Contents

CDR Liquid Index data as at 19 February 2007

Source: Credit Derivatives Research

CDR Liquid Indices
The CDR Liquid indices represent the CDS levels of the most-liquid names in their respective markets and ratings classes. Liquidity is measured by the number of bid-offers a credit receives. Index values are simple averages of on-the-run five year CDS levels.

CDR Global Market Depth™
The CDR Global Market Depth Index is a daily measure of how many names are actively traded. Liquidity is measured by the number of bid-offers a credit receives. Index values are counts of the number of names that exceed CDR's Liquidity Floor.

CDR Global Market Activity™
The CDR Global Market Activity Index is a daily measure of activity within the global CDS market. Liquidity is measured by the number of bid-offers a credit receives. Index values are simple averages of total bid-offers of all names that exceed CDR's Liquidity Floor multiplied by CDR's Global Base Liquidity Constant.

European LCDS re-think expected

Market practices mean US blueprint cannot be utilised

The small sub-group of dealers searching for an initial compromise on European LCDS documentation (SCI passim) is expected to give up its search this week. However, a longer-term project in this area is set to continue.

The move is unlikely to be welcomed by market participants who have been hoping for a strong compromise solution to attract all potential users to the LCDS market. A source close to the discussions says that, despite increasing external pressure for resolution, a way forward has not been found.

During their twice-weekly meetings over the past month or so, members of the sub-group are understood to have been able to reach agreement on many of the key issues but ultimately do not believe that they will be able to agree completely on everything. Consequently, the sub-group has finally concluded that it will be unable to achieve its stated aim of producing a draft template for a non-cancellable, reference entity-based loan CDS which is acceptable to all the members of the sub-group and could be presented to a broader European LCDS working group for discussion and approval.

The stumbling block is thought to be the greater levels of privacy in the European loan markets compared to the US. The intention of the European LCDS project as a whole had been to produce documentation based on ISDA's US form of Syndicated Secured Loan Credit Default Swap Standard Terms, but in light of the differences in market practice it was felt by the sub-group that the US-style contract would not be usable at the present time.

The sub-group is, however, understood to be unanimously in favour of the European Loan CDS project being continued and remaining a medium-term goal for ISDA and the broader working group. In the meantime, the sub-group has agreed that it would support and work on an evolution of the current European cancellable, reference obligation-based template that has been produced by Morgan Stanley. This template is a single-name version of the standard terms supplement that is used in relation to trading the LevX index.

The dealer sub-group is believed to be already discussing necessary changes to the Morgan Stanley supplement, including the possibilities for introducing an option for non-cancellability following a refinancing of the reference obligation. It is expected that the sub-group will report progress to the wider working group within the next two weeks.

MP

Further European CRE CDOs edge closer

Signs of life for sector as at least four deals get underway

There are at last signs of movement in the European CRE CDO market, nearly four months after the launch of the debut deal – the €342.5m Anthracite European CRE CDO 2006-1, brought by Morgan Stanley and managed by Blackrock Financial Management (see SCI issue 13). Four or possibly five deals are currently being prepped and the first to emerge is now expected in the reasonably near future.

Bear Stearns, Lehman Brothers and Prudential M&G are known to be working on CRE CDOs – with the latter believed to have begun beauty parading banks to lead its deal this week. Other names that are touted as potentially being involved in the early stages of prepping another new structure include Bank of America and Wachovia.

But progress on all of these transactions is being hampered by issues at the rating agencies, according to market sources. "One of the problems has been a lack of clear methodology from the rating agencies. Both Moody's and Fitch now appear to have formalised a methodology – it's still evolving a little bit, but they both have a pretty set methodology that works. S&P is definitely still a work in progress, with deals being looked at more by the CMBS group in conjunction with the CDO group – whereas the others are running the process out of their CDO group," says one.

"It is a slightly unusual situation in that S&P would lend itself more to a CRE CDO, given its greater coverage of the CMBS markets. But the unclear methodology is making people a little bit nervous about S&P," adds the source.

Most obviously affected by a slow-down in the development of the European CRE CDO sector is the Bear Stearns deal on behalf of Investec. The transaction was expected to be close behind Anthracite and brought at the end of last year. It is, however, thought to be still around four weeks away from launch.

The deal is believed to have experienced additional delays at the rating agencies specific to its unusual make-up. The manager is understood to be aiming to mitigate any possible investor concerns over potential future market illiquidity by offering the added attraction of rated underlying collateral. This will involve the inclusion of whole loan mortgage assets in the portfolio.

The two firms have already been in the market over the past week with a separate CDO of ABS – Pangaea ABS 2007-1. The €309.2m transaction is a managed cash arbitrage securitisation of structured finance assets, primarily mortgage-backed securities (RMBS and CMBS), CDOs, and to a lesser extent, commercial and consumer ABS. The portfolio also includes non-conforming or non-performing loan securitisations, whole-business securitisations and SME loan securitisations.

The deal comprises five rated tranches from triple-A through to double-B minus, along with an unrated subordinate piece of €12.2m, representing about 4% of the deal.

MP

Trading in anticipation

Dealers and investors positioning for expected tranche and CDS flows

The potential launch of as many as 24 credit derivative product companies (CDPCs) by the end of this year (see last week's News Round-up) is precipitating market activity in advance of the expected considerable volume of flows they will produce. The more immediate issue of the next index roll is more of a focus, however.

The launch of more CDPCs is likely to bring a large amount of senior protection selling in the 7y and 10y sectors to the market in both tranches and CDS. Some traders are working on the theory that this could produce some significant opportunities, particularly in the current overall tight spread environment. A number of possible kinds of trades have been mooted to take advantage of such flows, but so far 'keeping it simple' appears to be seen as the best option.

