Contents

Double-edged sword

Market turmoil impacts CDPCs for better or worse

Credit Derivative Product Companies (CDPCs) have become the latest structured credit sector to be affected by the current market turmoil. While some newer vehicles are struggling to issue auction rate notes, others are capitalising on opportunities presented by wider spreads and the increasing need for banks to offload exposures, however.

CDPCs fund via auction rate note issuance, which typically mature after the maturity date of the longest-dated swap written by the vehicle. The spread resets periodically, with the coupon payment set by investors' bids via an auction.

"More than one CDPC has found it hard to issue debt in the auction rate market recently, with the newer ones especially affected," notes one structured credit analyst. "But problems with auction resets are an indication of the overall credit environment, not an indication of any failure of CDPCs. While the overall picture is positive for vehicles, current difficulties in issuing debt means that their business plans generally aren't expanding as rapidly as had been hoped."

Investors have been bidding up to the maximum cap for auction rate notes and in some cases the auctions have failed, at which point sponsor banks have taken up the auction. A number of CDPCs have consequently begun to broaden their funding strategies by relying more on MTN and short-term issuance structures.

Such difficulties in funding have meant that some CDPCs have left the pipeline, but others are still moving forward with their plans. As the CDPC capital model is a core part of the rating process and the inability to issue debt at unfavourable levels is stressed within this, the situation is unlikely to impact vehicles' ratings.

Indeed, with the exception of Irish firm Structured Credit Holdings – which faced difficulties when the collateral it posted with its counterparties was subject to onerous repricing – CDPCs have otherwise largely been unaffected by the credit crunch because they hold synthetic exposures and aren't subject to market value triggers. With banks needing to provide more protection on their credit portfolio, reduce the market value volatility of their credit portfolios and spreads having widened to such an extent, there are opportunities for further vehicles to enter the market.

As Douglas Long, evp – business strategy at Principia Partners in London, explains: "The current market environment has underscored why the CDPC business model works: providing vehicles can attract the initial capital investment and term funding, they can now lock in wider spreads. Previously they were earning 20bp for the protection they provide; now they are earning 60bp."

The latest CDPC to be assigned Aaa counterparty ratings is Quadrant Structured Credit Products, sponsored by Magnetar MQ and Lehman Brothers (see SCI issue 55). The vehicle will provide credit protection on both single name and tranches of corporate reference obligations, having issued US$300m Aaa and US$100m Aa2 rated deferrable interest auction rate notes. At least two further CDPCs are believed to have entered the pipeline, including Pallium Investment Management sponsored by BMO Financial.

CS

Japanese investors look to correlation

Simpler structured products on the rise

Japanese investor concerns arising directly from the US sub-prime crisis have led to a preference for products exhibiting stable correlation and modest deltas. The theory being that such products make it easier to extrapolate marks and underwriters will provide more liquidity for the paper they're selling.

"An increasing number of investors have begun using valuation models internally and checking deltas periodically," confirms Takahiro Tazaki, director, head of structured credit research at Barclays Capital in Tokyo. "Market disruption has resulted in deep discounts for a broad range of both synthetic and cash products in the primary market – although there is currently better activity in secondaries."

While equity investors – mainly pension funds and alternative investors – remain active in the cash sector, the volume of first-to-default baskets and simple structured products being sold to small and medium-sized investors has grown significantly since the late summer. "Japanese accounts became typically more comfortable with less complicated products such as first-to-default – they've shied away from ABS CDOs and other such complicated correlation products," adds Tazaki.

These portfolios typically comprise liquid foreign and domestic financial names. In particular, some CDS participants are looking at Japanese consumer finance names, such as Promise and Aiful.

"Because of the civil rehabilitation filing of a mid-class consumer finance company, for the first time, the credit curve of such consumer finance firms became inverted. Correlation traders and hedge funds seem to be interested in hedging these names, even if they are more expensive," continues Tazaki.

And following the latest central bank liquidity injections, one Japanese mega-bank has returned to the market with a synthetic balance sheet CLO. The timing of the deal is thought to have been rescheduled due to the sub-prime crisis, with end-September judged to be a reasonable environment to get it away in – particularly as it was important to do so before the half-year accounting period ended.

Mizuho Corporate Bank's CuBic One Series 2007-1 transaction was rated by Moody's and comprised Y25.3bn CLNs (four classes rated Aaa to Ba2) and a Y248.4bn super senior CDS. Arranged by Mizuho Securities, the deal references a Y279.9bn portfolio of loans lent by Mizuho (and its wholly-owned overseas subsidiaries) to 128 non-Japanese corporates.

