Contents

CDR Liquid Index data as at 24 October 2006

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.

CDS on CMBS for Japan

New product in the pipeline

Tokyo-based dealers are prepping the first CDS on Japanese CMBS trades. Local investors are sceptical about the new products, but the target market is expected to be international firms.

"We and a number of other firms are working on synthetic CMBS products, but as yet no trades have been done. There will be within the next month though," confirms one dealer.

One Japanese investor is not convinced that there is a need for such products, however. "CDS on CMBS is not what we are thinking about at the moment," he says. "We have talked with other investors about something more like a total rate of return swap on a CMBS underlying - and that type of trade seems much more likely to be a success, I think."

He argues that CMBS is not yet a large enough sector in Japan to merit its own derivatives market. "I don't think there is an immediate need for investors to buy protection on CMBS using CDS and so it would also be difficult to write protection and make markets in such instruments as is done in the US. Here, it would have to be done on a case-by-case basis."

However, proponents of the new product see huge potential for demand, arguing that they will not only appeal to cash market investors, but also to a whole swathe of foreign investors who - because of regulatory restrictions - cannot access onshore cash products.

Indeed, it is expected that a wide array of non-Japanese investors will want to access the market through CDS on Japanese CMBS. "It's simple: Japanese spreads are wider than those for CMBS in Europe and the US, so there's an obvious relative value play there for the hedge funds. At the same time, real money will want in because the wider spreads offer great yield pick-up - they're not wider because of fundamentals but because of technicals," says the dealer.

"Ironically, it's the lack of foreign flows that have kept the spreads wide. So, once the market takes off, they'll come back in," he adds.

Meanwhile, it is understood that another type of structured credit product is being readied exclusively for the Japanese domestic market. A number of dealers are putting together managed CDOs of synthetic ABS and plan to offer them in yen-denominated form to local investors by year-end.

MP

European spreads grind tighter

No end in sight to current range

Hopes that some overall spread widening might take place over the last week were dashed as data and specific name concerns failed to move the market. Instead, dealers have busied themselves with marketing the new LevX index.

"Tranches have seen pretty thin trading and that really follows on from the indices being quite slow in what is usually a fairly dead week because of half-term in the UK, but with all the value squeezed out there is less activity anyway," says one trader.

"We're trading in a very narrow range and there doesn't seem to be a lot of volume going through at these levels," concurs another. "Investors think we are probably trading at the right levels here and everyone is trying to figure out where the next move is. Consensus is that we are at the tight end of the range, but it's difficult to see what is going to drive us wider."

Spreads in the 5y iTraxx Europe main index stood at 27bp at Tuesday's close, just shy of 1bp tighter than the previous week. Over the same period, the Crossover index narrowed just under 7bp to 257.8bp.

There were hopes that end of quarter earnings would produce a couple of disappointments that would lead the market wider, but that hasn't happened. Equally, there appears to be a widespread belief among structured credit investors that equity markets are too high. But without sharp moves there, credit spreads are likely to stay where they are.

It is a similar story for single name CDS, with very little support for any significant moves. "There are names that are the headlines but there is very little follow through," says a dealer.

A currently diminished CDO pipeline had also offered some false hope of easing of tightening pressure, but issuers instead looked to hedge recent bespoke CDOs. In any event, some believe the effect of the 'structured bid' is now overstated.

"These days there is always some kind of pipeline, but there is also protection buying coming from CDOs - it's much more of a two-way flow and much less dramatic than it used to be. The structured bid is there, but it's not like we suddenly gap in by 5bp because a big CDO prints anymore," says the dealer.

However, dealers have been fully occupied this week with marketing the launch of the new European leveraged loan index - LevX. The senior index is scheduled for launch next Monday, 30 October, and the subordinated index two weeks later (for more detail see SCI issue 10). "It's quite a big event because many people have been waiting for it to launch for some time," the dealer observes.

MP

Correlating opportunities

This week, Eric De Sangues, chief investment officer at Avendis Capital, answers SCI's questions

Eric De Sangues

Q: When, how and why did you/your firm become involved in the structured credit markets?
A: Four of us, all arbitrageurs, set up the company in February 2001. Two had a fixed income background with experience in Cash CDOs and other forms of securitisation and two came from the equity derivatives world with extensive skills in derivatives trading and pricing.

The idea from the outset was to provide equity and fixed income strategies to investors to help them post interesting returns irrespective of market conditions. Initially, we specialised in two types of strategy: structured credit and equity arbitrage, but today the core business of the company and where we have developed a core competency is structured credit.

We began by trading synthetic CDOs on behalf of clients first through managed accounts and then we built a family of hedge funds to rationalise the costs involved and to enable more investors to benefit from our strategies. Those strategies involved looking at the market in a different way to others.

We focussed on pricing models and tried to create some market neutral pricing models in the same manner as we had with convertible bonds in our previous jobs. Cash CDOs were not really traded this way at that time – they were typically traded on rating assumptions and analogy between products, which was, in our view, a far less rational way of looking at things.

We basically put our way of looking at options and derivatives products in the equity world into the fixed income world. Models were very far from being standard and there was a lot of discrepancy in prices between dealers so we realised there were a great deal of opportunities.

A further opportunity materialised with the 2002 credit crisis, which generated a lot of sellers of distressed synthetic CDOs. We took this opportunity to build a portfolio of distressed CDOs for private accounts and to launch the Avendis Enhanced Fixed Income Fund, one of the first credit correlation hedge funds.

Subsequently, we have also diversified into structured product management. We are now a well-established manager of single tranche synthetic CDOs and last week launched with Calyon a CPPI on index tranches called Cortex (see SCI issue 11), which is one of the first transactions of its kind.

Q: In your view, what has been the most significant development in the credit markets in recent years?
A: The standardisation of products, which has brought a lot of volume, has been key. Prior to the improvements in this respect, the lack of standardisation was causing a lot of problems in terms of secondary trading and in terms of the time it was necessary to spend studying each individual product.