As one trader observes: "Most of the more complex ideas that people are suggesting have so far been seen as being too structured, given the unknown timeframe for significant CDPC flows to kick off. A few step-up trades have been put on, for example, but not a great deal. There are, however, signs of investors trying to get in front of CDPC flows by buying protection on mezzanine tranches, particularly in the 10y to put on some cheap shorts. Equally, we have also seen them selling 10y senior mezz to, again, pre-position for CDPC-type flows."

Another dealer suggests that such trades are currently less of a focus for many market participants. He says: "A lot of focus is on the index roll on 20 March, with people thinking about what names are coming out and what names are going in. Now that we are heading into the roll period, most of the focus is on structured credit products which take exposure to the indices or on CSOs and CPPI products that take single name synthetic exposure. As we move to the on-the-run series on iTraxx and a new maturity date for liquid CDS, people start moving their positions out of the old series into new positions."

That said, he notes: "It's probably going to be a fairly uneventful roll, as it was last September. Now people are used to what happens and understand what's going on much better than they did with the first couple of index rolls, so fingers crossed it should go smoothly."

One possible fly in the ointment that was touted some months ago is the potentially large impact of CPDOs on roll activity. The dealer says: "A few months ago the street was positioning for the impact of CPDOs on the market, but based on the theory that the instruments would take off massively. They have, of course, been a reasonable success and a fair amount have been printed. But now it's a case of seeing what actually happens, as this is the first roll period since CPDOs hit the market."

MP

CPDOs stay on hold

Ratings methodology still under review, but end in sight

Fifteen banks have now applied for ratings on CPDO and CPDO variant products. The rating agencies are not yet able to respond as they continue to work on second generation methodologies, but market participants hope that new versions could be available by the end of the month.

The market remains in the same situation as it has been since last autumn. Following the launch of the first few deals, the rating agencies have effectively suspended CPDO issuance while they revise their approaches to try and address some of the potential structural weaknesses that their original methodologies and models didn't fully capture.

The agencies are understood to be making changes to the assumptions in their first generation CPDO methodologies, including those made in connection with spread volatility, the term structure of spreads and bid-offers. At the same time, they are trying to generate methodologies to deal with new variations on the CPDO structure, such as those referencing single names as well as indices and managed deals (see SCI issue 24).

Overall, the intention is to create a ratings methodology that is much more robust, but inevitably that will mean that returns will be lower than on first generation CPDOs, especially in the current tight spread environment. Nevertheless, dealers are sanguine about the changes.

Paul Levy, head of exotic credit structuring, EMEA, at Merrill Lynch, observes: "The development process of CPDOs is mirroring a lot of the development processes of leveraged super senior product. Essentially a good idea was created and was pursued as quickly as possible because the market commoditises ideas so quickly."

In the case of leveraged super senior (LSS), some firms did push ahead, but models have had to subsequently change. "The criteria for LSS that exists today are quite different to back in late 2004 when the product was originally being constructed," Levy says.

The same has happened with CPDOs, he argues. "Even before the first transaction had been closed we had examined carefully the methodology being applied to CPDOs and asked the rating agencies questions about some of the assumptions we saw implicit in the product," explains Levy

As a result, Merrill Lynch has been working closely with the rating agencies ever since, particularly focusing around the quantitative aspects of the methodology. In light of the firm's knowledge of the evolving situation with CPDO ratings methodology, it decided not to pursue its first generation CPDO product – REDI.

Paul Horvath, global head of structured credit origination, structuring and distribution at Merrill Lynch, explains: "We recognised before anybody else that the rating agencies were going to change their methodology. As a market leader – but I would argue that this should apply to any participant – we were not going to arbitrage our investors or the rating agencies by rushing into a product based on what was already an old model."

He continues: "We are very much committed to the CPDO product but committed to doing it in a stable manner. Once the new methodologies have been put in place, Merrill Lynch will be at the forefront of CPDO issuance."

MP

ABX tranches get nervous welcome

Buy-side stays on the sidelines amid correlation uncertainty

The meltdown in ABX spreads and resultant wide trading levels have spooked many in the market (see last week's SCI). Given this backdrop, it is no surprise that a nervous welcome met the launch of TABX – standardised tranches on ABX.HE indices – as many investors preferred to remain on the sidelines.

"We have thankfully avoided that mess," says one London-based fund manager. "There are too many challenges and I would be very surprised if it takes off," adds another portfolio manager at a London investment manager.

Jeroen Bakker, managing partner for the structured products business at Faxtor in Amsterdam suggests there could be a variety of reasons why activity has not taken off that quickly. "Given the state of the US mortgage industry and housing market, many are taking a wait-and-see approach, while others have not started trading in the TABX as their back-office systems are not yet up to scratch," he says.

More importantly, uptake has been hampered by a continuing lack of consensus over how to gauge ABS correlation. "People need to get used to how these tranches trade, and implicitly what kind of correlation you are trading and how correlated you expect the underlying reference obligations to be," says Hans Starrenburg, head of European structured credit, NIBC Credit Management in The Hague.

The portfolio manager is keen to see tranching develop but warns that this could prove difficult as correlation models are almost impossible to apply to TABX. "There is no concept like default, maturity or yield curve, it is not as straightforward as in the corporate market," he says. "Most dealers are quoting very rough levels and have no idea how to price the tranches. Investors need to find the delta, but the underlying ABX is trading too wide so it is almost impossible to hedge."

Many are still doing their homework before they step in and those that have seem to be playing it relatively safe by focussing their activity on the senior tranches. "At first glance, the senior tranches look a little bit more interesting at the moment, if you are not a very risky investor. In my view, being conservative and using the less risky tranches to gauge their behaviour before possibly moving down the capital structure is an arguable approach," says Starrenburg.

