Contents

European LCDS to go to the vote

Mexican stand-off makes result uncertain

Continued differences of opinion among the major dealers over the future path for loan CDS in Europe mean that a vote will now be taken among market participants as to how to proceed. The date of the vote has not yet been firmed up but is expected in the next week or so, and a likely candidate is next Monday, 18 December.

There is also uncertainty as to what exactly will be voted on. For example, it is not thought to be as clear-cut as simply whether a reference entity or reference obligation approach will be adopted.

One such approach will involve specifying a named reference entity and the other will define the reference entity as each obligor under the reference obligation from time to time. In both cases, the specified reference obligation may be one or more designated tranches, or all loans of the same lien under a facility as a whole.

One of the issues of concern in this respect is that the proposed reference entity approach for documentation might not prove to be fully satisfactory. The fear is that, unlike the reference obligation approach, sellers would theoretically be selling protection blind and be open to taking delivery of potentially highly undesirable paper.

In this case the vote could centre on how extensive a field of reference entity/reference obligations the market opts for as standard. Equally, the issue of cancellability - which was previously thought to have been resolved - will also likely be voted on.

As one dealer says: "Overall, everyone still has their preferences, so the vote could lead to an unexpected result."

Indeed, as another source observes: "It could end up coming down to who among those with a vested interest are the better advocates and how broad a constituency gets the vote. There are a lot of peripheral players who seem to be wavering from one side to another on an almost daily basis and not necessarily being consistent between issues."

Determining whether European LCDS contracts are cancellable or not is perhaps the most significant issue. For example, the International Association of Credit Portfolio Managers came out in favour of contracts being written on a cancellable basis (see SCI issue 16), but many CLO managers are opposed to such documentation.

However, at the centre of the debate are the major credit derivatives dealers who have aligned apparently counter-intuitively irrespective of loan book size. Notably, three of the potentially largest dealers in the contracts - Citigroup, Goldman Sachs and JP Morgan - are understood to be unwilling to become involved in the market until the European contract is made non-cancellable.

MP

Investors drive Asian structured credit

Asia Pacific buy-side is leading the market, while Japan's would like to

New investors are driving volumes in structured credit in the Asia Pacific region and – to a reasonable extent – seeing supply to meet that demand. However, in Japan there is still a demand-supply mismatch.

The growing presence of private banks and hedge funds is having a profound impact on the Asian fixed income market. As Greenwich Associates consultant Tim Sangston explains: "Every year, Greenwich Associates identifies a representative sample of the fixed income investor base in Asia. In 2005, that research universe included 40 private banks and hedge funds. By 2006, that number had increased to nearly 150."

Sangston's comments are based on the results of Greenwich Associates' 2006 Asian Fixed Income Investors Study. The report found that the influx of hedge funds and the surge in private banks' investable assets – thanks to Asia's booming economies producing increased personal wealth – are hastening the development and maturation of the local fixed income markets. In particular, these investors are driving growth in derivatives, notably credit derivatives.

Greenwich Associates' research shows that credit derivatives trading volume rose sharply for the second year in a row in Asia – nearly doubling from 2005 to this year. Hedge funds accounted for the substantial share of that volume.

They generated about US$5bn in flow credit derivatives over the past 12 months, but the impact of the hedge fund boom is seen more clearly in some of the more sophisticated products. In structured credit derivatives, which represent a relatively new asset class for Asian investors, Greenwich notes that hedge funds account for U$4bn in annual trading volume out of a total of US$14bn.

Supply of structured credit in the region looks likely to continue, according to a report released by Fitch this week. It says: "Singapore-based CDO asset managers are expected to continue to be active issuers in 2007. Fitch expects more balance sheet CLO issuance from regional Asian banks seeking to transfer part of their corporate credit risk in response to Basel II, with the trend of a greater percentage of the initial portfolio consisting of collateral from Asia Pacific likely to continue."

Furthermore, the rating agency notes: "Taiwan's structured credit market should continue to be dominated by locally originated CDO issuance, with more asset-backed commercial paper coming to market more regularly than term notes. CLOs backed by domestic balance sheet corporate loans and with liabilities issued either by way of term notes or ABCP are expected to be active issuers in 2007. In addition, CDO issuance driven by structured bonds that include US dollar assets such as RMBS, CDOs or principal protected notes or even principal-only securities are likely to continue issuance next year."

Meanwhile, Greenwich Associates' 2006 Japanese Fixed-Income Investors Study finds that Japanese accounts are keen to follow in the footsteps of other investors in Asia. However, market conditions are making this difficult.

"Investors are generally very interested in credit derivatives, but the big issue is they can't find the products available in the market. Plus, if the products are available, the spread is too tight, and liquidity is a problem as well," says Greenwich consultant Dev Clifford.

Such issues are likely to be exacerbated in Japanese structured credit next year, according to a report on the country also published this week by Fitch. While issuance is likely to remain stable, the agency believes that the clarified accounting treatment of synthetic CDOs under the new GAAP rules will significantly increase the number of potential CDO investors in Japan. In particular, regional banks are likely to increase their CDO investments.

MP

CDS activity jumps

Significant move in November's trading activity potentially driven by new market dynamic

Average liquidity in CDS jumped dramatically in November and the variation in individual liquidity also increased, according to Credit Derivatives Research (CDR). The spike in activity is likely driven by a variety of factors.