Another key development is the fact that the market has substantially increased on a swap basis and not on a cash basis. We now almost exclusively trade unfunded products, whereas when we began we were only trading funded products, which raised a number of concerns surrounding issues such as SPV trustees and collateral.

Q: How has this affected your business?
A: We are now in a world that enables us to more efficiently price the whole range and increasing variety of structured credit products. Part of the arbitrage opportunity of the past has now disappeared as a result, but at the same time volumes have increased so much it has enabled us to take different and, and I think more interesting, opportunities. For instance, increased liquidity has decreased bid-offer spreads making relative value plays possible.

Q: What are your key areas of focus today?
A: We think index tranches are the part of the market now where we can find the most appealing opportunities. Once again, the idea is that liquidity has increased so much and bid-offer spreads have decreased so much that it's now very efficient to use index tranches to express correlation views.

In the past it was possible to set up idiosyncratic plays by trading one CDO tranche against another or a single-name CDS, but it was not efficient to express correlation views this way because of the bespoke nature of the products involved.

Now, if you focus on index tranches they enable you to very efficiently play the dynamics of this market. Those dynamics are very interesting today because the market is still in a state where the flows are very imbalanced between hot money and real money.

Q: What is your strategy going forward?
A: Obviously developing assets under management in our hedge fund business is a key focus. But this is equally true for our structured products business.

With the launch of Cortex and similar structured products we will be able to attract more real money investors, which has the double benefit of diversifying our investor base, while also attracting new investors into the structured credit market more generally. We see repackaging correlation risk as a big trend developing in the market at the moment and we intend to participate fully in that trend.

Q: What major developments do you need/expect from the market in the future?
A: We expect to see more liquidity in synthetic CDOs and a more diversified investor base in the market more broadly as the aforementioned products are developed and their packaging makes them more suitable for a wider range of investors. For instance, we think that repackaged equity tranche risk products will develop – we are very bullish on rated equity structures if the current low default and low spread environment continues.

Another thing we think will continue in the future is the improvement in pricing models. We are all currently using different types of pricing models that are all essentially descriptive models as opposed to predictive. So, they do not take into consideration a lot of important aspects including the volatility of the risk factors that are input into models.

Undoubtedly improving models is still a work in a progress, but everyone in the industry is pushing hard for it to happen. We are doing our own work as well and think the natural extension for models today is to at least account for volatility in credit spreads, which will become very important in the future after such a long period without volatility in the market.

About Avendis
Based in Geneva, Avendis Capital was launched in February 2001 by four arbitrageurs as a privately held Swiss investment company. In less than five years, Avendis Capital has evolved into a strong alternative investment operation, with an expert team and institutional infrastructure.

Avendis-Group
Avendis-Group was formed at the end of 2005 to consolidate the entities created by Avendis founders. As of December 31, 2005 the Avendis Group of companies manage more than €2.2bn in combined assets.

MP

"We evolve with the market"

In this week's Provider Profile we talk to the securitisation technology group at Deloitte & Touche in New York

The securitisation technology group at Deloitte & Touche (D&T), although managed as a dedicated unit to enable it to react swiftly to client needs and market developments, is also an integrated part of D&T's broader securitisation practice. Indeed, it has developed on top of the foundation of Deloitte's broader experience in the securitisation business.

"Our U.S. organisation has well over 200 professionals dedicated to securitisation, and the other member firms of our global network double that number. However, our securitisation technology group's specific mandate is to focus on the development and support of software-based solutions for our external securitisation clients," says Mark Scherer, a principal with D&T in New York, and head of its securitisation technology group.

Hillel Caplan and Mark Scherer

"As we develop and enhance our securitisation software products, we're able to supplement our securitisation technology group's more than thirty dedicated business analysts and developers by drawing upon the deep resources and industry knowledge available to us here at Deloitte," adds Hillel Caplan, a senior manager with D&T in New York, and global product leader for its CDO Suite software system. Caplan himself illustrates this point as he transferred from D&T's dedicated CDO payment date verification group to the securitisation technology group back in 2000 to jumpstart and guide the original development of CDO Suite for a large U.S. bank that was entering the CDO trustee market. Caplan has led the CDO Suite product team ever since.

D&T's securitisation technology group has developed strongholds in a number of areas within the securitisation market. "We currently serve the majority of the global CDO trustee market, representing over 50% of year-to-date 2006 CDO issuance, many CLO and SFCDO managers, seven of the top ten issuers of credit cards, and many of the captive and independent auto finance companies," notes Caplan.

D&T serves the market via a family of three long-established securitisation software products – ABCP Suite (for sponsors of asset backed commercial paper conduits), ABS Suite (for issuers of, and trustees for, ABS and MBS vehicles), and CDO Suite (for asset managers, collateral administrators and trustee banks in the collateralised debt obligation market), and is in the process of developing a fourth system called Surveillance Suite.

"Surveillance Suite represents our first securitisation software product that is targeted specifically at the needs of investors in structured products. Our existing family of products is really focused on the parties that issue or administer securitisation programs," explains Scherer.

D&T's securitisation technology group is nearly twenty years old. "We have been doing this since we supplied software for one of the first credit card securitisations in 1988, for a large US bank," says Scherer. However, despite such a long history, the group hasn't grown top-heavy. "Significant business decisions are made by just three people, myself and our global software product leaders – Hillel Caplan on the CDO Suite side, and John Will on the ABS Suite side. We essentially have the nimbleness of a small company combined with the very deep pockets and vast, global resources of a large company" Scherer observes.

The large number of clients garnered by this streamlined business model has been achieved through D&T's collaborative and bespoke approach. "While we are very focused on maintaining a standard, common, core software platform, it is important to make sure that the software fits the individual needs of each client" says Scherer.