Bakker adds that the ABS CDO market may hold greater favour for many, as TABX only references 40 names compared to a typical CDO of ABS, which contains between 80 and 100 names. "Also, a manager in an ABS CDO can trade out of problems in case they arise and the excess spread in transactions serves as a loss protection, which is not available in TABX," he says.

Using the tranche market is more work than is currently necessary for many fund managers who have their hands full managing positions, tracking the ABX market, while analysing the performance of underlying transactions and which constituents are in the index. "There is enough volatility and juice in the underlyings of the ABX index to undertake interesting trades, so you do not need tranches yet, especially given the wide bid-offer. We probably need some improvement in pricing before it becomes interesting to use. The TABX does form a nice source of additional information though, for example, as a reference for mezzanine ABS CDO," says NIBC's Starrenburg.

Despite initial teething troubles, some in the market believe TABX will be beneficial to their business. "As a fund of funds we are not in the front line, though we do look forward to it, if it is a success. The main benefit to us, as with all liquidity increasing events in the market, would be that the funds we are looking at would have more opportunities to hedge their investments and also profit using TABX," says Erik Larsson, head of research at AlphaSwiss Asset Management in Zurich.

HD

Avendis appoints Moysan

The latest company and people moves

Avendis appoints Moysan
Avendis Investment Management has appointed Raphaelle Moysan as director of marketing and structuring in London. Moysan was previously director of marketing at credit-specialist asset manager Credaris.

Before joining Credaris, Moysan worked at Goldman Sachs, Merrill Lynch and JP Morgan in a range of fixed income derivatives marketing roles.

Lynch moves to Hong Kong
Patrick Lynch, head of credit trading at Morgan Stanley in London, is moving to a new role in Hong Kong next month, to run credit in Asia. His position will not be replaced and the existing management roles under Lynch will remain unchanged with the incumbent heads of desks reporting to Armins Rusis, head of credit in Europe.

Duarte joins WestLB
Alejandro Duarte has joined WestLB as a director reporting to Efstathios Margonis in the structured credit trading team. Duarte was previously an analyst in the quantitative modelling team at ABN Amro.

Regan leaves ABN
James Regan, head of ABS trading at ABN Amro in New York has left the firm. Geoffrey Chang moves internally to take over his trading responsibilities.

Nomura hires Scherr
Andrew Scherr has joined Nomura Securities International (NSI) as director, fund-of-fund origination in the fixed income division. Based in New York, Scherr is responsible for providing financing and investment solutions to hedge funds, fund of hedge funds, family offices and institutions.

He reports to Dean Chamberlain, md and head of fixed income sales at NSI. Prior to joining NSI, Scherr served for two years at Fortis Prime Fund Solutions (USA), where - as director of structured products - he led the fund-linked structured products business for the Western Hemisphere.

Correction
Part of the 'Internal moves at Bank of America' story in last week's job swaps section should have read: Nik Dhanani has moved from capital markets - where he covered hybrids - to trade bank and finance in both cash and CDS.

HD & MP

Fitch calls for increased competition

A round up of this week's structured credit news

Fitch calls for increased competition
Fitch Ratings has sent an open letter to arrangers, asset managers, issuers and investors encouraging them to respond to proposed new rules for rating agencies issued by the US Securities and Exchange Commission.

The Fitch letter says that in its proposed rules, the SEC found that "'notching' – the practice by which a rating agency punitively lowers the rating on an underlying security that it did not originally rate – is unfair, coercive and abusive and proposed prohibiting the practice, subject to one notable exception. Thus, the requirements by which Moody's and S&P rate structured credit products such as CDOs, SIVs, and Money Market Funds will have to change."

The agency argues that the new rules do not go far enough to meet the desired goal of a competitive marketplace. "The proposed rule, while it prohibits notching, essentially creates a 15% bucket for unnotched ratings by allowing a rating agency to refuse to rate any transaction if it rates less that 85% of the market value of the assets underlying the structured product," the letter explains.

Fitch says it believes that the proposed exception will allow the largest rating agencies to continue to suppress competition by compelling asset managers of CDOs, SIVs, money market funds and other structured vehicles to buy securities that carry their ratings. "There is no analytic justification for the proposed 85% threshold as Fitch and the other two major agencies have substantially similar ratings levels and performance trends," it says.

So, Fitch argues: "We believe that asset managers should be free to choose among securities rated by any rating agency without fear or uncertainty as to whether or not a rating agency will agree to rate their product. Therefore we believe the best policy for the SEC to promulgate is one that eliminates the requirement that an agency rate such a high percentage of underlying assets. We advocate a 66% threshold, which is the equivalent of a super majority."

For its part, S&P says: "In order for us to rate a security backed by a pool of assets, we must independently review the credit quality of that pool of assets. To insist that we be required to rely on the judgment of other rating agencies when evaluating the pool of assets is contrary to our commitment to providing the marketplace with objective, independent, high-quality ratings."

Moody's says that it too wishes to continue to rely on its own ratings. "We believe ratings produced by different rating agencies have different meanings and we believe that the market values this diversity of opinion. So any rules that would force an NSRO to treat each others ratings as interchangeable would limit that diversity of opinion," the agency explains.

SEC issued proposed rules on 2 February aimed at implementing the Rating Agency Reform Act of 2006 enacted on September 29, 2006 (see SCI issue 9). SEC is soliciting comments on the proposed rules until March 12 and Fitch, Moody's and S&P have all said they will be filing comprehensive comment letters by then.

Innovative synthetic duo launched
Two unusual tranched structures hit the market last week, both involving derivative technology rarely seen in European deals. Fortis priced Astir (Isara) – which references PAUG CDS – while Merrill Lynch's Edelweiss Capital includes equity default swaps.

Astir (Isara) is a high-grade ABS CDO that references a €1.5bn portfolio of CDS that utilise the pay-as-you-go (PAUG) template. Credit events under the transaction are bankruptcy, failure to pay principal, writedown and distressed ratings downgrade.