Domestic investors continue to be in a difficult situation in terms of asset allocation, however; the recent increase in 10-year JGB yields wasn't enough to solve their long-term investment dilemma. But, given that the leveraged loan sector is now picking up, CLOs may become more attractive for large Japanese banks if they can isolate residential risk from corporate risk.

Observers agree that current spread levels in Japan are fully priced, indicating that the impact of the sub-prime issue on the country's investor base is passing its peak.

CS

Distressed opportunities

Fund proliferation limited by lack of liquidity

Specialist funds set up to purchase distressed ABS CDO and leveraged loan assets are gaining an increasingly high market profile. However, it is likely that only a handful will be successful - not least because the prices being offered for such paper are too low for many sellers to stomach.

"These so-called vulture funds appear every five or six years and are part-and-parcel of the credit cycle. But it is unclear how much paper is actually being sold into these funds at this stage - it is difficult to find enough assets at the right price to make the arbitrage work," comments one structured credit strategist.

The funds investing in ABS CDO paper have been established by mortgage credit managers and are typically looking at mezzanine sub-prime tranches - the assets that have shown the greatest volatility over the last few months. The buyers are characterised by having significant trading experience, which is necessary in order to source the assets and manage them in the case of default.

"These funds are at the cutting edge - they are sophisticated investors who understand the underlying very well - and will continue buying assets for as long as the repricing in the market continues; certainly until the end of the year and possibly into 2008. But the longer the repricing goes on for, the harder it will be to find paper. The funds need to be already established to best capitalise on these opportunities," continues the strategist.

The sellers of such assets have been hit by mark-to-market losses or margin calls and are liquidating their portfolios; for example, hedge funds that are suffering from redemptions. "It is a good opportunity for those with deep pockets, but it is difficult to predict over the next few months how the market will turn out so they have to be willing to buy potentially at the bottom of the market," adds one asset manager.

Meanwhile, the sellers of the leveraged loan assets are typically LBO-arranging banks looking to free up warehousing lines. "Investment banks need to either sell to distressed funds or distribute paper via CLOs, which is a tough sell in the current environment. Because spreads have moved wider since the loans were originated, banks have been forced to sell these assets at a discount," the asset manager remarks. But some banks have also begun to tie in total return swap financing for the portfolios they sell, thereby reducing the funding costs for the acquirer.

The assets being sold in the leveraged loan space typically have high leverage multiples and are covenant-lite structures. The buyers for such paper are believed to be funds established by private equity houses.

CS

SIVs hit again

Two vehicles downgraded

Moody's and Fitch have moved to downgrade the debt issuance of two more SIVs in the past week. Meanwhile, a new research report offers a bleak forecast for the model.

Moody's has downgraded the ratings assigned to the MTN and CP programmes of Cheyne Finance from Aaa to Ba3 and Prime-1 to not Prime respectively. The agency downgraded the vehicle's mezzanine and combination capital notes on 5 September, and placed the ratings of the above senior debt programmes on review for possible downgrade.

The agency says that due to the current distressed market environment, Cheyne Finance has had to liquidate assets to repay maturing commercial paper and medium term notes. The deterioration in market value of Cheyne Finance's asset portfolio and the impact of crystallised losses on the rated notes caused Cheyne Finance to go into receivership on 5 September 2007.

This rating action reflects further deterioration in the market value of Cheyne Finance's portfolio, which has caused the net asset value of capital to drop significantly. The long-term rating assigned to the medium term note programmes is on review with direction uncertain to reflect certain restructuring proposals that are currently under consideration by the vehicle's receiver. Moody's review will focus both on the restructuring proposals and on the evolution of the market prices of Cheyne Finance's assets.

The agency notes that this rating action takes into account the current stressful market conditions. While the underlying assets of Cheyne Finance remain highly rated, the unprecedented illiquidity in the market for asset-backed securities has created a high level of uncertainty around the valuation of the assets, causing Moody's to consider a higher price volatility when determining the value of available capital.

At the same time, Fitch has downgraded Axon Financial Funding's MTNs, CP and mezzanine notes from triple-A to single-A, F1+ to F1 and double-B minus to triple-C respectively. Approximately US$2bn US CP, US$738m Euro CP, US$6.3bn US MTNs, US$25m Euro MTNs and US$890m of mezzanine notes are affected.

The rating actions reflect the fact that Axon Financial has not yet been able to secure any alternative sources of funding and has now entered a restricted funding operating state. Consequently, a sale of assets or committed liquidity draw will be required to meet maturing liabilities, which could result in further realised losses.