Dave Klein, research analyst at CDR, explains: "It is too soon to determine whether this rapid increase in liquidity is an anomaly or reflects new market dynamics. It is possible that the record CDO issuance in November and ramp-up (or expectation) of CPDO-type issuance has temporarily altered market dynamics."

CDR's Market Depth index has shown a gradual rise over the course of the year, but its Market Activity index has risen dramatically in the last two weeks (see SCI's data pages). The Market Activity index has generally shown a "day-of-week" effect, with bid-offer counts rising from Monday to Wednesday and then dropping from Wednesday to Friday.

This behaviour is also visible in the Market Depth index. In November, the Market Activity index still exhibited day-of-week differences (muted a bit by the US Thanksgiving holiday) but activity jumped dramatically.

"To analyse what changed, we considered market activity on two days, November 1 and November 29 (both Wednesdays). The number of liquid credits counted in the Market Depth index was 1181 on November 1 and 1225 on November 29, about a 4% difference. In contrast, the Market Activity index more than doubled from about 186 on November 1 to 399 four weeks later," says Klein.

He continues: "We see something remarkable when we group companies by the number of bid-offers counted on the two days. It is clear that the distribution has flattened dramatically since the beginning of November. This implies that more names are receiving greater attention in the CDS market and that the market is more discriminating in its attention to individual names."

The impact of this broadening of market scope has not been fully felt to date. "We have yet to see bid-offer spreads narrow significantly - another indicator of increased liquidity - and we would not be surprised to see some new names in January's roll of the CDR Liquid 50 North America IG Index," Klein says.

CDR tracks both market depth and market activity in liquidity indices. The CDR Global Market Depth Index is a daily measure of how many names are actively traded in the credit derivatives market, while the CDR Global Market Activity Index is a daily measure of activity within the global CDS market. Market Activity index values are simple averages of total bid-offers of all names in the Market Depth index multiplied by a constant.

MP

European REIT CDO debuts

New deal for real estate-hungry investors

The first European cashflow real estate investment trust (REIT) CDO, Taberna Europe CDO-1 from Taberna Capital Management, is currently being marketed by Merrill Lynch. The €600m deal comprises six rated tranches and a €23m subordinated piece, and has been warmly welcomed by European investors starved of real estate-linked structured credit assets (see also this week's News Round-up).

The proceeds from the note issuance will be invested in a static portfolio of approximately €600m of subordinated debt and senior unsecured debt securities issued by REITs and real estate operating companies (REOCs), as well as CMBS and commercial mortgage B-notes. The collateral will be selected and monitored by Taberna Capital Management – representing its eighth cashflow REIT CDO but the first to be backed by predominantly euro-denominated assets.

The portfolio comprises 55% REIT subordinate debt securities, 40% senior unsecured REIT securities, 2.5% CMBS and 2.5% B-notes diversified across various European countries.

60% of the collateral will be purchased before close, with the remainder bought over a twelve-month ramp-up period. Collateral added during the ramp-up period will be subject to credit quality, recovery rate and spread parameters, as governed by the transaction documents.

The transaction has a five-year reinvestment period for the CMBS and B-note components only, carrying an expected weighted average rating factor of double-B plus and an expected weighted average spread of 1.82%.

The deal's Class A1 and A2 tranches are both rated triple-A by Fitch and S&P, and have 9.7-year and 10-year WALs respectively. The Class B, C, D and E tranches are rated from double-A to double-B and all have a 10-year WAL.

Taberna Capital is expected to retain 50% of the equity and 100% of the Class E notes in the transaction, thereby more closely aligning its interests with those of the senior noteholders.

Derivative Fitch highlights a number of other strengths in the deal, notably an increased CMBS and senior debt collateral bucket (at 42.5%) relative to previous CDO transactions managed by Taberna Capital. Such collateral is more senior and more highly rated in comparison to REIT trust preferred securities. Overcollateralisation and interest coverage tests will redirect cash flows to more senior notes in the event of collateral deterioration.

Up to 15% of the portfolio may nevertheless be composed of perpetual securities, which have no stated maturity date. Limited historical performance data on the majority of the underlying collateral, together with limited diversity by obligor and industry should also be taken into account.

However, these concerns are mitigated by a number of conservative assumptions being applied by the rating agencies to the collateral that is expected to be added during the ramp-up period. Furthermore, the agencies' default and recovery rate assumptions account for the industry and obligor concentrations in the expected portfolio's assets.

Taberna Capital has a demonstrated track record with respect to CDOs of US real estate collateral and has implemented the necessary underwriting and monitoring infrastructure in order to support its first European deal. The manager is currently in the process of establishing a London office that will acquire local talent and assist in the management of its European real estate CDOs.

MP

Focussing on the details

This week, Dean Smith senior portfolio manager at Highland Financial Holdings Group, answers SCI's questions

Dean Smith

Q: When, how and why did you/your firm become involved in the structured credit markets?
A: Our firm and our professionals have been involved in structured credit for many years. For example, we executed some of the earliest ABS CDOs. Unlike many other deals in the 2000 and 2001 vintages, however, our deals performed very well because we stuck to sectors that we knew and understood well. More recently, we have been actively involved in single name ABS CDS, bespoke synthetic CDOs, plus other cash or synthetic structured deals.

Q: In your view, what has been the most significant development in the credit markets in recent years?
A: The single name, pay-as-you-go (PAUG) ABS credit default swap has fundamentally changed the market. Prior to its widespread acceptance in mid-2005, there really was no effective way to express a bearish view on individual ABS credits, or sectors. And on the long side, investments were limited to whatever bonds the Street happened to have in inventory. This was typically a very small part of the universe. Prior to the PAUG CDS, the decision for credit investors was either to buy a security, or to do nothing.