"As we've grown, we have also searched for creative ways to maintain a strong connection between our product team and our clients. For example, on the CDO Suite product we use the same 'wiki' software as the online encyclopaedia site Wikipedia does, to allow our client base to collaboratively develop and edit requirements for use in future development. Given the diversity of our CDO Suite client base in terms of asset classes and geography, this provides a broad, market-based answer to what our development approach should be on a range of topics. We provide secure, open access to the CDO Suite Wiki Site for all of our clients to express how they would like the software to develop. They do this based upon both their own experiences with the software, and their interpretation of market developments that will affect their future user requirements," explains Caplan.

D&T also publishes a listing of its development plans on the CDO Suite Wiki Site for the next several months to give clients a view of what they can expect. The overall wiki-based system is a better approach than calling each client to get their views separately as it enables true transparency on user requirements, according to Caplan. "On the CDO side, we doubled our client base in the last year and developed the CDO Suite Wiki Site along with our CDO Suite User Group as additional feedback mechanisms. We let our clients drive the process which ensures that we evolve with the market," he adds.

The results of this collaborative, yet standardised approach, says D&T, are that everyone benefits and becomes stronger. Otherwise, clients risk becoming an 'island unto themselves' as Scherer puts it. This fate befell a competitor of D&T's that made itself exclusive to a single client and eventually lost that client to D&T.

There is nevertheless a case to argue that in some situations it is right for some clients' development suggestions to be kept unique to them, otherwise they in turn might lose their competitive edge. Caplan acknowledges that this can happen, and says that D&T provides for it where possible: "There are areas of the system that are specifically designed to allow for customisations that remain proprietary to the client that requested it."

Providing open access for clients to its software source code is also a long-standing D&T principle, enabling users to feel fully involved in the software's development. This approach dates back to D&T's early days of developing securitisation software. "We provide great technology and service, but we also strive to make our clients as self sufficient as they choose to be to run and develop the software on their own. Source code access is an important part of that, and something that we believe we are unique in providing in the securitisation software space" added Scherer.

With the rapid pace of growth in the market, as well as increasing complexity in the deals that are done, Scherer remarks how it is 'disconcerting' to still see market participants working in tools like Access or Excel. He warns: "Organisations that rely on models that are hard to maintain, lack controls and may not be easily taken over by others if someone leaves may be incurring much more operational risk than they realise."

Equally, Caplan argues that newer market participants such as entrants into the CDO asset manager realm should demonstrate investment in a system. "Investors and the rating agencies want to see some substance; a true system demonstrates that you have made a real investment in your business. In addition, as time progresses, you will reap the rewards of that investment as you can grow rapidly without giving a lot of thought later to your systems, at a time when the cost in both time and money of conversion from tools like Excel into full-fledged systems grows."

Going forward, it is likely that the message of the need for scalability will be increasingly heard and adhered to and D&T's software group is well-placed to take advantage of that greater understanding. As Scherer concludes: "While we expect continued strong growth in the US, we also expect Europe to significantly contribute to our growth in the next few years. Fundamentally though, our success can ultimately be boiled down to the terrific continuity that we enjoy within our team, and we will make the investments necessary to ensure that our clients enjoy similar continuity for years to come."

JW

Athilon re-jigs as CDPCs' plans unravel

The latest company and people moves

Athilon re-jigs as CDPCs' plans unravel
Following the exit of ceo Paul Sclafani, CDPC Athilon Asset Acceptance has promoted coo Patrick Gonzalez to ceo and Joseph Bauman has been appointed chief administrative officer. Gonzalez was a member of the management team that founded the firm in 2004. Prior to joining Athilon, Bauman worked on a consultancy basis to WestLB's CDPC Mandalay and previously was co-founder and former chief financial officer of Primus Guaranty.

No reason was given for Sclafani's departure, but it is understood that Athilon has recently shelved plans to sell itself in the absence of the right offer in terms of both size and conditions. At the same time, Matthew Cooleen, ceo of Deutsche Bank's CDPC, has also left his job.

Sources suggest that his departure was a result of a rift in strategy and has caused Deutsche's plans to be put temporarily on hold. The bank is expected to move forward with its CDPC next year, however.

Meanwhile, CDPC initiatives elsewhere appear to be struggling in some form or other. Mandalay has been shelved completely, while other bank-sponsored companies are believed to be having difficulties with the rating agencies because they are proposing more complex structures than existing CDPCs. Additionally, some non-bank sponsored initiatives are understood to be having problems raising the necessary capital.

RBS adds another strategist
Gregory Venizelos, formerly a credit derivatives strategist at ABN AMRO, will join Royal Bank of Scotland's structured credit strategy team on 8 November. Venizelos is the latest in an expected number of hires to be made by RBS (see SCI issue 10) and will report to Siobhan Pettit, who heads the team.

Pentagram cio joins Babson
Babson Capital Management has hired Mark Turner, president and cio of Pentagram Investments Partners, as an md. He will be responsible for expanding the firm's global macro absolute return focus.

Turner will manage a group of five professionals and the team will be based in Babson Capital's Boston office, with responsibility for three global macro portfolios comprising initial assets that total US$100m. He will also retain ownership of Pentagram, the company he founded in 2000, and will continue to have management responsibilities for the current Pentagram portfolios that will be sub-advised by Babson Capital.

Prior to founding Pentagram, Turner worked for three years as a senior partner at Schooner Asset Management, an alternative investment boutique. Before that, he served as an md at Caisse des Depots, as an md and cio of global fixed income at Putnam Investments, and as md of global fixed income at Scudder Stevens & Clark.

Deutsche promotes Morley
Deutsche Bank has promoted Daniel Morley to global head of emerging market debt. Morley, previously head of CEMA structuring, will report to Kay Haigh, head of emerging market debt trading.

Quintas joins Lehman
Juan Quintas has joined Lehman Brothers as head of ABS correlation trading. He was previously at Citigroup, where he headed up trading and risk management for the global ABS correlation desk. Quintas now reports to Georges Assi, global co-head of CDO and structured credit, and Gordon Sweely, global head of principal finance and ABS trading.

