDOs are too highly levered; investment-grade credits experience rapid credit deterioration; investment-grade CDO managers pick the yieldiest credits for any given rating category; and investment-grade CDOs are credit "barbells."
Salomon Smith Barney strongly disagrees with all four points. We believe that the credit tranches (i.e., equity and second loss tranches) issued by investment-grade CDOs offer significant unrealized value, a view that we expound upon in our recent report "Investment Grade CDOs." In short, we consider IG CDO equity and triple-B tranches cheap.
Investors should consider adding investment-grade CDOs to a well-diversified portfolio for a number of reasons. Chief among these reasons are: stable risk-adjusted mezzanine and equity tranche returns; lack of correlation of returns with those of other asset classes; low default rate and default rate volatility; manageable transitions of collateral to default; greater liquidity and obligor/industry diversity; no ramp-up risk; noncall for life feature; and maturity profile of investment-grade bonds.
Stable risk-adjusted CDO equity and mezz returns
Lower default rate volatility leads to more stable risk-adjusted returns on investment-grade average CDO equity. For perspective, Table One compares historical returns and Sharpe ratios for widely-held asset classes with returns and ex-ante Sharpe ratios for CDO equity projected from our simulations.
Based on our simulations, the ex-ante Sharpe ratios that we have calculated for investment-grade CDO equity returns compare very favorably with those of other asset classes. The average internal rate of return (IRR) for investment-grade CDO equity was 16.5% (ranging from 14.1% to 18.6%), the average Sharpe ratio was 2.3 (ranging from 1.4 to 2.9), and the average standard deviation of returns was 5.1%.
Lack of correlation
with other asset classes
Although there is no established CDO performance index, the performance of the typical investment-grade CDO is inextricably linked to the behavior of the U.S. high-grade bond market. We used the Salomon Smith Barney Corporate Index as a proxy for the performance of corporate bonds. The SSB Corporate Index is virtually uncorrelated with traditional equity indices and negatively correlated with alternative investments such as hedge funds and private equity. From an asset allocation perspective, this suggests that investing in IG CDO equity is an excellent way to diversify within the traditional and the alternative-investments portions of a well-balanced portfolio.
Nonetheless, one must be cautious when engaging in correlation analysis across illiquid asset classes, which by definition are difficult to mark to market.
Low default rates and low default rate volatility
According to Moody's, the average annual default rate for Baa3 credits from 1983 to 2000 was 0.46% (standard deviation 1.16%) versus an average default rate for B2 credits (typical of high-yield CDOs) during the same period of 7.41% (standard deviation 6.08%).
According to market data and ratings agency reports, default rate volatility increases dramatically at the lower-end of the credit spectrum. In fact, the volatility of B2 default rates is more than ten times higher than the volatility of Baa defaults. Moody's makes the same point in a recent report: "As one moves down the ratings scale . . . default risk increases, and the risk that the default rate will differ from the historical average also increases." Default rate volatility is extremely important when analyzing CDOs, given the granular nature of the typical CDO portfolio - most CDOs contain between 80 and 120 names.
Although it is not surprising that Baa default rates are lower than Ba default rates, investment-grade CDOs are more highly levered than high-yield CDOs. Thus, the critical question for investment-grade CDO note holders is the following: is there value given the greater leverage in the typical investment-grade CDO? In the "Methodology" section of our report we answer this question with a resounding "yes."
Transition, not cliff!
The vast majority of investment-grade firms do not default, and unlike high yield companies, those that default generally do not fall off a "cliff" from investment-grade directly to default. Credit problems for investment-grade companies have historically tended to develop much more gradually. For example, our analysis shows that four years prior to default, the average investment-grade bond was rated single-A minus and just prior to default it was rated single-B. During the intervening 48 months, the average rating slid eight rating subcategories. By contrast, four years prior to default, the average high-yield bond was rated B+ and just prior to default it was rated CCC+. The average rating on these bonds declined only three rating subcategories before default.
This gradual decline in credit quality, coupled with the deep, liquid investment-grade bond market, often give investment-grade average CDO managers ample opportunity to trade out of credit-impaired names. The same cannot always be said of the high-yield market.
As the following chart illustrates, since 1980 only seven firms have dropped from investment-grade to default within a month.
If a single security comprises a large percentage of the asset base of a CDO, its default early in the life of a deal can significantly impair CDO equity returns. Thus, a high degree of industry and obligor diversity is crucial to the success of any CDO. This diversity is much easier to achieve in the investment-grade market than it is in the high-yield market. In addition, where the IG CDO manager engages in trading after the close, such trading is done in a much deeper and more liquid market. As a result, selling becomes a much more realistic exit strategy (i.e., less price volatility) for an investment-grade CDO than for a high-yield CDO.
The breadth and depth of the investment-grade market dwarfs that of the high-yield market. The investment-grade market contains 1,227 firms with an aggregate market value of $1,689 billion. The high-yield market contains 25% fewer firms (883) with an aggregate market value that is less than one-fifth ($258 billion) the market value of the investment-grade market. The investment-grade market also possesses more industry diversity.
With respect to industry diversity, the largest industry in the investment-grade market (banking - 628 firms) has almost nine times the firms in the largest sector of the high-yield market (service/other - 74 firms). The number of firms in the top ten investment-grade industries (2,549) is more than five times larger than the number of firms in the top ten high yield industries (490). Finally, with respect to the smallest number of firms per industry sector, the investment-grade market has 20 industries with fewer than 20 firms versus the high-yield market, which has 46 industries with fewer than 20 firms. The high-yield market has only 61 industries in total (i.e., more than two-thirds of the industries have fewer than 20 firms).
No "Ramp-Up" risk
Because of the depth of the investment-grade market, most investment-grade average CDOs are fully "ramped up" at closing, and this feature can benefit both rated note holders and CDO equity holders. That is, the absence of a ramp-up period reduces the risk that credit spreads will tighten during the collateral accumulation process. Because other markets are not nearly as deep and liquid, the same cannot be said for high-yield CDOs, leveraged loan CLOs, or multi-sector CDOs. A CDO that accumulates assets after closing is exposed to the risk that credit spreads will tighten, and the CDO manager will be forced down the credit spectrum to meet minimum average coupon and margin requirements and generate sufficient returns to CDO equity holders. In this chase for yield, some managers will take on a disproportionate amount of credit risk, and this has the potential to affect all note holders in the capital structure.
bonds non-call for Life
Many arbitrage CDOs have five-year reinvestment periods during which note holders are paid only interest. During this period, the collateral manager reinvests all principal collections in new bonds and loans, and it must do so while complying with minimum ratings, coupon, margin, and tenor restrictions. In a declining interest rate environment, or when credit spreads are tightening, there is a risk that the arbitrage projected at closing will deteriorate. This reinvestment risk is exacerbated in high-yield CDOs because the typical ten-year high yield bond is callable after five years. In general, investment-grade CDOs have far less of this uncontrollable portfolio turnover because most investment-grade bonds are noncall for life.
Finally, in our opinion, obligor diversity is even more important than industry diversity and this issue is extremely important in investment grade CDOs given the leverage in these structures. All other things equal, a larger number of smaller positions are better, but at what point do the benfits of obligor diversity trail off? Based our our simulations, we have found that increasing the number of bonds above 100 provides little incremental risk reduction in the form of decreased volatility. This result supports our view that the leverage inherent in IG CDOs makes it risky for a manager to have less than 100 bonds in position and that no position should exceed about 1.25% of the portfolio.