Naturally, the CDO was itself a landmark innovation. It may be the ultimate credit derivative, as it combines fundamental credit risk with dynamic structural credit features, and portfolio/correlation effects. While the CDO has been around for some time, we would argue that they are still not very well understood.

Q: How has this affected your business?
A: The PAUG CDS is a core instrument for us. Even if we don't have a position in a particular security, seeing where the CDS trades is highly informative. In fact, with Mezzanine CDO managers, who are usually buy-and-hold investors, taking such a large share of the new issue HEL offerings, sometimes the CDS is the only way to trade a particular name.

Most importantly, the CDS allows us to act upon a negative view we have on a particular credit, and permits the execution of all types of relative value trades that would have been impossible only a short time ago. Bid-ask spreads are still wider that we would like, and can limit our ability to put on trades that look attractive in principle, but we expect that over time liquidity and transparency will improve.

However, a large share of the single name CDS business takes place via bid-lists, rather than dealer axes. This means that investors need to be able to respond very quickly to sometimes very extensive lists of names on which some party wishes to buy or sell protection. This has necessitated building databases of "pre-run" analytics on hundreds of securities in order to be responsive. These runs need to be updated frequently in response to changes in rates, or new information contained in remittance reports.

Q: What are your key areas of focus today?
A: Our major focus from a credit perspective is on the effects of regional housing markets. The headline news on national housing price averages or inventories is of almost no use for actual trading activity. The diversity across the 350+ MSAs is simply too great to make the national average a meaningful data point.

We look at loan level detail, and build up credit analysis from this detail based on a number of regional economic factors, combined with loan-specific attributes. Differences in regional performance will be an important feature of the landscape for the foreseeable future. We have been focusing on this for some time – the market in general is slowly coming around to this view.

Q: What is your strategy going forward?
A: Our strategy is simple – buy deals that have adequate credit enhancement, sell deals without it. Easy enough to say, but the devil is in the details. For example, there is a current focus in the market on the poor early performance of the 2006 vintage of sub-prime mortgage loans.

We agree that many 2006 deals are off to a weak start, but that is only half the analysis. The key questions are: what does that early performance indicate about the future; and how does that compare to the credit enhancement of the actual securities?

We commonly buy securities where we foresee relatively high mortgage default rates, but if there is adequate subordination, excess interest or other forms of credit enhancement, even fairly large projected losses may not be a problem. Frequently, such securities can be bought cheaply, since the market is fixated on collateral performance.

Contrariwise, we see deals in the market where projected losses are very low, yet mezzanine or subordinate securities can be highly vulnerable to write-off or downgrade. Weak structures can undermine the total return, even for a high quality collateral pool.

Q: What major developments do you need/expect from the market in the future?
A: The structural complexity of most CDOs makes rigorous quantitative analysis extraordinarily challenging. Consequently, what most investors resort to is scenario analysis, in which various stresses to defaults, losses or other fundamental factors or stated variables are assumed, and the resulting effects on CDO tranches is observed by the analyst.

The problem with this approach is that it does not take likelihood into account. This deficiency can produce poor investment decision making – both from "false positives" and "false negatives".

The market therefore needs a better approach to risk in ABS portfolios and CDOs, one that draws on modern portfolio theory and arbitrage free pricing techniques. Tranching of the ABX (expected to commence in January 2007) will help, but is only a part of the solution, since the index references such a limited number of credits.

The downside of an agreed upon analytic framework, of course, is that some of the inefficiencies in the market will be arbitraged away. Exploiting these inefficiencies constitutes an important part of our strategy.

On balance, though, a better model will help the market. Option trading was a backwater before Black-Scholes, and synthetic corporate credit trading didn't really take off until the Gaussian Copula model was articulated.

Pricing transparency also needs a great deal of improvement. The tradition in the cash ABS markets has been for auction results to be widely disseminated, at least with respect to runner-up or "cover" prices. This enhances liquidity for everyone, and facilitates risk taking and market making.

PAUG CDS auctions have seen a much lower level of disclosure, especially from hedge funds. We understand why a seller might think it is in his interest to not provide "colour", but ultimately we view this as short-sighted.

We frequently talk to dealers who decline to participate in auctions held by non-disclosers, on the grounds that if one goes through the trouble of coming up with bids or offers, they ought to be entitled to know where the market clearing levels are. After all, it works for Sotheby's and the US government.

About Highland Financial Holdings Group
Founded in 1998, HFH Group is an investment management firm specialising in fixed income alternative investment products. The firm's strategies emphasise relative value, performance stability and liquidity and cover the entire credit spectrum of the short duration MBS/ABS markets.

HFH was established with one goal in mind: to provide consistent, superior absolute investment returns independent of the overall market. In the eight years since, HFH has pursued this goal through a combination of focused research, rigorous investment selection and disciplined risk management.

HFH was awarded Hedge Fund Manager of the Year in 2004 by Risk Magazine for its commitment to risk management and transparency and was a MARHedge 2006 Performance Awards finalist. HFH is currently registered with the Securities and Exchange Commission as an Investment Advisor.

Deutsche ramps up Asian structured credit

The latest company and people moves

Deutsche ramps up Asian structured credit
Following the arrival of Sajid Javid as head of global credit trading, Asia (ex-Japan) at Deutsche Bank in Singapore (see SCI issue 10), the bank is believed to be reorganising and revamping its structured credit operations.