Phoenix rises in London
New York-based interdealer broker Phoenix Partners Group has opened a London office. The new operation will initially focus on voice-brokered credit default swaps. The business is led by Alex Hucklesby and Graham Smith, who both formerly worked for Mint Equities.

Phoenix was founded in New York last year, and currently has about 200 clients.

Derivative Fitch launches
Fitch Ratings has launched Derivative Fitch, the first specialist rating agency designed to focus on the credit derivatives market. In forming the new agency, Fitch says it will consolidate over 100 professionals from its global CDO and structured credit ratings groups and related products, analytics and modeling groups.

In addition to ratings, some of the products and services that will operate under the Derivative Fitch brand include: the Default VECTOR model, Risk Analytics Platform for Credit Derivatives (RAP CD), Valuspread, Stability Scores and FitchCDx.

MP

CMBX rolls and adds double-B tranche

A round up of this week's structured credit news

CMBX rolls and adds double-B tranche
The CMBX Indices will roll today, October 25. In addition to the new series of the existing triple-A to triple-B minus indices, a new double-B class will launch and will function identically to the alternate indices.

The creation of the double-B index will create new trading potential for investors and allow the opportunity for exposure to an additional credit class, according to CMBX data provider Markit Group. The dealer group for the indices will also continue to provide daily marks on the most recently off-the-run version of CMBX.

CDO diversification a boon for bond investors
Wachovia Securities' structured products research team has published a report that analyses CDO performance versus corporate bonds. Uniquely, as part of the analysis the team has developed a new metric which factors in risk and return – an aspect that has never been included in prior industry metrics.

CDO investors have been well rewarded over the past three years: spreads have compressed in most sectors; ratings transitions have been stable; and liquidity has increased, allowing investors to take profits. Although there is a dearth of publicly available data on the secondary CDO market, Wachovia's analysis indicates that – based on new issue spreads – the higher risk premiums awarded to investors in the sector have resulted in superior performance relative to swapped corporate bonds.

But buyers of CDOs have not benefited equally from this superior performance. Dragged down by housing fears, the mezzanine notes of ABS CDOs (both high-grade and mezzanine structures) have underperformed notes in other sectors.

On the other hand, the recent benign credit environment has rewarded mezzanine CLO investors the most – although the Wachovia analysts believe that triple-A notes offer a better risk/reward tradeoff going forward. Indeed, the team is especially cautious on double-B rated notes, where spreads have compressed by around 460bp over the past three years. In a more uncertain credit environment, lower-rated note investors are especially vulnerable to low collateral recoveries.

The bank's findings support the idea that investing across different CDO sectors is an excellent means of diversifying a portfolio, with corporate bond and equity investors also likely to benefit from such diversification opportunities.

EFG brings debut Greek SME CLO
Deutsche Bank has launched the first Greek SME CLO – Anaptyxi 2006-1. As part of its wider objective of supporting SME lending in the country, the European Investment Bank subscribed to the Class A notes, with the originator of the deal – EFG Eurobank – agreeing to use the equivalent amount to finance new SME projects that meet the EIB's eligibility criteria.

Rated by Standard & Poor's and Moody's, the €1.75m AAA/Aaa rated Class A notes priced at 17bp over three-month Euribor. The €150m A/A1 Class B notes came at 40bp over, while the €125m BBB/Baa1 Class C notes and the €225m BB/Ba2 Class D notes printed at 75bp and 250bp respectively. The funds raised by the securitisation will be used to cover part of the EFG's funding programme.

Joining Deutsche Bank as joint-lead managers on the transaction were EFG Telesis Finance and Merrill Lynch International, with DZ Bank acting as co-lead.

Avoca continues tightening trend
Avoca Capital's latest transaction – Avoca CLO VI through Credit Suisse – has priced at record tight spreads at the single-A, triple-B and double-B levels. The single-A notes priced at 55bp over Euribor, 1bp tighter than Alcentra's Jubilee CDO VII which Barclays Capital brought to the market earlier this month (see SCI issue 10) and 3bp tighter than the equivalent tranche of the Carlyle Group's CELF Loan Partners III – through Deutsche Bank – from September.

The triple-B tranche came 10bp and 5bp tighter than CELF and Jubilee respectively, setting a European spread record at 135bp over. However, the double-B notes matched the tight levels set by those transactions, printing at 335bp over.

Unusually, Avoca CLO VI included a single-B rated note, reducing the equity slice to around 7.4% of the total capital structure compared with 10% for a typical CLO. The portfolio comprises senior and mezzanine European loans, and high-yield bonds – with a minimum of 85% of the pool expected to be senior secured loans.

CDO equity sophistication continues apace
CDO arrangers are eyeing more creative ways to package and sell the equity pieces of their transactions, according to analysts at Fitch Ratings. The agency believes that different flavours of zero-coupon equity and rated equity on investment grade corporate synthetic transactions will continue to emerge.

In the zero-coupon equity structure, the investor makes an upfront payment (for example, 30) and if there are either minimal or no losses over the term, the coupon is rolled-up and the investor receives 100 at maturity. Such trades are appealing to hedge funds because of the low up-front funding requirement.

Rated equity trades, meanwhile, are an offshoot of CDO combination notes – typically a mix of senior notes and equity that are designed to fit a specific investor's needs. Rated equity trades differ from combo notes in that they only reference the equity portion of a transaction.

In these trades, the investor puts in 100 upfront to purchase an equity tranche (for example, 4-5% thick) that is supported by excess spread. The rated coupon is nominal (70bp over Libor) and the counterparty also makes a reserve contribution (7-10%) which is used to offset losses and pay the coupon over the term or at maturity.

Away from the CDO equity space, Fitch analysts also note that the hunt for extra yield has led to the creation of more varied portfolio blends in CDOs. For example, while residential mortgages still comprise the majority of the assets in structured finance CDOs, market participants have been discussing adding corporate buckets in these transactions. Similarly, traditional CLOs with some exposure to structured finance assets are being considered, with single-property and middle market loans appearing in the form of CDOs of trust preferred securities (TruPS).

