Ratings

Standard & Poor's has now assigned recovery ratings to more than 1,300 secured loans with 2,362 rated tranches. Chart 1 shows the frequency distribution of these ratings by loan since recovery ratings began in December 2003. The median rating is in the '3' recovery rating category, denoting 50%-80% recovery, and the mean rating also falls near that category at '2.78'.

Recovery ratings show considerable dispersion in recovery prospects, even for secured loans. Standard deviation stands at 1.34, or nearly 50% of the mean, representing more than one recovery rating category. This contrasts with the more tightly grouped distribution generally associated with historical secured loans. In the historical data on secured loan recovery recorded in Standard & Poor's LossStats Database, standard deviation is less than 40% of this mean. This dispersion may well be characteristic of the recent market for secured loans, as LossStats data also shows an increasing dispersion of recoveries in recent years. This pattern, Standard & Poor's concludes, makes it particularly important that lenders and investors in leveraged loans distinguish between loans with high recovery potential and those with lower recovery prospects. In the current highly liquid market, recovery prospects cannot be estimated simply based on the fact that a loan is secured, as both recovery ratings and historical recovery data confirm that the class of secured loans in the current market includes increasingly diverse recovery prospects.

Recovery ratings continue to congregate at the lower end of the recovery spectrum. Loans with recovery ratings of '4' and '5', indicating the potential for less than 50% principal recovery, account for more than 28% of loans. This is materially above the historical average of 17% in the LossStats Database. Standard & Poor's attributes this pattern of reduced recovery prospects to trends characteristic of the current highly liquid market, in which lenders and loan investors continue to reach down the credit spectrum for increased yield.

Second-lien loans have emerged as an ongoing form of senior secured debt in U.S. (and recently in European) loan markets. Standard & Poor's recovery ratings for second-lien loans show that most of the loans in the lowest recovery rating bucket of '5' (indicating the expectation for less than 25% principal recovery) were second liens.

Also, the aggressive expansion of dividend cash-outs (a common feature of LBO-driven secured loans) has put downward pressure on potential recovery values. In our view, this levered dividend monetization represents a material drag on recovery because it increases debt while often failing to increase the value available to secured lenders in the event of a default. Of the 93 secured loans identified by Standard & Poor's Leveraged Commentary & Data as dividend cash-out deals that were assigned recovery ratings, more than 50% were rated '4' or '5'--about twice the percentage of overall secured loans with recovery ratings at this level.

Recovery ratings indicate few patterns across industry sectors. Chart 5 shows the average recovery rating in each sector arranged in ascending order. The data indicates that the sectors on the left of the chart (from high tech to health care), with loans that are generally secured by few hard assets, on average have lower recovery ratings than the sectors on the right, which are secured for the most part by hard asset pledges. But the differences between average recoveries are fairly small, with dispersion within the sectors generally exceeding differences between the sectors. Recovery rating distributions were distinctive in a handful of sectors, notably power, energy, minerals, and real estate, which showed both average and median recovery at the '1' recovery rating (see chart 6). But other sectors showed no distinguishing patterns. This seems to indicate that the determination of recovery prospects tends to rely more on issue-specific parameters than sector-specific ones.

Average recovery ratings moved little from month to month over the period since their inception in December 2003 (see chart 7). Since January 2004, average ratings ranged from a low of 3.18 in March 2004 to a high of 2.24 in December 2005. Volatility increased somewhat at the end of 2005 and beginning of 2006. This may be characteristic of late-stage expansion of the new-issue market as it tests the level of recovery prospects that lenders and loan investors are willing to accept. Still, the distribution of ratings did not shift materially for more than one or two months at a time. This relative consistency could reflect the current particularly liquid leveraged debt market, which has persisted since the inception of Standard & Poor's recovery ratings, or it may reflect other factors. Forces driving monthly movements would become more clear should the highly liquid state of the leveraged loan market turn.

One of the most important questions regarding post-default recovery is the extent to which it is correlated with default risk. Chart 8 shows the scatter plot between recovery ratings and corporate credit ratings (i.e., Standard & Poor's traditional default risk ratings) on each loan in our recovery rating portfolio. The chart shows little relationship between the distribution of recovery ratings and default ratings. This confirms the premise underlying secured lending at the lower end of the credit spectrum. Many loans to borrowers carrying corporate credit ratings of 'B' or below are often structured to provide high post-default recovery, in order to offset potentially higher default risk. At the same time, loans to borrowers with less default risk (those with corporate credit ratings from 'BB' to 'BBB') in some cases have lower recovery ratings in the current market, as these borrowers saw less need to offset default risk with structures designed for strong recovery.

Recovery ratings do not yet address to what extent recovery prospects change with changes in default rates. To date, Standard & Poor's recovery ratings have been issued in a period characterized by strong and expanding liquidity in the debt markets, driven, in part, by cyclically low and declining default rates, as measured by the portion of debt issues in default. How recovery ratings and actual recoveries will change in a less benign period can be determined only once market liquidity has turned. We expect that ratings and actual recovery levels over time will show a correlation with default rates, as has been observed in the LossStats Database of actual recovery data (see Chart 17 in "Annual 2005 Global Corporate Default Study And Rating Transitions," published Jan. 31, 2006 on RatingsDirect).

One of the key features of the recent leveraged debt markets has been the increasing role of institutional investors, both directly and through structured vehicles (such as CDOs). While institutions held less than 25% of new secured loans in 2000, their share of the loan market grew to greater than 65% by 2005. Together, these trends have helped to expand the demand for leveraged loans and contributed to substantial liquidity in both the loan and distressed debt markets.

It is too early to tell how these institutional holders will view loans should the current market liquidity turn in response to credit or other events. The limited duration of the last market reversal in 2001-2002 and the rapid strong return of demand for leveraged loans show the potential for these new entrants to remain a major factor in the leveraged loan market over the long term. However, some industry sectors that have been aggressive borrowers in the leveraged loan market recently (such as merchant energy in 2000-2001 and telecom somewhat earlier) have proved that they can pull out of the market just as fast as they moved in, if not faster. On balance, however, Standard & Poor's does not expect that such rotations will change the overall trend toward greater institutional participation in the syndicated bank loan market.

One of the principal drivers for high recoveries in secured loans relative to other debt classes has been the cohesion that many bank groups have maintained as they worked toward maximizing recovery for senior secured bank lenders. Banks maintained this cohesion because they represented par holders for the most part, and focused on achieving full par recovery and avoiding the need to recognize losses.

In the current U.S. loan market, as institutional holders typically collateralize debt obligation structures and prime funds, private equity funds, and hedge funds replace banks as the chief holders of bank paper, Standard & Poor's expects that there is the potential for the cohesion among lending groups to decline should a loan face default and restructure. Institutional holders of bank paper often have shorter holding periods than banks. As a result, should a loan become distressed and enter default, the group of lenders holding the credit in current markets often will include institutional holders--at least some of which will have acquired the credit not at par, but at a substantial discount. Therefore, the incentives for restructuring and workout may be very different for institutional holders (particularly non-par buyers) than for the earlier bank groups. We note that every default and restructuring is different, and few loans with substantial non-bank holders have completed restructuring to date. However, we expect that the changing dynamics among non-bank holders will have the effect of limiting lender cohesion, which, in turn, could potentially reduce recovery prospects.