Javid relocated from Deutsche Bank in London, where he was global head of emerging market structuring, and is now regionally responsible for all cash and derivative credit trading, CDOs, securitisation, structured finance, convertibles and for the commodities business. The unifying nature of his new role is seen as a clear signal of the bank's plans to expand its business in Asia.

The two previous co-heads of credit trading are staying at Deutsche Bank, with Wai-Hong Luck now taking a senior prop trading role, while Chetan Kumar Shah will continue to run structured credit finance. However, Raj Shourie, former head of Deutsche's Asia securitised products group, has left the bank.

Deutsche is understood to be in the process of making a number of hires in structured credit for the first quarter of next year and to have plans to increase head count further going forward.

Truett to retire from CDPC
Charles Truett, head of portfolio management at Primus Guaranty and member of its operating committee, will retire from the company effective 31 March 2007.

Thomas Jasper, Primus ceo says that the company expects to review external and internal candidates as it works to fill Truett's responsibilities during the first quarter of 2007. Before joining Primus in 2001, Truett served for over 30 years at Bank of America and Continental Bank, holding a number of senior credit management positions.

Calyon announces its structured credit intentions
Calyon formally announced yesterday that its credit markets & CDO product line would grow from its current 180 professionals to more than 240 globally in 2007. At the same time, the bank outlined a number of temporary measures to cover recent leavers from its US credit operations.

"The growth of our credit markets & CDO activities will come from a balanced mix of synthetic correlation and cash CDO credit products, with a particular emphasis on North America, where we intend to double our CDO activity next year," says Loïc Fery, md and global head of credit markets & CDOs at Calyon.

The bank also confirmed that a group of professionals from its London office are temporarily working out of Calyon's CDO structuring, sales and trading team in the US. The move – which includes Edouard Bremond, head of European cash CDO structuring – is in response to recent departures from the US CDO structuring unit.

Calyon says that it expects to appoint a new head of US cash CDO structuring within the next few weeks, and will also add others to its New York CDO team.

Goldman takes Amaranth 17
Goldman Sachs is understood to have hired 17 traders from failed hedge fund Amaranth Advisers. The team includes 14 credit specialists in New York and three in Singapore, and will be led by Gregg Felton, who managed Amaranth's investments in debt markets.

Barclays Capital makes two senior US hires
Barclays Capital has recruited Bill Archer as an md and head of US leveraged credit risk. He will report to Ian Prior, md and chief credit officer for the Americas.

The bank has also appointed Vincent Breitenbach as an md and head of credit research, Americas. Breitenbach will be responsible for US investment grade, high yield, securitisation and Latin America corporate credit research. Based in New York, he will report to Mark Howard, md and co-head of research.

Archer joins from Goldman Sachs, where he was an md and co-chairman of the firm's capital markets committee. Prior to that, he worked at Deutsche Bank for three years and at Bankers Trust for 18 years.

Breitenbach joins from Countrywide Financial Corporation, where he was an md with responsibility for fixed income investor and bank relations, rating agency relationship management and internal capital structure advice. Before that, he worked at Lehman Brothers for more than eight years, where he was an md in corporate bond research covering specialty finance companies, US banks and securities firms.

MP

Dexia makes WISE move

A round up of this week's structured credit news

Dexia makes WISE move
Dexia Credit Local is marketing the first securitisation of wrapped infrastructure bonds – the £63.75m WISE 2006-1 deal.

The transaction is structured as a partially funded synthetic CBO and transfers the credit risk associated with a pool of £1.5bn infrastructure bonds originated under both private finance initiative/public private partnership projects and direct or indirect government-regulated entities that provide services to the general public, for example, the supply of energy or water. The initial reference portfolio comprises 31 reference obligations that are wrapped by triple-A rated monoline insurers.

Rated by Standard & Poor's, the deal consists of £30m triple-A rated Class A notes, £22.5m double-A rated Class Bs and £11.25m double-A minus Class C notes.

Markit launches LCDS pricing service
Markit has launched a pricing service for loan credit default swaps (LCDS). The service provides same day and T+1 spreads for over 300 reference entities and tiers traded in the European and North American markets.

Markit draws prices from contributing dealers and cleans them to create a composite which is made available at 4:00pm daily in London and New York. Sell-side firms using the service will see spreads on a particular reference entity when there is a minimum of three dealers making markets in that name, while buy-side firms will be able to access even thinly quoted entities.

The firm will also offer valuations on LCDS. Markit says that the LCDS pricing service complements Markit's role as official calculation agent for LevX and LCDX, the North American LCDS index, which is expected to launch in the first quarter of 2007.

Tim Frost, principal, Cairn Financial Products, comments: "It is a sign of the increasing maturity of the product that Markit has been able to extend its well respected pricing service into the LCDS market. We would expect the availability of this pricing service to stimulate further growth, and we look forward to Markit launching its LCDS parsing service soon."

Fitch asks if Europe can deliver on CRE CDOs
Fitch Ratings says that the strong reception for the recent debut European real estate CDO demonstrates the high demand for leveraged exposure to European commercial real estate. Further issuance is inevitable; however, the agency questions whether the strength of demand is strong enough to overcome significant barriers to entry for all of the CDO managers hoping to bring similar deals to market.

In a comment, the agency notes that despite the considerable market excitement surrounding BlackRock's €342.5m Anthracite Euro CDO 2006-1 transaction (see SCI issue 13), it is questionable whether the transaction will herald a flood of issuance.