While in the restricted funding operating state, the SIV is not permitted to issue further senior notes and will have to sell assets or use committed liquidity to redeem maturing funding. Axon Financial has used most of the available excess committed liquidity, but currently holds sufficient committed liquidity to meet its liquidity test requirements.

Axon's portfolio currently comprises 31% RMBS (7% prime, 18% near-prime, 1% sub-prime and 5% closed-end seconds), 21% monoline-wrapped securities, 19% CDOs, 16% CMBS, 9% cash equivalents and 4% other ABS. The portfolio has a geographic exposure of 92% to the US and 8% to the UK. Currently, 96.5% of Axon Financial's portfolio is rated triple-A equivalent, 3.2% double-A equivalent, and 0.3% single-A equivalent.

Fitch notes the very high credit quality of the portfolio assets but, of late, some of the assets have experienced downgrades and their market values have come under extreme pressure.

That pressure has been widespread throughout the SIV sector, according to a new report by Deutsche Bank analysts. The report in part examines the various methods of re-funding the vehicles have tried to utilise.

"Replacing CP funding with the use of liquidity or repo facilities clearly provides only temporary respite, assuming the short-term financing market for SIVs does not fully recover," the report warns. However, some of the recapitalisation measures used appear to be more meaningful solutions to the asset-liability valuation/maturity mismatch and squeeze in ABCP funding.

Even so, the Deutsche report concludes: "Such measures are designed primarily to finance an orderly "roll-off" of assets rather than re-establishing the SIV business model. Lest we see a full normalisation of asset and funding markets therefore, the current generation of conventional SIVs managing legacy, pre-crisis assets are very unlikely to be incremental investors going forward, in our view."

MP

Value participation

Hans-Michael Schania, CDO management, Alex Tsymbal, US ABS and SIV management, and Alex Rothlin, ABS management, at Aurelius Capital answer SCI's questions

Q: When, how and why did your firm become involved in the structured credit markets?
A: Aurelius Capital is 100% owned by its employees and was founded in November 2006 after 18 months of negotiation with an institutional investor (a financing arm of an Austrian province that manages all the financing for the State of Lower Austria), which wanted to outsource expertise in managing structured finance assets. We were mandated to manage a CDO equity fund on its behalf. With the strong institutional sponsor at our back, we wanted to capitalise on the long-term experience and track record in origination, structuring and managing structured finance assets.

Q: In your view, what has been the most significant development in the credit markets in recent years?
A: Global spreads have tightened over the last several years, which attracted a whole swathe of new investors to structured credit. The amount of money consequently available to place forced banks to produce new investment vehicles: the increased demand was met by increases in supply. However, not everyone who participated in the market understood the risks – and those who made the wrong assumptions, particularly about the liquidity of the asset class, suffered the consequences.

Q: How has this affected your business?
A: The resulting credit crunch has slowed our growth plans, although we are in line with our initial business plan. To go forward we need to educate our customers about what was happening. Although we expected problems to appear in the sub-prime sector and so avoided participating in sub-prime RMBS and mezzanine ABS CDOs, no-one expected the situation to become as severe as it has done. But the result is that the long talked-about tiering between asset managers is actually emerging now: it is much easier to distinguish good managers from mediocre and bad managers.

Q: What are your key areas of focus today?
A: We closed a US$300m CDO-squared called Aurelius Capital CDO 2007-1 on 10 September, having priced it successfully at the beginning of August. It is predominantly backed by CLOs, as well as a small percentage of ABS CDOs.

We decided to invest in better quality assets at the same spread target rather than increase the return on equity, even though we own the majority of it. Thanks to strong support from our clients and sponsor, it closed successfully.

We like the leveraged loan asset class and the CDO-squared structure, and we always participate in the equity piece of our own transactions besides investments in third-party managers. But we're looking to diversify our issuance too; for example, we're currently analysing the commercial real estate sector. We need to be able to adjust to a changing marketplace: right now there is an opportunity to invest in high quality assets at wide spread levels via ABS funds.

In general we target sophisticated institutional investors where some of them have typically never participated in ABS before but have analysed the market for a while and believe that the rewards of investing in the asset class compensate for the risks involved.

Q: What is your strategy going forward?
A: We have taken a conservative approach to our business plan and envisage launching one CDO a year. Besides that we've got another two to three projects in the pipeline which we expect to launch over the next three to four months.