Capital structure comprises €150m triple-A Class A1 notes which priced at 15bp over Euribor, €30m triple-A rated Class A2s at 35bp and €30m single-A rated Class Bs at 100bp. Rated by Standard & Poor's and Moody's, all tranches have a seven-year WAL.

The portfolio consists of 204 predominantly RMBS exposures, with US securities making up 53% of the total and European names accounting for the remainder.

As well as arranging the deal, Fortis acts as CDS counterparty, repo counterparty, interest rate swap provider and portfolio manager. The deal is callable by the issuer quarterly from January 2010 and annually from January 2014. Taxation call, together with the termination of the interest rate swap, CDS or repo funding agreement is another source of possible early termination for the transaction.

The notes may be partially redeemed pro rata after the reinvestment period, subject to S&P's SROC test and Moody's CDOROM test, and the class B subordination may not be decrease by more than half of its initial amount.

Meanwhile, Edelweiss Capital Series 2007-1 comprises US$10m A2 Moody's-rated Class A notes, US$3m Baa2 Class B notes and a US$75m Baa2 equity default swap (EDS) between Merrill Lynch and an investor. This unfunded swap is related to the same theoretical tranche as the Class B notes.

The ratings address the expected loss posed to investors by the legal final maturity in 2013. In the case of the swap, the expected loss is calculated as a percentage of the maximum exposure under the contract. The transaction involves Edelweiss selling protection on a portfolio of EDS while buying protection on another portfolio of EDS.

An EDS is a derivative contract between two parties, a protection buyer and a protection seller, whereby the former has the right – in exchange for the regular payment of a premium – to receive from the latter a protection amount in case some form of trigger event occurs. One example of a trigger event is if the reference entity's stock price falls below a certain percentage of its initial value. The protection amount is set at a fixed percentage of the reference entity's notional amount.

In this transaction the EDS contracts' trigger events stem from equity-linked events and are defined as a drop of a reference entity's stock price below 35% of the stock's initial value. Furthermore, equity default events may not occur at any time other than at the weekly observations carried out during the last three years of the trade, with the protection amount paid as a function of how many times the stock price trades below the strike.

Moody's finalises US CLO methodology
Moody's has published its final report on changes to its US cashflow CLO rating methodology. The changes incorporate Moody's new probability-of-default ratings (PDRs) and loss-given-default assessments (LGDs) for speculative-grade corporate issuers.

Moody's will begin applying these changes to CLOs that close after May 31 2007. Before that date, CLOs may elect to adopt the changes or continue to apply Moody's current methodology.

Incorporating these measures brings new precision to its CLO ratings analysis, according to the rating agency. Under the new methodology, CLOs will use Moody's PDRs and LGDs for each debt instrument in the CLO's collateral pool. Historically, CLOs used Moody's instrument-level expected loss ratings to make general assumptions about default probability and recovery.

Moody's does not expect its new methodology to result in materially different ratings when applied either to existing CLOs or to new CLOs. For CLO surveillance, the agency indicated that it would initially test CDOs under both methodologies to verify continuity between the two approaches.

The proposed changes will not affect Moody's European CLO rating methodology. It expects to publish a report describing its approach to incorporating the PDR/LGD initiative into its European CLO rating methodology as the PDR/LGD initiative is implemented in Europe. In the meantime, Moody's will continue to use its Recovery Rate Toolkit to analyse CLOs with substantial non-US exposure.

Deutsche announces upsized closure of EM CLO
Deutsche Bank's Loan Exposure Management Group (LEMG) doubled the size of its CRAFT EM CLO 2006-1 at closure last week from US$500m to US$1bn. The execution of the upsized CLO augments an already existing six-year risk distribution platform, which is utilised to hedge highly illiquid emerging market exposures on the bank's balance sheet.

As a result of the tap, an additional US$60m of notes were sold into the market, consisting of equity and mezzanine (double-B and triple-B) tranches of the capital structure. The structure includes an innovative feature which lowers hedge costs through a fixed recovery rate mechanism.

The upsized US$1bn platform is expected to further facilitate new lending by Deutsche Bank to local companies domiciled in emerging markets, and to increase derivative capacity in these markets.

S&P takes rating actions on European synthetic CDOs
Standard & Poor's says that it has taken credit rating actions on 32 European synthetic CDO tranches. Specifically:

• Ratings on nine tranches were removed from credit watch negative and lowered.
• The rating on one tranche was lowered and remains on credit watch negative.
• The rating on one tranche was removed from credit watch negative and affirmed.
• Ratings on 18 tranches were removed from credit watch positive and raised.
• Ratings on three tranches were raised.

These rating actions and the credit watch updates follow two reviews. The first review was of credit watch placements made on 14 February after S&P ran its January 2007 month-end SROC (synthetic rated overcollateralisation) figures. The second review was of the ratings on tranches that have been on credit watch negative for more than 90 days.

MP

Trading ideas - is the bell tolling?

Dave Klein, research analyst at Credit Derivatives Research, looks at a capital structure arbitrage trade on Toll Brothers Inc

Credit and equity risk are unambiguously linked as the risk of debt holders not receiving their claims is akin to the risk of equity prices falling to zero. Both credit and equity risk are directly tradable with liquid instruments such as CDS and equity puts.

We continue to analyse hedging CDS with specific (deep out-of-the-money) equity put options. (Please also see CDR's Trading Technique article – Credit Risk Models – on the firm's website, which provides significantly more detail on this relationship.)

The overall logic is that owning a corporate bond is equivalent to owning a Treasury bill and writing a put option to the issuer. This put option gives the firm's equity holders the right to default on or before maturity of the debt, in exchange for transferring ownership of the firm's residual assets to debt holders.