However, the agency warns that these mixed pools carry additional risks. In the case of corporate buckets in SF CDOs, corporate assets follow a different default pattern from structured finance assets and tend to be more vulnerable to idiosyncratic risks such as management and business line changes (leveraged buy-out risk), as well as changing demographics.

MP

Trading ideas - Mo' Money Clues?

Tim Backshall chief credit derivatives strategist at Credit Derivatives Research suggests a curve steepener trade involving Marsh & McLennan Companies Inc

It's been a turbulent few months for Marsh & McLennan Companies Inc (MMC) with spreads seemingly anticipating much of the actual news. After insisting for months it would not consider a sale or spin-off of Putnam, its investment management unit, MMC abruptly reversed course in late September. The company said that in response to "repeated inquiries," it has decided to do a "market check" on the value of Putnam. MMC said it had not settled upon "any specific action," but shareholders will likely be disappointed if the company does not take action after publicly announcing its intention to consider its options.

Even if Putnam is sold or spun off, it is not clear that shareholders will be satisfied especially now it seems likely that outside bidders may be interested in the whole company. An outright sale of Putnam would trigger a significant tax bill and so the possibility of a spin-off becomes more likely (and only adds to the uncertainty). A spin-off is less favourable for credit investors as there would be no cash infusion and further any debt reduction (applied to the Putnam unit) would be minimal compared to MMC's total debt.

While Moody's believes that MMC will 'protect its financial flexibility', we feel given operating cash flow, pressured by restructuring and settlement charges, has been negative in recent quarters impacting liquidity, concerns over the Putnam unit's future, and the growing impatience among MMC's shareholders (still waiting to get back to pre-2004/2005 insurance scandal levels), that MMC is a weak credit with significant event risk. Rather than play the outright, and face the cost of protection and roll-down, we prefer the bear steepener which while offering less upside, controls downside and provides positive carry and roll-down giving us time to be right on this troublesome insurer.

Coming Apart

Over the past couple of years, debt and equity markets have generally agreed on the fortunes of MMC. Exhibit 1 shows that MMC's credit spreads have traded consistently in line with its equity price until recently. It is evident that equity investors have certainly been more vociferous in their views and have shown significantly more volatility than the rather more positive credit investors over the past year.

Exhibit 1

Spreads were considerably wider back in 2004/2005 on the back of the market-timing and contingent commission crises but as investors memories shortened, spreads rallied. The equity price (green line), which is inverted, did follow the spreads at this time but with less of a jump, rallying to over $33 per share from less than $25. For the next year, equity prices sold off along with spreads with equity prices significantly more volatile than spreads.

During 2006, we saw MMC spreads ignoring the overall market flow. While CDS widened modestly over the first eight months, equity prices sold off dramatically back to late 2004 lows. This drop puts more pressure on management to be friendly towards the shareholders and we have seen a rally over the last couple of months driven by overall equity market flows. At the same time, spreads moved wider – this clear divergence (indicated by the arrows in Exhibit 1), anticipates significant changes to the capital structure as the two markets come unhinged.

The spread differential (mid-to-mid) between fives and tens has behaved consistently as seen in Exhibit 1, as spreads have tightened and widened. The price action this last few days has spiced up the curve's volatility and offers us an interesting entry opportunity at current levels, especially considering the change in relative (rather than absolute) differential.

Differentiated Differentials

MMC has excellent liquidity in the CDS market across the curve and now trades actively in five-, ten-, three- and seven-year maturities. MMC is a member of the new Dow Jones CDX NA IG Series 7 index and bid-offer differentials have been dropping and liquidity is excellent. As spreads have moved over the past year so the curve differential (between five- and ten-year CDS) has moved largely consistent with it, as seen in Exhibit 2.

Exhibit 2

Mid-to-mid absolute differentials had remained modestly high (compared to its peer group) and consistent between 25 and 30bps until early summer when spreads started to widen further and the curve (as we would expect), moved steeper towards the mid-30 basis points levels.

Percentage differentials have behaved just as consistently over the past year – rising as spreads drop and dropping as spreads rise. The percentage differential is also much more reactive to the changes in the curve and this is one reason we use the percentage differential when judging crossover or wider trading credits as the curve fair value is more consistent across wider spreads.

As with the absolute model, the percentage differential ground higher as spreads slowly tightened through 2005. As spreads moved wider this year, and volatility picked up so we saw the percentage differential dropping and become more volatile. The recent rumours and results from last month drove the percentage differential lower as spreads approached one-year wides and the differential approached one-year lows.

These changes in relative-value triggered one of our curve models and we feel indicates not only that the curve is too flat (given the spread levels) but as investors in CDX7 begin to analyse the components more fully, they will see MMC out of line with its corporate peers (even when we account for the insurance premia). This mid-to-mid analysis is significantly enhanced when we investigate the actual bid-to-offer prices of curve trades.

The Real Deal
We prefer positive carry (or low-cost) trades (and relative-value opportunities) that are somewhat market neutral and at these levels we see a little too much downside for a naked short although we could potentially hedge with some exposure to the index. We still prefer the steepener as our positively economic shorting trade (especially when our market bias is to the widening and steepening side) when we are concerned on downside and so investigating the 5s-10s relationship provides us with both relative-value and positive carry.

One more item of note with the steepener is that it buys us time – if we are modestly negative on the credit but do not expect a major move soon, it does not cost us to maintain a position that benefits from under-performance. The position provides us with a cushion on our timing of any negative view and offers good upside from just rolling down the curve (if nothing happens).

Moving from the mid-to-mid (rough) curves to the bid-offer adjusted curves in Exhibit 3 paints a much clearer picture of where MMC's curve is trading. The differential between the standalone levels and the peer group's fair value, indicated by the dotted lines in Exhibit 3, point to an interesting divergence in 5s-10s. The 5s-10s flatteners (red circles) and steepeners (green triangles) have traded consistently flat to their peers over the past year with levels rising over the last few months.