"Investor demand looks far from being exhausted, which will doubtless whet the appetites of a range of potential issuers eager to replicate the attractive returns apparently in store for BlackRock," says Euan Gatfield, director, Fitch Ratings. "However the barriers to entry into the club of RE CDO managers will be high."

GEM integrates with T-Zero
Global Electronic Markets (GEM) has announced a license sale and successful integration of the trade comparison component from its FpML Mediator software platform to T-Zero. This integration enables T-Zero's buy-side clients to perform trading activities, such as assignments and early terminations on CDS trades, with the assurance that they are in synchronisation with the trade state maintained by the DTCC's Deriv/SERV derivative trade confirmation service.

When a T-Zero client wishes to perform a transaction on an existing CDS, T-Zero retrieves the deal's details from DTCC and uses GEM's component to compare the client's version of the trade with that retrieved from DTCC. This enhancement saves clients the effort of searching for the applicable trade and minimises the operating risk of selecting the wrong trade.

MP

Trading ideas - drugs and digs

Tim Backshall, chief credit derivatives strategist at Credit Derivatives Research, recommends a pairs-trade between CVS Corp and Host Hotels & Resorts Inc

We believe that a superior way to outperform the broad credit market is to combine strong fundamental analysis with detailed technical (or quantitative) modelling. This report describes a simple process that we update monthly, which offers outstanding historical performance and a critical insight into risks and opportunities in the investment grade credit markets.

Fundamental analysis is undertaken by Gimme Credit. On a daily basis, each analyst reviews each credit and assigns a score of -1, 0, or +1 based on expectation of balance sheet performance over the coming months.

This credit score is not a relative value measure and is not designed to judge whether the market has fully priced in the fundamental outlook. These scores should be thought of essentially as meaning deteriorating, stable, and improving fundamental outlooks respectively.

In order to try and identify value, we overlay quantitative techniques. Our quantitative analysis is based on a number of models and parameters. These models range from simple to complex and span from intra-sector to cross-asset class. Many of these are unique to CDR and we describe one useful measure below.

Crossing the capital structure
Credit market participants have long recognised that the equity markets can provide insightful and timely information to aid in the credit selection process. From the earliest models of the capital structure produced by Merton (1970), a link, however strained, has been established between equity prices, equity volatilities, and bond yields or credit spreads.

One of the most frequent uses for these so-called 'structural models' has been to calculate a default probability. It is beyond the scope of this article to detail the methodology (see CDR's Trading Techniques 'Credit Risk Models' features for more details), but it is enough to know that by using option-pricing theory we can calculate an implied probability of default from the market's price of equity and equity volatility.

This default probability (DP), most often assigned a one-year horizon, is useful as a relative value measure. Its derivation from a highly liquid and transparent market - the equity market – can provide important signals for the management of credit.

Credit Default Swap (CDS) prices can also be manipulated to produce an implied default probability. The simple formula (see CDR's Trading Techniques 'Basis Trading' feature for details) tying CDS prices, default probability and recovery rates is as follows:

CDS Premium = Probability of Default * (1 – Recovery Rate)

So, as an example, if Albertson's is trading at 283bps for five year protection, ignoring discounting and compounding (for simplicity's sake) and assuming a recovery rate of 40% (industry standard), we have a probability of default of 4.72%.

Identifying the trade
Armed with our two measures of default risk, one from the equity market and one from the CDS market, we can begin to examine any relationships. At first, as has been found by many practitioners, the results appear disappointing. Given the risk neutral nature of the CDS market models and the real-world nature of the equity market models, relating the two is not simple.

We take a different and unique angle of attack. Instead of comparing absolute levels of default risk, we will compare relative ranks. It is reasonable to expect that an issuer with a high ranking CDS level (wide CDS level) relative to its peers should also have a high ranking equity-implied default probability (high default risk). We compute percentile rank scores for each issuer based on each measure and look for inconsistencies.

In our search for outliers we normally look for rank differences of greater than 40 percentage points between the equity and CDS markets. In previous month's analyses, we have had to drop that differential to as low as 20 percentage points for deteriorating credits in order to encompass enough names to generate the 'pairs' trade. This month, given our extended universe and increased dispersion, we can revert to our original strategy of 40 percentage points.

We note that there are only a few candidates in each of our 40% differential ranges and highlight that with the number of opposed credits in each segment. This is especially evident in the high-ranked CDS-implied default risks where there is considerable 'capital structure change' and the equity-implied model is unable to cope well with these changes.

More importantly, an analysis of the chart below also shows that the upper right and lower left corners are seeing more clustering than usual – especially in the upper right.

This portion of the chart (upper right) is where investors in both markets agree that the issuer is a high risk. It is interesting that the tendency for investors across the two markets to agree on the demons but be less in sync on the angels could be simply a reflection of momentum traders in both markets chasing the idiosyncratically risk-challenged issuers and ignoring the stable or low-event risk names.

We also note that in the upper left quadrant there are a number of credits with low equity risk ranks ('good' credits) but high CDS risk ranks ('bad' credits). These reflect the increasing number of LBO-prone names and the divergence that we see in the case of an LBO.

Traditionally we would expect spreads to tighten as equity prices rise and vice versa but in the case of an LBO it is quite different. Equity prices rise due to the premium that private equity will pay to get the deal done but spreads will sell-off dramatically at the merest whisper of an LBO as the company is expected to raise leverage considerably. This raises a critical issue in our selection criteria and one that we will see is very relevant this month.

