S&P Viewpoint

The weakness in the leveraged finance market in the past few weeks emphasizes the risks faced by banks that underwrite and invest in these assets. The initial change in investor sentiment came as the huge primary issuance pipeline provided an opportunity to push back the most aggressive transactions, particularly covenant-lite and payment-in-kind (PIK) structures. Further impetus has come from the problems in the U.S. subprime mortgage market and the resulting spillover to other asset classes. A few high-profile leveraged finance issues were consequently postponed, and financial sponsors chose to reprice or add covenants to others to get them away. Certain banks have sustained losses on unsold underwriting positions, but the sums involved are not material at this point.

The jury is out on whether the correction marks a much-needed return to more rational risk-based pricing, or the beginning of a severe downturn. The fact that sensibly structured and priced deals are still being closed indicates that the turn in the market may be relatively moderate. It also underlines the shared interest of financial sponsors and banks in working together where possible to place the transaction pipeline. Our central expectation is that activity will remain soft at least through the summer, but remain underpinned by the generally supportive economic outlook. We expect a moderate increase in corporate defaults from the current cyclical low, but a harder landing cannot be ruled out. Heightened risk sensitivity among investors also raises the possibility of further falls in prices and liquidity. For the purpose of this article, Standard & Poor's Ratings Services has stress tested the impact of a severe downturn on banks' risk exposures and revenues. Our analysis shows that this would certainly be a painful scenario, but banks should be able to absorb the effects at their current rating levels. A widespread deterioration across the credit markets would cause rating pressure, however.

For some time, leveraged finance has been a focus of our rating discussions with the banks that are active in this market. They include not only the major players that underwrite the primary transactions, but also smaller banks, particularly Europeans, that invest in these assets.

We have previously highlighted the potential rating implications of a market downturn (for more information, see "The Outlook For Banks' Leveraged Finance Revenues And Risks," published on Oct. 25, 2006), and the purpose of this article is to provide an update on our views in light of recent developments.

Before June 2007, the leveraged finance market appeared ripe for a correction as LBOs increased by size and number, and investors' clamor for assets eroded risk-based price differentials. Banks' underwriting criteria were similarly frothy, with leverage multiples moving ever higher (see chart 1), interest coverage falling, and new issuance increasingly skewed to higher risk assets such as cov-lites, PIKs, second-liens, and even equity bridge loans. The market had effortlessly navigated previous bumps in the road such as equity and oil-price volatility, the May 2005 downgrades of Ford Motor Co. and General Motors Corp. (both rated B/Negative/B-3) , and the Amaranth hedge fund collapse. It was finally undermined in June 2007, however, by a combination of the rebalancing of supply and demand, concerns over rising global interest rates, and greater risk sensitivity among investors in light of the weakness in U.S. subprime mortgages.

 Chart 1

The softening of the market has taken slightly different forms in the U.S. and Europe. Sentiment changed more quickly in the larger U.S. market as the need to place the mammoth deal pipeline--estimated at $215 billion by Standard & Poor's Leveraged Commentary & Data--shifted the balance of power away from sponsors and arrangers toward investors. Investors responded by pushing back against the most aggressive structures and pricing, particularly cov-lites (which represented 35% of U.S. institutional issuance in the first half of 2007) and PIKs. Some high-profile new issues such as U.S. Foodservice were consequently postponed, but others proceeded after sponsors added covenants, changed pricing, and/or boosted fees to give more comfort to the investors and arrangers. U.S. secondary market pricing has also slipped and indices now stand below par (see chart 2). The main difference in Europe is that cov-lites are much less prevalent than in the U.S. and primary issuance has therefore suffered a little less disruption, although there have been compromises on pricing and covenants. Another factor in Europe's favor is that the pipeline, despite reaching a record level, is nothing like the burden of its U.S. counterpart. It is now very probable that cov-lites will fail to take hold in the European market (and will retreat in the U.S.), but cov-loose transactions were well established in Europe, and offer little more protection to investors than cov-lites. Secondary market prices also fell in Europe and liquidity is tighter even for the most frequently traded names, but indices remain slightly above par as some investors have reportedly re-entered the market after seeing value in the credits that had fallen the furthest.