Exhibit 3

The curve basis (steepener to flattener differential) has also been significantly tighter than the market reflecting the strong liquidity in MMC's curve. This improves our overall economics (given easier exit levels).

The steepeners (green triangles) have traded cheap (flat) to fair value for much of the last year but current levels show that 5s-10s are actually the most divergent they have been – at over 10bps. The combination of the tighter basis (meaning it is cheaper to cross the bid-offer of the curve trade), steepener cheap to fair value, historically wide divergence from fair value, and recent modest pull-back, encourage us to continue looking for relative value here.

The jump in spreads after the recent results, and accompanying modest steepening of the curve provides us with a positively economic trade in this potentially troublesome credit. As the curve has pulled-back a little in the last couple of days, below the top of the last year's trading range, we feel comfortable that on a standalone basis this trade is in our favour and further will benefit from any further deterioration in spreads.

Constructing the Curve
Our models are constructed on mid-to-mid levels to provide us with as much consistency as possible across the curve – especially when trying to compare investment grade and high-yield names together. In reality we must ensure that the relationship still holds when bid-offers are accounted for.

Our data provider, aggregating a number of dealers' runs on a real-time basis throughout the day and offering us a critical eye on the actual intra-day trading, provides us with exceptionally tight markets in many of the market's most liquid names – as we received best bids and best offers. We use this 'latest' bid and offer to drive Exhibit 4.

Exhibit 4

This model provides a much more real-world approach to judging how effective our trade could be. The green line divides the credits into positive and negative carry trades – steepeners positive below the line and flatteners positive above. This line is approximate and does not account for the slight convexity differential as spreads widen, but is certainly useful in deciding where to trade.

Comparing MMC to its peer group of Insurance credits, we note (in Exhibit 5) that it is trading at the high end of its range in five-year spreads but is actually the cheapest to fair value (green arrow). Also, we have added the fair value curves for the Insurance sector (dotted curves), that show both the tighter bid-offer of the Insurers (driven by investor appetite) and the fact that the curve is a lot flatter overall compared to the market (solid curves). Given its potential weakness and uncertainty over its future, we are extremely surprised that investors are willing to treat it so well – letting its curve trade so flat. Even though it appears in line with the Insurance fitted curve – it is pulling the long end down as the only credit trading that wide.

Exhibit 5

If spreads remain where they are and we get no more surprises from MMC, we will still roll comfortably down the curve, earning our carry and picking up time value. On a standalone basis, we are trading close to one-year wides, have a deteriorating fundamental outlook idiosyncratically but are prone to event risks (which we view as increasingly likely), and are trading modestly off the high-end of its curve history.

If we combine this standalone perspective with the peer group comparisons, which have MMC at fair value 10Y levels around 20 basis points higher given the current 5Y levels it is evident that something has to give with MMC – it cannot sustain a curve this flat at these levels. Even more surprising when one considers the somewhat unambiguous consequences of the idiosyncratically deteriorating fundamentals and outcomes of any capital structure changes as being negative for credit investors combined with any potential impact of a reduction in the technical bid (on potential downgrades). Maybe it is our paranoia but we believe that the slope will steepen as the market digests these risks (and takes a closer look at MMC and continues to discriminate weak credits more effectively among the insurance sector) pricing MMC's intermediate-term debt accordingly, especially if the more endogenous technical bid's impact on the long-end is reduced as we expect.

We therefore recommend a duration-weighted curve steepener in MMC's fives-tens segment of the curve. Duration-weighting here is around 1.6-to-1. This trade benefits from significant positive carry and roll-down, and offers a low-cost time-insensitive approach to place a negative bet on MMC.

Risk Analysis
This trade is duration-weighted to ensure positive carry as well as to reduce our exposure to absolute levels. We are therefore hedged against short-term movements in absolute spread levels, profiting only from a curve steepening between the fives and tens. We could adjust the weighting to take advantage of the apparent richness in the ten-year but we feel more comfortable with the risk-reward of the duration-neutral steepener.

The carry cushion protects the investor from any short-term mark-to-market losses. This trade has positive roll-down thanks to the curve shape and tightness of bid-offer.

Entering and exiting any trade in these maturities carries execution risk, but this is not a major risk with this credit in these maturities as they are increasingly liquid.

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 MMC and the bid-offer spread costs.

MMC, is a member of the Dow Jones CDX NA IG Series 7 index, and is regularly among the top 150 quoted credits by dealers on a daily basis – especially recently. Bid-offer spreads are relatively wide at around five and ten basis points in five- and ten-year CDS respectively.

Fundamentals
This trade is significantly impacted by the fundamentals. The technical flatness of the credit and positive economics are helped by an event risk-prone outlook which, in our view, overwhelms the deteriorating idiosyncratic fundamentals of the credit.

Summary and Trade Recommendation
MMC's recent results were underwhelming and continue to put pressure on liquidity as cash flow remains negative. A change in tack on the possibility of a spin-off or sale of the Putnam business unit remains more negative for credit investors than positive (cash to share buybacks or tax bills) and as the debt and equity markets diverged recently, the likelihood of significant capital structure changes seems high.

As spreads approach one-year wides, but curves still significantly flat to fair value, the impact of any continued disruption in its service-oriented business could worsen results and leaves us with a conservative bear view. The curve trade offers better economics than the outright short. The duration-neutral steepener allows us to earn carry while offering us a time-insensitive bearish skewed position that benefits from any further deterioration in fundamentals (and indeed negative events) and will be considerably less affected, than an outright position, by any significant (or surprising) improvements.

Sell $16mm notional Marsh & McLennan Companies Inc. 5 Year CDS protection at 71bps and
Buy $10mm notional Marsh & McLennan Companies Inc. 10 Year CDS protection at 106bps to gain 8 basis points of positive carry.