An equity-implied rank significantly lower than the CDS-implied rank should be interpreted as: the CDS premium is higher than its default risk would deserve, relative to its peers. On a relative basis, we would expect these names to tighten relative to their peers, based on our historical analysis of this asset selection process.

The chart graphically displays the individual issuers, their equity- and credit-implied default risk rank and their fundamental outlook. We look for credits with major rank differentials (top left segment or bottom right segment) that converge with our fundamental outlook, i.e. a green triangle (improving fundamentals) in the green shaded area (tightening expected from quantitative models). The blue dotted line joins the two issuers – notice the deteriorating name is just inside our 40% segment – this is still the most significant divergence in three months.

A further point of interest is the migration of all issuers towards the bottom-left to top-right diagonal – i.e. getting closer to agreement between the two markets as spreads widened out this month. It is perhaps especially noticeable at the extremes (as we mentioned above) – top-right clusters of deteriorating credits and bottom-left clusters of improving credits. It is maybe not so surprising that the first credits to become more aligned in risk between the two markets would be the most extreme – very strong or very weak credits – but we still significant dispersion over any argument of market alignment.

We then analyse the list of the names (Deteriorating, Stable, and Improving fundamental outlooks respectively) covered in this report and highlight the important next step – convergence between the technicals and the fundamentals.

This has resulted in the selection of two credits:

CVS Corporation (CVS), through its wholly-owned CVS Pharmacy and Revco subsidiaries, operates a drug store chain in the Northeast, Mid-Atlantic, Southeast and Mid-West areas of the United States.

Host Hotels & Resorts Inc. (HST) is a real estate company which owns or holds controlling interests in upscale and luxury full-service hotel lodging properties.

Please note that Host Marriott LP is the issuer of bonds and Host Hotels & Resorts Inc. is the issuer of equity – we use them interchangeably when discussing the CDS contract which appears to reference both.

We recommend a pairs trade between Host Hotels & Resorts Inc. (Improving) and CVS Corp. (Deteriorating). We are pleased that the relative-value trade is significantly positive carry without us being exposed to any curve trades (as we have had to resort to in the past). At current levels the short HST protection (long credit) and long CVS protection (short credit) position offers around 93 basis points of positive carry.

Risk analysis
This 'pairs-trade' carries a direct risk of non-convergence. In other words, there is the possibility that the names will not tighten and widen as expected. However, based on historical performance of the technical indicators, the cushion provided by the positive carry, and the conviction of the fundamental analysts, we believe these risks are well mitigated.

Liquidity
Both CVS (a member of the CDX.NA.IG 7 index) and HST (a member of the CDX.NA.HY 7 index) offer good liquidity in the CDS market. They consistently rank in the top 200 issuers by quoted volume on a daily basis. They also have tight bid-offer spreads, regularly less than five basis points and sometimes down to three basis points. We see no concerns with execution of this trade.

Fundamentals
This trade is based on the convergence between our fundamental analysts' recommendation and our unique model-based outlook. For more details on the fundamental outlook for each of HST and CVS, please refer to Gimme Credit.

Summary and trade recommendation
We use a bottom-up analysis of the most liquid (CDS-based) investment grade credits combined the fundamentals with the technical outlook from CDR's proprietary models to highlight anomalies across the market. With spreads extremely tight but dispersion remaining higher than at previously low spread levels (possibly driven by LBO concerns), bullish investors seem to be discriminating more effectively than in the recent past providing more opportunities for relative-value trades. The increase in liquidity since the last roll allowed us to extend our universe to over 300 IG and XO credits and while we notice increased alignment of debt and equity markets' risk assessments at extremes of strength and weakness, there is still significant uncertainty (and therefore opportunity) in the more moderate credits.

The two most outstanding names, in terms of relative value and strong convergence between technicals and fundamentals, are Host Hotels & Resorts Inc. (improving balance sheet and stable credit ratios) and CVS Corp. (management clearly willing to sacrifice credit quality) – Improving and Deteriorating outlooks respectively – and are the same as last month. At current levels a matched maturity trade offers close-to-duration neutrality and strong positive carry.

Buy $10mm notional CVS Corp 5 Year CDS protection at 20bps and

Sell $10mm notional Host Hotels & Resorts Inc. 5 Year CDS protection at 113bps to gain 93 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).

Credit range accruals - part 3

In this series of three articles, Shreepal Alex Gosrani, Emiliano Formica and Domenico Picone of the structured credit research team at Dresdner Kleinwort examine range accrual structures in a credit context

Expected performance
We use here our proprietary model (the features of which we explain in the appendix together with aspects of its calibration) to price and to assess the break-even position of several credit range accrual structures.

? Price – it is the credit range accrual spread over Libor (in the table below) that is paid to the investor.
? Breakeven – We calculate the proportion of days in an interval (e.g. quarterly) that the iTraxx needs to close within the range for the investor to break-even with respect to a benchmark like Libor [Libor/(Range coupon)], i.e. a number of days such that the investor earns exactly Libor¹. Considering the stylized example presented on page 8, we see that a break-even position would mean that the iTraxx needs to be observed within the range about 60% of the time².

Here we show some potential spread levels as calculated with our model for the five year maturity together with the break-even proportion of days that the iTraxx needs to finish in the corridor assuming that Libor is 3.00% and that the current level of 5Y iTraxx Europe is 33bp, as it was on 18 July 2006. For pricing, we use the market-implied forward curve of that day.