 Chart 2

There are a number of routes through which the weakness in U.S. subprime mortgages has been transmitted to the leveraged finance market. First, there are concerns that U.S. house price weakness will increasingly undermine consumer spending and consequently dilute corporate creditworthiness and economic growth, initially in the U.S., but ultimately across the world. This is evidenced by poor sales figures from U.S. retailers such as Home Depot Inc. (BBB+/Stable/A-2) and Sears Holdings Corp. (BB+/Negative/--). Another transmission mechanism arises from the wider dispersion of risk in the modern credit markets. For example, some hedge fund liquidity has left the leveraged finance market in recent weeks, partly because they see better value elsewhere, but also because they have had to pay margin on discounted U.S. subprime mortgage assets. Finally, an adverse change in sentiment in one market tends to immediately affect others that are perceived to be similarly risky. For example, the inclusion of U.S. subprime mortgages in a number of CDOs has triggered tighter liquidity across the CDO asset class. This is important to the leveraged finance market as most loan distribution to end investors takes place through CLOs.

The outlook for corporate credit quality should be the single most important determinant of future leveraged finance pricing and liquidity. Standard & Poor's does not at this point have widespread concerns in the corporate sector, but it is clear that an upturn in defaults from the current cyclical low cannot be far away (see chart 3). Probable triggers include the pressure on consumer spending as discussed above, coupled with increased corporate leverage and a squeeze on profitability from higher interest rates and input costs. The rise in corporate leverage is not only due to the growth of LBOs, but is also the result of some investment grade companies choosing to gear up to improve balance sheet efficiency. These companies were also motivated by the wish to avoid being victims of LBOs, but sponsors' firepower (in terms of the number and size of deals) will surely be lower following the current market correction. For existing leveraged borrowers, near-term debt servicing is protected by bullet repayment structures, interest rate hedges, and weak or nonexistent covenants. Together with the generally favorable global macroeconomic environment, this points to a moderate increase in the corporate default rate from the current level. However, it may also drive a sharper spike in defaults if the economy should deteriorate unexpectedly, as it may delay the identification of credit problems and therefore reduce the time available for restructuring.

 Chart 3

The key question at this point is where the leveraged finance market goes from here. Is the recent softening a deflation of the previous bubble and a return to more sustainable levels of pricing and leverage, or the beginning of a severe downturn? The changes seen to date certainly indicate a more rational market, with investors starting to discriminate again on the basis of leverage multiples, covenant protection, and borrowers' creditworthiness. For example, there was little pricing distinction between 'BB' and 'B' credits a few months ago, or between cov-lites and traditional loans, but a gap has now reopened. Demand for primary issuance has softened as investors can now wait to see how transactions trade in the secondary market before deciding whether to buy them. Several existing CLO managers continue to plan new transactions, but some new managers have put launches on hold. All things considered, there are tentative signs that a resistance level may be emerging in market pricing for sensibly structured deals, but there is evidently a great deal of nervousness among participants and negative news could spark further declines. Possible triggers might include a further deterioration in U.S. subprime mortgages, leading to a general loss of confidence in CDOs, or one or more high-profile corporate credit events.

The leveraged finance boom of recent years was the result of both structural and cyclical factors (see chart 4). The main structural factor was fund managers' increased asset allocation to private equity, which provided sponsors with huge war chests to finance new transactions. The principal cyclical factor was the greater availability of bank credit at ever-reducing margins, which was itself facilitated by strong demand for higher yielding assets from institutional investors. Sponsors consequently increased their M&A activity, and underwriting banks earned substantial fees while quickly offloading most or all of the risk to external investors. Even if the current market hiatus turns out to be a mild correction, it seems certain that the LBO market's glory days are over and the future environment will be more challenging for sponsors and banks alike. This has clear consequences for the risk exposures and revenues of the banks that we rate.