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

Copyright © 2006 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).

FTD baskets and correlation - part 1

In this series of three articles, Shreepal Alex Gosrani, Priya Shah and Domenico Picone of the structured credit research team at Dresdner Kleinwort, examines the mechanics and characteristics of first-to-default basket swaps along with their potential use in correlation trading

Mechanics

A first-to-default (FTD) basket is a product that allows an investor to either buy or sell first-loss protection with reference to a basket of credit exposures. The typical size of the basket ranges from four to ten names. The contract is contingent on the occurrence of the first default in the reference portfolio.

As with a single-name CDS trade, the FTD contract consists of two legs, a premium leg and a default leg. The premium leg represents the spread paid to the protection seller and the default leg represents the impact of the first potential default. The contract is akin to a standard CDS agreement whereby, at inception, the spread is calculated such that the values of the two legs are set equal to each other. In this way, the FTD contract is an unfunded credit derivative.

Protection and credit risk
The protection buyer is in a short credit risk position as he is effectively shorting the credit risk of the first default in the basket. Correspondingly, the protection seller enters into a long credit risk position and is paid a premium for bearing this credit risk.

Similarities to vanilla CDS
Somewhat similar to a CDS contract, the protection seller, as the bearer of the FTD risk, receives a regular premium until the contract matures or the first default in the basket occurs, whichever happens soonest. So, if no default occurs, the FTD basket expires at the contract's maturity.

Upon default of the first credit the FTD protection buyer delivers the reference credit obligation to the protection seller in exchange for the notional amount (physical settlement), or alternatively, receives the loss given default (LGD) amount (cash settlement) from the seller. The LGD amount in the case of a FTD is simply the par amount of the defaulted credit less recovery.

Equity tranche similarities
Additionally the FTD resembles the equity piece of a synthetic CDO, particularly if we look at the behaviour of its premium as the default correlation changes. The equity piece of a CDO (e.g. the iTraxx Europe 0-3% tranche) and the FTD are long correlation products, both products experiencing a decrease in their premiums as correlation increases. However, an equity tranche does not terminate when the first default occurs, but when the notional is fully exhausted by the losses occurring in the synthetically referenced portfolio, or naturally at maturity of the transaction.

Funded FTD
A FTD basket can also be created in a funded format where an upfront payment, typically par value, is made by the protection selling investor. In this case it is called a credit-linked note (CLN) that references a FTD basket. The par value is then used to buy high quality collateral. If there is no default by maturity, the protection seller receives the par value of his initial investment.

When there is a default in the FTD basket, the high-quality collateral is liquidated and the proceeds are used to cover the loss suffered by the protection buyer with the remaining balance being returned to the protection seller. Through the CLN format the protection buyer is no longer exposed to the potential default of the protection seller but instead is exposed to the relatively remote default likelihood of the rated collateral, i.e. the protection buyer in such a case has mitigated the counterparty default risk of the protection seller, though of course still retaining the FTD exposure.

Benefit of credit rating
FTD CLNs are also rated by ratings agencies. The likelihood of one default in the basket will drive the rating. Equally important for the rating are: the credit quality of the collateral and the bankruptcy remoteness of the SPV.

In a CLN format, the SPV uses the note's proceeds to buy the high quality collateral, which is liquidated in the event of a default in the reference basket. In this way, the SPV permits a rating which only depends on the intrinsic credit quality of the basket. In addition, the rating allows comparison of the CLN premium with the premium paid by other products with the same rating.

The effect of dependency or correlation

Different industries may have strong or weak default correlation depending on how dominant the economic cycle is in affecting their performance – dominance of systematic risk (higher correlation) versus idiosyncratic risk (lower correlation) therefore has an important impact. Moreover, we also expect the default correlation to change over time, as the effects of the economic cycle (growth and recession) tend to dominate the performance of many firms.

As mentioned earlier, like synthetic CDO equity tranches, selling protection on the FTD basket is a long correlation position. A single default within the basket will result in the FTD terminating and therefore the probability of no loss is the key driver determining the fair FTD spread that the protection buyer pays.

As default correlation within the basket increases the likelihood of all credits surviving increases driving down the FTD premium. A similar argument can be used for a second-to-default (STD) basket. In the case of an STD the probability of paying a loss is one minus the sum of the probability of one or no defaults.

For an FTD, to better understand this concept of correlation, we consider below what happens to the FTD premium when the default correlation moves from zero to one.

Independence case – default correlation equal to zero
Assume that the FTD basket is made up of five uncorrelated names, each with a notional of $10m and flat credit curves. A no-arbitrage argument can be used to calculate the spread on this FTD basket.

In order to hedge the spread risk of this FTD basket a protection seller would have to buy protection via CDS contracts on each of the five names. In the event of a default the FTD protection seller would compensate the buyer and the contract would terminate.

At the same time, the CDS hedge on the defaulting entity would cover the loss suffered by the seller of protection on the FTD position. In addition, as the default correlation is zero for the trade's duration and the credit curves are flat, the CDS spreads of the remaining four names would not change and, crucially, can then be unwound at no additional cost.

Upper bound
Therefore, for zero correlation the spread of the FTD basket must be equal to the sum of the spreads (like an upper bound to the FTD spread) of the five individual CDSs, otherwise an arbitrage opportunity arises.

Perfect dependence case – default correlation equal to one
Assume now that the basket is made of the same five names with the same notional values as before, but with a default correlation among the assets equal to one. This means that if one default was to occur, all the other names would also default.

In this case the protection seller hedges to mitigate the payout the he would be liable for upon occurrence of the first default. If the first default does materialise, due to the other four entities also defaulting, the FTD protection seller would still be able to exit the trade at zero cost as the payment received from the CDS protection matches the liability due on the FTD protection leg. The name most likely to default in the basket is naturally that with the highest spread.