In this section, unless otherwise specified, with the term 'static range accrual' we will refer only to the static structure where the range is adjusted to the spot level of the index. The performance of the static range accrual with the barriers adjusted to the forward will not be object of this analysis.

Static credit range accrual
For the static range accrual we have:

When the range is widened to 20/70, the spread drops considerably to 148bp, and this is clearly because of the higher likelihood of finishing in such a range as evidenced historically, and due to the market-implied forward curve.

On the other hand, with a wide range like 20/90 the spread is only 35bp and hence, iTraxx has to be observed 90% of days within the barrier for a break-even investment.

Dynamic credit range accrual
For the dynamic range accrual with the corridor adjusted to the new spread level every six months (when the index rolls) we have:

If we consider a corridor with the upper and lower barrier set respectively at 20% above and 20% below the current spread level, our model returns a spread over LIBOR of 128bp. This means that the iTraxx has to trade 71% of days within the range for the investor to break-even. Widening the corridor to -30% and +30% the spread drops to 57bp, and the index is expected to trade within the corridor for 86% of the time.

Looking at the historical annual volatility
Considering that the historical annual index volatility is around 43%, which equates to a semi-annual volatility of around 30%³, it seems that the probability of observing the iTraxx trading outside the corridor has been somehow underestimated. However, a thorough analysis of the index behaviour revels that the jump component contributes for more than 50% of the overall index volatility. Without including the jumps in the series the historical annual volatility of the index drops to 21% (15% semi-annual) and therefore a corridor of -30% and +30% is indeed wide enough to contain the index for most of the time.

Only when very large jumps occur we can actually observe the iTraxx breaching the corridor unless a relatively high positive or negative drift of the process push rapidly the index towards the range boundaries. As we will show later, in this context, a key role is played by the market forward curve used to calibrate our model.

Sensitivity analysis
The performance of the static and dynamic range accrual mainly depends on two key variables used in our model:
? Spread volatility, clearly affected by parameters like average jump size and jump arrival frequency
? Market forward curve that is used to estimate the long-term mean of the mean reverting process specified

In an earlier table, we listed all the parameter values that constitute our 'base case'. However, the future values of these variables may diverge from those historically observed, and may lead to different estimates in the future. Hence to complete our study, we prepared some sensitivity analysis and stressed those variables from their base cases. To complete our analysis we also tested the performance of these two products with maturities longer than five years.

As explained before, a static credit range accrual can be seen as a view on future spread volatility and future spread levels of the index. We expect, therefore, the spread to be particularly sensitive to changes in both spread volatility and forward curve.

For the dynamic credit range accrual, on the contrary, we expect the spread to be particularly sensitive only to changes in spread volatility since the corridor is periodically readjusted to follow the new index levels.

The impact of the index volatility
To study the impact of the index volatility on the pricing spread for both type of product, we made vary from -50% to 50% (with respect to the base cases) the average jump size and the jump arrival frequency (see table on page 19) of both positive and negative jump leaving the Garch parameters unchanged. The results obtained for the static range accrual are listed in the table below.

? Static credit range accrual - Plotting the fair coupon over and ö, we can see that clearly the spread tends to increase as and do. A higher value of means that we expect to observe on average larger jumps while with a higher value of we expect to observe a greater number of jumps occurring throughout the life of the product. In both cases the probability of observing the index breaching the corridor tends to increase and consequently the reward for the investor tends to be higher. As displayed in the charts below, for every choice of range, the higher the volatility of the index the higher the coupon paid to the investor.

? For the dynamic range accrual, as expected, we obtained similar results (see table below).

With higher index volatility due to either larger jumps or greater jump frequency the probability for the iTraxx to trade within the range decreases and therefore investors have to be compensated with a higher spread. The charts below clearly show the relationship between the spread of the dynamic credit range accrual and the volatility of the index.

The impact of the market forward curve
Looking at the evolution of the iTraxx in the last two years and comparing it with the path followed by the iBoxx during the last 7 years, it is easy to note that the iTraxx seems to be near the end of a sustained period of tightening and, therefore, is expected to revert back to a somewhat higher levels in coming years. This is confirmed by the upward sloping forward curve implied by the spot-term structure.

The risk-neutral long-term mean derived from the market forward curve plays a key role in the pricing of a credit range accrual. Here we want to study the impact of different forward curves on the spread value for both static and dynamic range. In particular we are interested in the performance of the static and dynamic credit range accrual in terms of the spread that has to be paid to the investors under the assumption of a flat or a downward sloping forward curve (see charts below).

Calibrating our proprietary model to these two different forward curves we obtained a risk-neutral long-term mean m* equal to 3.4 for the flat and 2.6 for the downward sloping forward curve, while for the market forward curve we compute 4.16 (see relevant table). Notice that these values refer to the natural logarithm of the index spread, which is the process modelled. More details about the model and how it can be calibrated to a given forward curve are available in the appendix below.

The results obtained for both static and dynamic credit range accrual with these new inputs are displayed on this page. The two strategies react differently to changes in the forward curve. As expected the static tends to be more sensitive than the dynamic structure to changes in the market forward curve.

? Static credit range accrual – we can see that the lowest spreads are naturally achieved with a flat forward curve. With a drift that tends to keep the index to its original level, the probability of observing the iTraxx breaching the barriers becomes very low and, therefore, the product can pay only a very small extra spread over Libor. With an upward or downward sloping forward curve the likelihood of observing the iTraxx inside the corridor deteriorates since the drift tends to push the index towards the boundaries of the corridor.