As perceived credit risk rises, we would expect firm volatility to rise (this would increase the value of the put option) thus increasing the likelihood that equity holders would default on the company's debt. The payoff profiles of an out-of-the-money equity put and a CDS are very similar and thus misalignments in that relationship are potentially profitable.

The below trade on Toll Brothers Inc (TOL) takes advantage of this relationship by selling equity puts and buying CDS protection.

Delving into the data
When considering market pricing across the capital structure, we compare equity prices and equity-implied volatilities to credit market spreads. There are a number of ways of accomplishing this, including the use of structural models that imply credit spreads (through an option-theoretic relationship) from equity prices and the analysis of empirical (historical) relationships between the two markets.

The first step when screening names for potential trades is to look at where equity, equity options (implied volatility) and credit spreads stand in comparison to their historic levels. Exhibit 1 shows the past year's levels for TOL's stock (inverted scale), put implied-volatility and CDS spreads. Notice the high correlation between the CDS levels and the inverted stock price.

Exhibit 1

Comparing market levels over time give a rough feeling for how each security moves in relation to the others. In order to judge actual richness or cheapness, we rely on a fair value model. Given that our trade is a combination of CDS and equity puts, we consider the empirical relationship between CDS and equity-implied volatility.

Exhibit 2 plots equity-implied volatility (y-axis) vs. 5Y CDS premia (x-axis). If the current levels fall below the trend line, then we view CDS as too rich and/or equity puts as too cheap. Above the trend line, the opposite relationship holds. At current levels, TOL is very close to fairly priced.

Exhibit 2

We next consider this relationship over time, finding a "fair value" for implied volatility based on the CDS level. We do this for two reasons.

First, we want to assure ourselves that the implied volatility is not too low to fair value. Second, we want to see how often implied volatility and CDS converge to fair value.

Exhibit 3 shows the trend for TOL over time. We note that TOL implied vol has dropped significantly this year and that implied vol has consistently moved above and below fair value. Although TOL has recently moved back to fair value, the fact that the implied vol has traded both above and below fair value increases our confidence in the CDS-Vol relationship.

Exhibit 3

Hedging default risk with equity puts
Our analysis so far has pointed to a potential misalignment between the equity-implied-volatility markets and credit spreads of TOL. It would appear that we should buy protection (sell credit) against a short position in equity volatility. The most suitable way to take this position is to trade CDS against Equity Puts.

As default approaches, we see CDS rates increase (to points upfront) and equity prices fall close to zero. In this situation, an equity put will increase in value as CDS rates increase (and equity prices fall). If we expect equity volatility to rise also, then this will be a double-whammy for the put option value.

However, when a company finds itself under increasing shareholder pressure, credit spreads can widen while share prices increase. This is good news on both legs of a short credit/short equity put trade, and we believe this is the situation in which TOL finds itself.

If we can buy CDS protection cheaper than the relative costs of selling equity puts, then we can position ourselves correctly for the trade and benefit from positive carry for the life of the option. This is what makes capital structure arbitrage opportunities difficult – finding the right relationship misalignment with the right economics.

TOL's current price levels across the capital structure, as seen in Exhibit 4, offer us exactly the right combination of positive carry (in the longer-dated options), hedged default risk, and long equity option, short CDS. Translating the cost of our puts into basis points, we find that the trade has an overall equivalent carry of 103bp for the first two years and positive present value. We have omitted the details of our calculations in this report but will happily share them with interested clients.

Exhibit 4

The main trade risk here is that the stock price could fall below the put strike without the occurrence of a credit event that triggers a CDS liability. Thus the holder of the long CDS/short put position could potentially be liable for a put pay-off without a proportional change in CDS value.

In our case we like the positive carry of the hedged position of around 103 basis points until the option expiry. The mismatch in maturities (three-year CDS versus two-year hedge) also provides us with a natural short forward credit position. This position implies a two-year forward one year long protection position.

We therefore recommend buying protection in three-year CDS and selling longer-dated LEAPs (put options) on TOL for a 103 basis point pick up in carry, hedged default risk, and long forward position.

Risk analysis
This position does carry a number of very specific risks.

Recovery and 'Default' Stock Price Assumption: In the default scenario the cheapest-to-deliver CDS obligation may have a lower than expected market value and the stock price might not fall to US$0.75 as assumed.

CDS and Equity Put Timing Mismatch: When the CDS durations exceed that of the put, investors could be exposed to an un-hedged CDS protection payment liability. However, shorter dated CDS may be available that can be tailored to the equity put expiration and/or longer dated puts may be offered in the over-the counter equity options market.
Alternatively, at expiry or over the duration of the put contract, the short CDS position may be re-hedged with puts of later expiration. Moreover, the trade may possibly be unwound profitably before the put expiry.

CDS Present Value (PV): The CDS PV is an expected value, but not necessarily a realised outcome. In practice, the CDS may trade on an up-front and/or running basis.

Corporate Actions: Spin-offs and Private Equity Buyouts, for instance, could force an early settlement of the equity puts, leaving investors with un-hedged long CDS positions.

Mark-to-Market: In our view, credit and equity derivatives markets operate largely independently and this can lead to trade opportunities. At the same time, however, any relative mis-pricing may persist and even further increase, which could lead to substantial return fluctuations.

Overall, frequent re-hedging of this position is not critical but the investor must be aware of the risks above and balance that with the strong positive carry with minimal default risk over the options life.

Liquidity
Liquidity is a major driver of any longer-dated trade – i.e. the ability to transact effectively across the bid-offer spread in the bond and CDS markets. Our data on liquidity, created from the volume of bids, offers, and trades we see each day, provide us with significant comfort in both the ability to enter a trade in TOL and the bid-offer spread costs. TOL is a highly-liquid name, even in the three year maturity and bid-offer spreads are narrowing to around five basis points.