Lower bound
Hence, for 100% correlation, the spread of the FTD basket must be equal to that of the riskiest name in the basket since the contract is contingent only on the occurrence of the first default, and this can be treated like a lower bound.

The fundamental role of default correlation is fully appreciated when looking at the MTM function of FTD positions, hedged and un-hedged. The default correlation measures the changes to the default probabilities, and hence the CDS spreads, of the remaining names in the basket, given the occurrence of a default by one of the entities.

Modelling default correlation
As mentioned above, a FTD contract is akin to a vector of CDS contracts that automatically unwinds upon the first default in the vector at no further cost to the protection seller. When the two default correlation limits of zero and one are used, and assuming flat credit curves, the FTD basket is priced easily without the use of any analytical model. However, between these two limits, a model becomes necessary since the basket premium depends on the default dependency or correlation among the assets.

The model calculates the fair FTD basket spread by finding the spread that equates the default and premium legs at the outset. The default correlation is a key parameter of a FTD model. We outline a modelling approach in part 2 of this series of articles. In the following chart, we show what happens to the 5-year FTD basket spread as default correlation varies from zero to one.

We have analysed the following basket based on five North American CDS spreads whose levels (as of 20 September 2006) are shown below. For pricing, we assume flat credit curves, a 35% recovery rate, notional of $10m for each name, and we have used a one-factor Gaussian copula model¹ to calculate the fair spread.

For reasons explained above, when default correlation is zero the 5-year FTD spread is equal to 321bp, i.e. the sum of the CDS spreads. We then calculate the spreads of such a basket under various correlations ranging to 100%. At 100% correlation the FTD spread is equal to 178bp, the highest of the CDS spreads in the basket.

In sum, the FTD spread depends on the probability of paying a loss, which decreases with correlation. Correspondingly, the lowest FTD spread is attained when the default correlation is one. As explained above, the spread should reflect the probability of paying a loss. As correlation rises, this probability falls, which can be seen in the following loss distribution for the portfolio described in the table above.

Above, we see that as the correlation rises, the probability of surviving together, i.e. zero losses on the basket, rises. We show the loss distribution for 0%, 50% and 95% correlation among the five names detailed in the table above.

FTD and maturity
As maturity increases, the probability of one or more losses increases which drives up spreads at each correlation. This is highlighted in the chart below which shows how FTD spreads evolve with both maturity and correlation.

FTD performance as correlation changes
The FTD is priced with a correlation assumption at the outset. Baskets can therefore be used effectively as tools to trade correlation. Market participants can structure a FTD basket with entities of their choice and then proceed to trade the correlation of this basket.

Using a pricing model, we can ascertain how changes in correlation affect the MTM value, and hence the performance, of the FTD contract. A protection buyer benefits from a decrease in correlation as the probability of an entity defaulting increases which drives up the fair spread, increasing the MTM value of his position. Similarly an investor who has sold protection would benefit from an increase in correlation.

The following table and chart show some MTM changes to a short correlation trade (protection buyer) resulting from different correlation scenarios assuming the above un-hedged basket is priced at a correlation of 30% at inception. In the table, when correlation goes down, the short correlation position makes money.

In the graph below, we see of course that there is a greater dollar impact with a longer maturity for a given change in basket correlation.

Other factors affecting the FTD basket spread
In addition to correlation and maturity there are a number of other factors which also influence the FTD basket spread.
• Number of reference credits in the basket
• Individual default spreads of these credits
• Credit curve steepness
• Expected recovery
• Degree of basket homogeneity in terms of individual spreads and credit ratings

By focusing on a simple, homogeneous basket, in which all entities initially have the same spread of 100bp and the same notional amount of $10m each, we look at the impact of key factors.

Number of reference credits
As the number of reference credits increases, there is a greater likelihood of a single default occurring within the basket and, assuming constant correlation, there results an increase in the compensation demanded by the protection seller. We show this for the homogeneous basket in the following chart under various correlation assumptions for a five-year term.

Credit quality and expected recovery
As credit quality and recovery rates increase, the value of the FTD default leg decreases driving down the fair spread. The following two surfaces show the impact on both spread and delta as correlation and recovery change.

Consistent with our explanation earlier of the impact of correlation on the basket spread, we see that spreads decrease with rising correlation, as the risk of no losses increases. A higher recovery rate also reduces the impact of any default on the protection seller, driving down spreads.

Looking at delta, an increase in correlation reduces the sensitivity of the FTD to the underlying spreads thereby reducing each entities delta. Rising recovery rate also implies a lower delta.

This is because higher recoveries entail higher implied default probabilities and, thus with more risk in the portfolio, FTD deltas go down². As evident from the graphs spreads and deltas are generally more sensitive to basket correlation relative to recovery rates.

Homogeneity of reference credits
A FTD basket structure with credits of similar spreads or credit quality is not driven by a single spread. On the other hand, a single distressed credit within a basket would have a dominant effect on the FTD spread.

This is demonstrated in the chart below (for different correlations) which assumes a basket of five credits, one of which is currently trading at an increasingly wider spread to the other four which are held at 100bp. As can be seen, when the single distressed spread widens, the FTD spread is increasingly dominated by this single name.

Notes
1. We refer readers to Laurent and Gregory, Basket Default Swaps, CDOs and Factor Copulas (2003), and Hull and White, Valuation of a CDO and an nth to default CDS Without Monte Carlo Simulation (2003) for more on such models. In part 2 of this Research Note we will show some key equations.

2.
where s, R and t are the CDS spread, recovery rate and time interval respectively for each name, and where p modelled as the (risk-neutral) default probability which increases as recovery rises. For a proof see Meneguzzo and Vecchiato, Copula Sensitivity in Collateralized Debt Obligations and Basket Default Swaps Pricing and Risk Monitoring (2002). For more on delta, see Delta-hedging: a primer, published by Dresdner Kleinwort on 12 April 2006.

© 2006 Dresdner Kleinwort. This Research Note was first published by Dresdner Kleinwort on 3 October 2006.