In either the case of an upward or downward sloping forward curve, the iTraxx is expected to significantly move from its actual level over the next years, the corridor must be carefully selected to keep the probability of breaching the barriers low. The set of ranges chosen in the test are clearly derived from the assumption of an upward sloping forward curve and, therefore, the lower bound is relatively close to the current spread level. This explains the relatively high spread paid by the product under the assumption of a downward sloping forward curve.

? Dynamic credit range accrual – we can notice that, although the lowest spread is obtained with a flat forward curve, the spread obtained in all the other cases is not so different from that value. Since for this type of structure the corridor is readjusted typically every six months, the level of the spread does not have as strong an impact on the performance of the product and, therefore, the spread tends to be rather stable for movements in the forward curve. The slightly higher spread under the assumption of a sloping forward curve is due to the fact that between two reset dates the index tends to be pushed towards the upper/lower barrier anyway and this makes clearly increase the probability of breaching the corridor.

The impact of a longer maturity
Here we want to analyse the performance of a static and dynamic credit range accrual with maturities longer than five years. In particular we priced the static and dynamic structure with a maturity of five, seven and 10 years over the usual set of ranges, leaving the remaining parameters set to the 'base case'. The pricing spreads obtained for each configuration are displayed below.

For the static range accrual we can immediately observe that the spread tends to increase as the maturity does while for the dynamic version of the product the spread tends to remain flat. This behaviour was absolutely predictable.

? Static credit range defined at the outset, the longer the maturity the more chance you have to observe the iTraxx trading outside the corridor and therefore the higher the compensation for the investor. Another way to see the same concept: seven or 10-years volatility is clearly higher than 5-years volatility and, therefore the spread paid by the structure must be higher as well.
? Dynamic credit range – With a corridor that is, instead, reset every six months to the current level of the index, the investor is constantly exposed only to a collection of short-term forward volatility that are not affected by the length of the investment period. For this reason, just increasing the maturity, the riskiness of the product tends to remain stable and so the pricing spread paid to the investor.

? In summary – The table below summarizes what we have discussed so far. With a higher index volatility obtained either by increasing the average jump size or the jump arrival frequency the spread paid to the investor tends to be higher for both static and dynamic barriers. With a longer maturity only the static range accrual pays a higher spread while the dynamic structure tends to remain flat. Moreover, when the forward curve moves from a flat orientation for both products the spread tends to be higher although for the dynamic range accrual this difference is only slight.

Conclusion
Credit range accrual is a product that pays a coupon for each day a pre-defined credit reference rate fixes within an agreed range or corridor. The product is principal-guaranteed and, therefore, the investor bears only coupon-specific risks.

The range
The barriers, typically monitored on a daily basis, can be of two types: static or dynamic. In the former the range is defined at the inception in terms of absolute value and remains fixed throughout the life of the product. In the latter, on the contrary, the barriers are defined in relative terms as a percentage of the current spread level and are periodically re-adjusted. A different definition of the corridor clearly leads to different sensitivity of the product to changes in variables like index spread volatility and implied forwards.

Static credit range accrual
We showed that a static definition of the range must be preferred by those investors who believe the CDS index will either remain stable or see limited widening over the investment period. An investor who decides to opt for a static credit range accrual is clearly short long-term volatility since with a higher volatility there is more probability that the iTraxx will be observed outside the specified range.

When spread volatility is less pronounced, however, there can be a higher return for the investor since more coupon is likely to accrue. With this type of structure the investor is also exposed to future spread levels. This is another source of risk that permits the investor to enhance the yield pick-up even further. The upward sloping forward curve implied by the spot term structure expresses the market consensus about higher future index levels and, therefore, those who invest in such a product will receive a higher compensation for expressing a range-bound view.

Dynamic credit range accrual
The dynamic credit range accrual provides the investor with a different risk-return profile. The investor who buys this structure is not exposed to the long-term index volatility but only to the short-term volatility between contiguous resetting dates. Assuming that the corridor is re-adjusted every six months (when the index rolls), we see that investors bear the risk of the six-month volatility and are only slightly exposed to the actual spread level.

With a dynamic specification of the corridor, the performance of this product, while being somewhat insensitive to changes in the market forward curve, strongly depends on the index spread volatility realised through the life of the product. As for the static range accrual, the lower the volatility, the higher the probability for the investor to gain an extra spread over the market benchmark. This type of product must clearly be preferred by those investors that believe in a rise in the spread level but don't foresee an increase in the index volatility.

Maturities
We also analysed the performances of both types of products for maturities longer than five years. In particular we noticed that the static credit range accrual tends to pay a higher premium for longer maturities to compensate the higher probability of observing the index breaching the corridor while the dynamic version of the product tends to return for each maturity the same coupon. Notice that the latter remains a view on six-month volatility while the former becomes a view on a longer-term volatility (in simple terms, five years volatility < 10 years volatility).

Summary
In summary, credit range accrual is a very simple product that offers an alternative way to boost yield by exploiting a view. In a spread-widening environment, where bespoke tranches would suffer a MTM loss, a range note can be a viable alternative especially for those investors that expect spreads to move largely sideways and a dynamic range structure, as presented here, benefits from re-setting barriers periodically.

© 2006 Dresdner Kleinwort. All Rights Reserved. This Research Note was first published by Dresdner Kleinwort on 10 October 2006.