Equity puts are available through LEAPS and from analysing historical volumes it seems trades can be done in reasonable size. A potential alternative is to transact in an OTC trade rather than through the exchange to better tailor positions. The strike we have chosen has had volume recently. We do note that there has been a great deal of activity today in the US$15 strike which may be an easier place to trade (although we would recommend selling 42,100 puts).

Fundamentals
This trade is somewhat impacted by the fundamentals. We need to consider both TOL from a credit perspective as well as shareholder pressures on the name.

Kathleen Shanley, Gimme Credit's Homebuilders expert, maintains a deteriorating fundamental outlook for TOL. She notes that TOL's recent report on preliminary results were gloomy with falling revenues but slightly improved cancellation rates. As a credit, Kathleen believes that TOL is still susceptible to the housing bubble as well as shareholder pressure.

We have heard rumours that TOL is in talks with private equity. The market seems to have reacted to this news with a sell-off in the CDS and increased volume in out-of-the-money puts. We note that, if the rumours are substantiated and TOL follows the typical pattern for LBO credits, we should expect a further widening of CDS and a rally in equity. This is good news for both legs of our trade.

Summary and trade recommendation
As investors view homebuilders in a better light, they become more and more attractive from a private equity perspective (as we have discussed in the past, our LBO viability screen has many homebuilders in the top 50 most-likely). With spreads still tight combined with today's rumours on TOL, we believe this is an excellent time to put on a capital structure arbitrage trade, betting that equity will outperform credit. While it is dangerous to trade on rumours, especially after the market has already moved, the economics of today's capital structure arbitrage trade are strong enough to warrant the risk. Given our deteriorating fundamental outlook for TOL, the trade is a good way to go short the credit without taking an outright position.

Buy US$1m notional Toll Brothers Inc 3-Year CDS at 34bp

Sell 31,200 Toll Brothers Inc Jan-09 20 Strike Equity Puts (312 lots) at a cost of US$24,960 (250bp of notional or 137bp of CDS DV01 equivalent) to capture 103bp of equivalent carry to option maturity.

For more information and regular updates on this trade idea go to: www.creditresearch.com

Copyright © 2007 Credit Derivatives Research LLC. All Rights Reserved.

Note: This article is intended for general information and use and does not constitute trading advice from Structured Credit Investor (see also terms and conditions, below, section 12).

CLO Overlapping

The issue of overlapping within a multiple CLO portfolio is tackled by Priya Shah and Domenico Picone of the structured credit research team at Dresdner Kleinwort

It is important for investors to have an understanding of the combined exposure held in the different names in their portfolio in order to be comfortable with the risk involved in the event of a default or rating action. We therefore propose an algorithm to calculate a risk weighted combined exposure to individual securities.

Background
Over the last few years, the popularity for holding CLOs as part of an investment portfolio has increased substantially. CLOs, in addition to strengthening the overall portfolio diversification, also provide attractive returns in a relatively low risk and volatility environment. The tranching of the capital structure enables investors to purchase CLO tranches to suit their own particular risk/return preferences, and is therefore appealing to a wide range of different investors.

As the popularity of CLOs has increased the number of new CLO issues has also risen, driven in part by existing CLO managers issuing new deals but also by new managers eagerly entering the market. On average, most CLOs reference between 50 to 100 different leveraged loans, although some are more granular than others, and whilst the supply of leveraged loans has grown the increased competition from the various CLO managers and other investors, combined with the current high levels of prepayments, has put some strains on supply.

As a result, investors holding multiple CLO tranches may have some overlapping exposure to various popular names. We therefore believe that an investor should have an understanding of the combined exposure held in the different names in their portfolio in order to be comfortable with the risk involved in the event of a default or rating action by one of these securities. In particular, the extent of overlapping should be noted by multiple equity and junior mezzanine tranche holders who are primarily affected by the first default losses.

Below, we initially outline some results of two overlapping surveys undertaken by S&P, one focusing on the US CLO market and another on the synthetic CDO market, where overlapping should also be considered. We then detail how combined exposure to individual names can be calculated, taking account of the tranche purchased and the capital structure of the CLO. Some methods of mitigating the impact of any overlapping are then detailed towards the end of this section.

Some surveys

US CLOs
In April 2005, S&P published a survey titled 'CDO Spotlight: Is overlap affecting diversity in the US CLO market?', which detailed the results of an overlapping study on US CLO managers. In total, 15 US CLO managers were reviewed with a total of 67 cash flow CLO issuance between them.

The key findings of this study were:

? The average pair-wise overlap – calculated as the (par balance of the shared name) / (par balance of the CLO portfolio) – within CLOs issued by the same manager was roughly 50% with a range of between 29% and 73%.
? Within CLOs issued by the same manager overlap was greatest between CLOs issued within one year of each other.
? The study found that the average pair-wise overlap between two different CLO managers was approximately 35%.
? CLOs ramping up in the same period are also more likely to share similar collateral, and one would expect a higher overlap between CLOs of the same vintage but different managers. Although this was investigated by S&P, the small sample size available resulted in non-conclusive findings.

This survey was based on CLOs issued by August 2004 since which time the CLO market has evolved and grown significantly. The numerical survey finding may therefore not be totally accurate for today's CLO issuance. However, it does highlight the importance of monitoring the extent of the overlap in an investor's portfolio.

Synthetic CDOs
A second overlapping study undertaken by S&P in September 2005 focused on synthetic CDOs, titled 'CDO Spotlight: Overlap between reference portfolios sets synthetic CDOs apart'.

In contrast to some other structured asset classes, such as consumer ABS, synthetic CDO portfolios are also exposed to a degree of overlapping. Many synthetic CDOs are bespoke trades; however, they differ only marginally to standardised index tranches and therefore often refer to a number of high quality securities which constitute the index. As the credit index is comprised of only a limited number of names this leads to a degree of overlap between synthetic CDO transactions.