 Chart 4

The shift from a "take and hold" to an "originate and distribute" underwriting model has undoubtedly been a positive development for banks' credit risk profiles. The emergence of hedge funds and other institutional investors allowed banks to be more selective over the risks they retain, and nowhere has this been more apparent than in the leveraged loan market (see chart 5). The weakness of this business model, however, is that it encourages banks to underwrite larger and more risky exposures, and thereby increase their vulnerability to an abrupt, adverse change in investor sentiment. There is an argument that the institutional investor community is now so broad and diverse that demand will remain reasonably firm through a market cycle, but this view seems rather optimistic given that many of the new breed of investors target short-term relative value.

 Chart 5

Banks' heightened vulnerability to a turn in investor sentiment is not only due to the increased number and size of leveraged finance transactions, but also because they take longer to complete. Sponsors have traditionally acquired private companies or subsidiaries of public companies, and the selldown of these transactions to the targeted hold level has typically taken three or four months (see chart 6). As sponsors turned their sights on large public companies, however, the distribution speed slowed as acquisitions became dependent on regulatory approvals, go-shop clauses, and so on. Banks were therefore carrying larger underwriting pipelines when the market turned, and we expect this will contribute to mark-to-market losses in third-quarter earnings, particularly for those that were caught with unsold positions in aggressive structures such as cov-lites and PIKs. Banks often say they only underwrite exposures that, if necessary, they would be willing to hold in full on their balance sheets. This may be true of some, but they would clearly not choose to hold large, overvalued positions such as cov-lites or high-yield bonds in a declining market, and investment banks in particular are constrained in funding illiquid assets.

 Chart 6

Most banks increased their risk management focus on leveraged finance during the past year in the expectation that a correction was inevitable. The most risk-conscious lenders have undertaken more detailed reviews of borrowers, and tightened limits on underwriting pipelines, transaction sizes, sectoral exposures, selldown timetables, and hold positions. The performance of these teams is typically assessed according to their return on capital and balance sheet usage, and they are therefore incentivized to maximize the velocity of debt through their books. We saw few signs of a fundamental reduction in risk appetite prior to June 2007, however, as banks were keen to preserve their share of the revenue pool. Indeed, the introduction of equity bridges shows that some banks were still willing, albeit reluctantly, to use their balance sheets to maintain good relations with sponsors.

In addition to higher credit losses, another potential threat to banks is litigation from investors. The cov-lite phenomenon emerged as the demand for leveraged assets exceeded the supply and investors stopped discriminating between the underlying risks. In some ways, this was not a particularly surprising development: since loans are increasingly traded like bonds, why retain covenants on loans when they do not exist on bonds? Nevertheless, we see cov-lites as a classic top-of-the-market development and they suggest that some investors may not fully understand what they have bought. The resulting litigation threat is unquantifiable, but is most likely to occur if a borrower defaults shortly after syndication and it can be established that the arranger knew, or should have known, of the underlying cause. This makes it even more essential that banks undertake rigorous credit analysis of their transactions.

So far, this section has primarily focused on underwriters, but banks also act as investors in the leveraged finance market (see chart 5). The reported numbers partly reflect the residual pieces of transactions that underwriters typically retain, and also include loans acquired by smaller banks, particularly Europeans, that wish to gain leveraged finance exposures. The true positions held by these smaller banks are probably overstated by primary market shares because excess demand for assets enabled them to flip exposures into the secondary market and make easy profits. This opportunity has now ended, and investing banks could be left exposed if they are sitting on significant portfolios. As public disclosure in this area is poor, we surveyed the rated European banks earlier this year to determine the size of their exposures and the quality of their risk management practices. A number of small and midsize European banks maintain portfolios of senior loans and highly rated leveraged loan CDO tranches, but they have limited exposure to high-yield bonds, junior loans, and lower rated CDO layers. For most banks, the portfolios are not material in the context of their overall earnings and balance sheets. In some cases, the widely held view earlier this year that the market was overvalued led to selective disposals before the recent turn in the market. Some midsize European banks hold sizable leveraged loan portfolios, however, and we will continue to pay close attention to the performance of these assets. With regard to the residual pieces retained by underwriters, commercial banks traditionally kept 10%-15% of benchmark transactions, but this proportion fell with the increase in average loan sizes and the strengthening of investor demand. For example, we found that the most risk-conscious banks typically held about $20 million of a 'B' exposure, which is low by historical standards and should benefit their credit and spread losses if the market downturn becomes more severe.