In its survey S&P analysed European synthetic CDO transactions, by focusing on the most commonly held credit securities. The study found that the top three securities were referenced in more than 80% of transactions whilst each of the top 40 names appeared in more then 50% of issued CDOs. Alternatively, by looking at the full capital structure of a typical transaction, they found that on average each credit security was also referenced in 29% of the other transactions.

Therefore, as with CLOs, it is important for an investor holding multiple synthetic CDO tranches to regularly monitor the actual exposure to individual names, to ensure that the impact of deterioration in one or more of these names is sustainable.

Calculating exposure to individual names

An investor in his/her portfolio may hold several CLO tranches, credit funds and even on occasion direct leveraged loan investments, giving exposure to several different credit securities. As the number of different investments rises there in an increased probability that some names will appear more than once, causing some overlapping. In the chart below we show a hypothetical investor's exposure to shared names via different investment vehicles including multiple CLO tranches.

A standard approach for calculating the investor's overall allocation to any security i is simply:

where:
Wi = Overall portfolio allocation to security i
Pj = Investor's allocation to portfolio j
Sij = Portfolio j's allocation to security i

Although this approach is appropriate for direct or credit fund leveraged loan investments it is not fully accurate when investment is through a CLO. For CLOs the tranched capital structure causes some distortion to the overall risk undertaken.

Depending on which tranche is held, the investor is exposed to different degrees of risk to the underlying collateral; the equity tranche, for example, bears the first losses and is therefore more at risk from an increased exposure to an individual name on default, whilst a senior tranche holder benefits from a higher degree of credit enhancement and is not as concerned. It is important therefore to take this sort of risk differential into account.

Below we propose an algorithm which may be adopted by investors in order to calculate their risk allocation to the individual names held through their CLO investments. The algorithm accounts for the dependency on the tranche held by taking into consideration the level of credit enhancement offered by more junior tranches and their width.

The proposed approach places more weight on thinner tranches with lower credit enhancement relative to more senior, wider tranches. The algorithm is followed by an example showing the application of the various steps to a hypothetical investor portfolio.

Calculation algorithm
? Step 1: For each different CLO tranche held in the investor's portfolio current details on the total CLO and individual tranches sizes are needed in order to calculate the credit enhancement and tranche width impact: 

Why credit enhancement and tranche width?
Credit enhancement gives an indication of the level of subordination offered by more junior tranches. Over time if defaults occur the losses are first borne by the equity tranche holder, but senior tranches are also impacted as a result of a drop in their credit enhancement.

For tranches with zero or little credit enhancement, losses have a leveraged impact, the extent of which is determined by the width of the tranche held. The smaller the tranche width the more levered the exposure and hence the bigger the impact of any loss.

Credit enhancement and tranche width are therefore important when trying to identify the varying risk profiles of the different tranches and hence we propose to use them as weights when estimating the risk borne to a particular security.

Synthetic CDOs
For synthetic CDOs, the tranche's detachment point indicates the impact of the tranche width, whilst the tranche's attachment point is equivalent to the level of credit enhancement.

? Step 2: The next step is to identify the number of different securities held through the various investments, and the CLO's allocation to each name. For each different security i held, once the CLO portfolios that it is held in have been identified the resultant risk exposure can be calculated as:

Where, as before:
Wi* = Overall, risk-adjusted, portfolio allocation to security i
Pj = Investor's allocation to portfolio j
Sij = Portfolio j's allocation to security i
and an extra weighting factor of

By dividing by the tranche width impact factor, greater weight is placed on smaller tranches which are more levered and hence more risky, whilst (1-CE)² places more weight on the junior more riskier tranches. As the impact of credit enhancement is not linear we propose to use a squared factor. Over time, if defaults occur in a particular CLO, junior tranche width and overall credit enhancement decreases, thereby increasing the risky allocation to all the securities held in that particular portfolio.

Current estimates for credit enhancement and tranche width should be used. It should be noted that similarly rated tranches issued by different CLO managers may have different values for the weighting factor, depending on the overall capital structure. When exposure is either through direct investment or a credit fund the credit enhancement is zero whilst the tranche width is one and therefore the formulae simplifies to the standard form.

? Step 3: This type of weighted allocation needs to be calculated for all individually securities held across the various investments. However, one impact of using such a weighting system is that the total exposure, summed across the individual names, is now no longer equal to 100%,

The final step therefore involves rebasing this exposure to 100% whilst still maintaining the calculated ratios to give the resulting allocation to each different name:

The investor can then judge whether this final risk exposure is tolerable. Names held through equity and mezzanine tranche will be weighted more than senior tranches, particularly if the tranche is quite narrow.

Waterfall
For this approach we have not adopted the waterfall mechanism directly in order to calculate the risk weights. Although the waterfall is more accurate at reflecting the impact of any losses or collateral repayments on the capital structure, our aim was to introduce a method which is quick and straightforward to calculate.

Through the waterfall, tranches are amortised periodically following repayments and losses, and in the case of CDOs with delayed draw features the overall note size can also increase when any outstanding commitment is called. By using current credit enhancement and tranche width as the weighting factors, the impact of any such amortisation or expansion is indirectly taken into account in the above algorithm.

Hypothetical investor example
Suppose a hypothetical investor's portfolio is comprised of four different CLO tranches (two equity, one mezzanine and one senior) and one credit fund, which between them give exposure to 20 different credit securities. The application of the proposed algorithm to calculate the risk exposure is shown below. It should be noted that in reality exposure would be to a much larger universe of loans, but we have used 20 in order to show the algorithm steps more clearly.

Findings
The numbers in the 'Rebased Weight (W_i^R)' column in the above table show the estimated risk exposure the investor has to each of the different securities.