Credit hedging of leveraged finance exposures has been slow to take off because the syndication market was so liquid and there were structural issues to resolve, such as the novation of hedges to new loans after refinancing. The industry has worked to address these points and certain banks were opportunistically growing their name-specific and macro hedge books prior to June 2007, but overall activity across the industry remained relatively low. This picture is unlikely to change in view of the increased cost of hedging in the current market.

Turning to the outlook for banks' revenues, the leveraged finance fee pool is likely to stabilize or fall as sponsor led M&A activity slows in the new market environment. Although sponsors have huge amounts of unspent funds, their internal rates of return will be pressured (assuming stock-market valuations remain high) by their higher cost of capital and the need to provide greater amounts of equity to new transactions (as underwriters' leverage tolerance declines). Banks' revenues will be further squeezed as refinancings and recapitalizations slow, and if they choose to bring in more subunderwriters to mitigate risk on new transactions. The ability to adjust cost bases in these circumstances would be vital.

For the purpose of this article, we have conducted simple stress tests to illustrate the effects of a severe, but plausible, downturn scenario. Banks are exposed to both name-specific credit events and generalized spread widening, and the former could well trigger the latter in today's nervy markets. Underwriters face the highest potential exposures, but investing banks could also experience sizable losses relative to their earnings and capital.

The life of a leveraged finance underwriting commitment has three basic stages. First, banks compete to win business from the client, which might itself be in a competitive M&A situation. Second, once the mandate has been won, it becomes an accepted and reportable commitment, but it is usually unfunded and some of the terms are still subject to negotiation. Third, on completion of the underlying transaction, it becomes a hard, funded commitment. Accepted commitments feature material adverse change clauses and, in some cases, price flexing, but banks are only partially protected since the absolute spread is capped. Sponsors and banks have a shared interest in the efficient distribution of the outstanding pipeline, however, and some sponsors have recently added covenants and changed other key terms of accepted commitments to help banks launch the transactions. In contrast, other sponsors have taken a harder line and required banks to adhere to the terms of their original commitments. Hard commitments fully expose the lender to credit and spread events.

The recovery ratings published by Standard & Poor's corporate group help us to estimate loss-given-default (LGD) and potential spread widening in a severe downturn (see chart 7). If the expected LGD on a given loan is 20%, for example, we can make a simple assumption that it is unlikely to trade much less than 80 cents on the dollar in a reasonably liquid market. The recovery ratings indicate that senior loans are generally well collateralized, but junior tranches and high yield bonds/bridges are more vulnerable. The post-default workout of a leveraged borrower would be tricky in today's market given the variety of investors holding the paper and the likely differences in their incentives. For example, an investor that had hedged its position would presumably favor default over restructuring. The large number of investors in each deal means that borrowers' repayment difficulties are likely to become apparent more quickly, although bullet structures may disguise the emergence of problems, and the erosion of maintenance covenants gives arrangers and investors less opportunity to intervene and find a solution.

 Chart 7

The most severe single-name credit event would be the default of a benchmark underwriting transaction during the syndication process. Table 1 shows the result of a simple stress test based on this scenario. In spite of lengthening selldown timetables, a default during syndication is highly unlikely as it would occur relatively soon after the bank's internal credit approval. It is a conceivable situation, however, and the most probable trigger would be an idiosyncratic risk such as fraud.

Table 1

Single-Name Credit Event Stress Test
	Exposure (mil. $)	Assumed average LGD (%)	Estimated loss (mil. $)
Loan	1,700	30	510
Bridge to high yield	800	45	360
Total	2,500	35	870
LGD--Loss-given-default.

For an underwriting bank, the effect of generalized spread widening would be most significant if it occurred abruptly while several benchmark transactions were in the syndication process. Table 2 shows the potential impact if three such deals were affected. This is again a severe scenario, but it is plausible given the increased number and size of deals in underwriters' pipelines.

Table 2

Generalized Spread Widening Stress Test
	Exposure (mil. $)	Assumed position sold at par (%)	Assumed average loss on sale of residual position (%)	Estimated loss (mil. $)
Aggregate loans	5100	70	20	306
Aggregate bridges to high yield	2400	50	30	360
Total	7,500	64	23	666

Default stresses are also relevant for the "buy and hold" portfolios maintained by both underwriting and investing banks. As with the other scenarios, the potential exposure depends on the size, rating distribution, sectoral diversity, and granularity of the portfolio, as well as any credit hedging or other offsets. The typical senior leveraged loan rating lies between 'B+' and 'B', and the average three-year cumulative default rates for these rating levels over the past 25 years were 12.0% and 19.5%, respectively. These numbers indicate the potential gross medium term losses in a normal market environment, but there should also be sizable recoveries on senior positions.

These stress tests show that a severe downturn in leveraged finance could lead to material losses that would emerge over a number of months. It must be emphasized that the assumptions we have used are indicative estimates and should not be regarded as definitive. In interpreting the results of these stress tests, we take into account the significant earnings strength and diversity of the leading players in the market.

We can also stress test the potential loss of earnings from a pronounced market downturn. Most banks do not disclose their total leveraged finance revenues, but Credit Suisse Group (A+/Positive/A-1) revealed that it made $2.1 billion in 2006, which was a 50% increase on the prior year and represented 13% of its investment banking division's revenues. The industry pretax margin for leveraged finance is likely to be about 40%-45% and it is therefore a more profitable business line than many others. Using the simple assumption that other banks' revenues relative to Credit Suisse are proportionate to their relative underwriting volumes according to league tables, it is possible to estimate the percentage of recurring groupwide revenues that the major commercial and investment banks earn from leveraged finance (see chart 8). A severe but plausible stress would be a 50% decline in leveraged finance revenues, which is similar to the fall already seen in U.S. subprime RMBS origination. According to our numbers, the worst-affected bank in these circumstances would be Credit Suisse, with a 3.8% fall in group revenues. Although this is a sizable amount, it does not indicate a high dependence on the future direction of the LBO market. Staff compensation forms a large part of the cost base of banks' leveraged finance units, and they should therefore be well placed to adjust expenses reasonably quickly in the event of a fundamental change in revenue prospects.

 Chart 8

We have not to date changed any bank ratings as a result of leveraged finance activity, but we are closely monitoring developments in this market and in related sectors. The central expectation behind our ratings is that the environment will remain soft at least through the summer, and that banks caught with the most aggressively structured or priced transactions when the market turned will suffer relatively modest losses over the coming months, particularly on high-yield bond and junior loan exposures. Underwriters and sponsors have a shared interest in maintaining a smooth market, however, and will continue to work together as far as practical to place the existing pipeline. Although the golden age of leveraged finance volumes and returns may now be at an end, our revenue stress test suggests that the major banks can handle a return to normality. Underwriting criteria also show a more sustainable picture, with borrowers' new issuance debt leverage about 1x to 1.5x EBITDA lower than a few weeks ago. Similarly, the death knell appears to have been sounded for cov-lites, PIKs, and possibly equity bridges, and this is a welcome development for the banking industry going forward. Corporate defaults will certainly rise from the current low level and banks must therefore pay close attention to basic counterparty credit risk analysis. The macroeconomy is the ultimate determining factor and, although there are sectoral concerns, particularly in the U.S., the outlook remains broadly supportive.

We are not complacent, however. In particular, we see the risk of a harder landing in the economy, and of further spillover into leveraged finance from other markets. The major banks could handle the outcome of our risk exposure stresses at their current rating levels, but ratings could be pressured if this event coincided with general weakness across the credit space. Equity markets may also be affected if the changes we are seeing in leveraged finance cause sponsors' M&A activity to slow materially. We will therefore continue to pay close attention as the current market correction evolves.