As seen, compared with the overall CLO's allocation the equity tranche allocation is significantly higher in order to account for the leveraged impact of the first losses. CLO1 had a thinner equity tranche width compared to CLO2, and hence the corresponding equity tranche weights are slightly greater. The senior tranche in contrast benefits from a large degree of subordination and is much wider and hence the corresponding weight allocation is less.

Difference to standard approach
The final column in the table shows the allocation to each security if the standard approach is adopted. As seen, there are some notable differences, e.g. Security 1 is held in both the equity tranches and therefore to account for the greater risk, on default, the algorithm approach results in a significantly higher allocation.

In contrast, Security 13 is not held in either of the equity tranche allocation nor the mezzanine tranche and therefore the algorithm results in a risk exposure of only 0.5% compared with 2.7% using the standard approach.

It should be noted that the algorithm does not show the percent of the portfolio that will be lost if a particular security defaults. Losses will be experienced in the equity tranches and the credit fund but for the non-equity tranches the risk is as a result of a drop in credit enhancement which indicates a greater probability of being affected by future losses, in particular the mezzanine tranche.

Algorithm weight allocation across different types of capital structures
The following table details the extra weighting factor application to tranches from different capital structures, highlighting the importance of using both credit enhancement and tranche width in the algorithm.

We have created three different CLOs with varying capital structures to emphasise the flexibility of the algorithm. The three CLOs can be considered to have an increasing fixed allocation to senior secured loans and the remaining in second lien or mezzanine loans, starting from CLO A with 75% senior secured, CLO B with 90% senior secured and CLO C with 100%.

Observations
? Of the three equity tranches CLO C is the thinnest, with a tranche width of 7.0% and therefore results in the highest weighting factor of 14.29. As the equity tranche size for CLO A and CLO B are identical and, being an equity tranche, there is no credit enhancement, the two tranches are equivalently risky and the algorithm therefore results in the same weighting factor of 8.33.
? Focusing on the junior tranches, CLO A and CLO B benefit from the same credit enhancement of 12%. However, CLO A's sub tranche is much thinner and therefore more risky which is reflected in the higher weighting factor of 14.56 compared with CLO B's sub tranche factor of 4.08. Although both CLO B and CLO C sub tranches are of equal size (7.0%) the lower credit enhancement offered by CLO C results in a higher weight.
? The senior tranche for CLO B and CLO C are equivalent from a risk perspective, based on the algorithm which provides a weighting factor of nearly 0.9 for both. Although CLO C has a much lower credit enhancement than CLO B, 30% compared with 35%, this is compensated for by the wider tranche, reducing the leveraged impact of any losses.
? The level of credit enhancement and tranche width therefore drives the value of the risk weighting factor. As we move up the capital structure the risk, and hence the weighting factor, for all three CLOs decreases.
? The flexibility of this analysis therefore provides a means for comparing the risk of two different CLO tranches with equivalent rating but with different underlying collateral allocation, for example different senior secured loan allocation as per the above analysis.

Wider portfolio
The analysis undertaken here focuses on any overlapping arising across different leveraged loan exposures. In addition to leveraged loans, investors may also hold other types of exposure to a particular corporate name, e.g. corporate bonds and equity shares.

To account for this level of overall risk exposure the analysis can be further extended to these asset classes. The seniority of the different instruments, in the event of a bankruptcy, should be taken into account by basing the allocation on suitable weights such as recovery rates, with loans ranking higher to issued bonds and equity holders receiving only the residual asset value.

Ways of managing and reducing impact of overlapping

By investing in multiple CDOs investors can benefit from increased diversification. However, as the number of CDOs held increases some degree of overlapping cannot be avoided. To manage and reduce the impact of any overlapping, the following techniques can be adopted.

Measuring diversity using a Herfindahl type measure
One way to quantify the overall diversity of an investor's portfolio is to then calculate a diversity measure based on the Herfindahl index, similar to Moody's diversity score. This index effectively converts a pool of unequal loans into an equivalent number of equal size loans, measuring overall diversity.

A diversity index (DI) can be calculated as the reciprocal of the sum of the squared rebased weights (shown in in the Rebased Weight (W_i^R)' column in the hypothetical portfolio example):

For the hypothetical portfolio the diversity index is calculated to be 16.2, indicating that although the overall portfolio is exposed to 20 different entities, in varying sizes, this is effectively equivalent to an equal sized allocation to only 16 names, due to the effect of overlapping.

As new investments get added to the portfolio, we would expect the diversity index to increase, although at a decreasing rate, due to increasing probability of overlapping names.

Manager skill
The CLO manager skill plays an important role. A skilled manager is better able to identify and replace weaker securities before deterioration in quality or defaults impact the CLO portfolio.

However, as indicated by the S&P survey, CLOs issued by the same manager, particularly ones issued close together, have a degree of overlap and hence investment in multiple CLO tranches managed by the same manager should be kept to a minimum.

Diversification across different managers also gives a broader exposure to different investment and management styles thereby reducing the impact of any overlap.

Diversification across vintages
With time the portfolio structure of CLOs evolves and therefore allocation across different vintages will further benefit diversification. Not only are the managers sourcing from different leveraged loan pools but the different structural features also add value. Recent years have seen synthetic buckets, short buckets, mezzanine and second lien loans all being used to enhance spreads and hence returns achieved by the portfolio.

Capital structure and regional spread
The investor's portfolio should also be well spread out between different parts of the capital structure. As seen senior tranches provide a larger degree of subordination and therefore are less affected by any overlapping exposure. Investment across different regions also benefits from a broader range of collateral.

Regular monitoring
As overlapping is difficult to avoid completely, we would further recommend that investors regularly monitor their exposure to the various securities to ensure they are comfortable with the risk borne and the impact of any losses in the event of a default is sustainable.

© 2007 Dresdner Kleinwort. All Rights Reserved. This Research Note was first published by Dresdner Kleinwort on 13 February 2007.