One of those would be the execution of an ABS fund with no leverage, so it's not subject to margin calls. The idea is to establish the type of fund that should have been sold to investors all along.

The current crisis has demonstrated that ABS isn't really a money market instrument – it becomes illiquid as soon as the market turns. We are trying to educate our investors about the fact that no structured credit product will ever be ultra liquid because they are ultimately over-the-counter investments.

In general, the funds we are working on are characterised by reasonable lock-up periods: we are looking to set those between two to three years, with quarterly redemptions after that. But it really depends on the investment; for example, there is a three-year commitment for a CDO equity fund we're planning and a two-year commitment for the ABS fund.

The more generic the fund, the less commitment we'd expect from investors. As long as we educate them properly, it becomes logical to have lock-up periods because the value participation occurs over the long term. The objective is to liquidate the funds when they have sufficient value to give a good return for investors.

Q: What major developments do you need/expect from the market in the future?
A: We expect the market to now begin asking banks to add value – linked to both credit and liquidity considerations. Investors are unlikely to get fair valuation levels on equity pieces that are purchased even recently. The banks need to begin providing investors with a way of understanding how much an asset is really worth, both from credit quality and liquidity perspectives.

How they do this depends on the investor's horizon: if they need daily valuations (as in market value vehicles), then traders need to provide daily bids; if they are a buy-and-hold investor (such as insurance companies and long-term funds), then they should receive timely information about any deterioration in the underlying assets which might impair their investment.

Additionally, such arrangements need to be agreed in advance, so that investors are aware of what they are actually getting when banks talk about valuations and what those valuations imply. Unfortunately, there is no evidence of this happening as yet, so it is up to investors to ask for it.

About Aurelius Capital
Aurelius Capital is a Vienna-based asset manager specialising in credit and structured products. The firm's main products and services are managing and structuring CDOs, ABS and long-term structured product investment vehicles.

The team of Aurelius Capital has an average credit experience of more than 10 years, with members having joined from well-known investment banks and hedge funds based in London and New York. The management team brings a long and stable track record through different credit cycles and has managed portfolios throughout difficult market environments.

Bank expands securitisation unit

The latest company and people moves

Bank expands securitisation unit
nabCapital, the institutional banking and capital markets division of National Australia Bank, announces that Chris Leslie and Sandie Staimesse have joined its London-based Securitisation Group, with primary responsibilities for securities arbitrage sourcing and execution. Both roles report to Tony Gioulis, head of securitisation UK/Europe.

Leslie joins nabCapital as a director. Latterly with Gordian Knot, he has more than seven years of structured credit experience, with a main focus on senior CDO investments and commercial mortgage-backed securities.

Staimesse joins nabCapital as an associate director. Most recently with WAMCO, prior to that she worked at Citigroup Asset Management.

Gioulis comments: "Chris and Sandie are joining us at a pivotal point in the evolution of the asset-backed market. Their extensive sector experience will enable us to capitalise on current market opportunities, ultimately enhancing our established position as a major securitisation player in the European arena."

Index trading head quits
Ian Gladen has left his role as head of credit index trading at Bank of Montreal. His destination is not yet known.

Manager adds in-house analyst
Former Lehman Brothers ABS CDO analyst Michael Koss has joined Scottwood Capital Management as a senior investment analyst. He will cover the asset-backed securities sector in a new role at the firm, which in light of current of current market conditions decided to bring such expertise in-house on a full-time basis.

PIMCO hires prop trader
PIMCO has appointed Luke Spajic as head of European credit strategy and senior portfolio manager. He will be based in London and report to Scott Mather, md and head of portfolio management Europe at PIMCO.

Spajic is joining PIMCO from Goldman Sachs, where he was an executive director and a proprietary trader specialising in macro trading across a broad array of global markets including credit and emerging markets. Prior to that role, he was a vp at Highbridge Capital where he focussed on trading non-US credit.

Babson promotes in loans
Babson Capital Management has announced that Russell Morrison and Marcus Sowell, who are both mds with the firm, have been named co-heads of the US bank loan team. They succeed Thomas Finke, who last month was named president of Babson Capital. Morrison and Sowell will continue to report to Finke.

Morrison and Sowell joined the Babson bank loan team's predecessor, First Union Institutional Debt Management, in the spring of 2000. Babson Capital acquired the business unit from First Union/Wachovia Corporation in June 2002.

Merrill heads exit
Head of European credit, Dale Lattanzio, and head of fixed income trading, Osman Semerci, have both left Merrill Lynch. Market speculation continues that there will be more senior credit exits in the near term.

Such speculation is understandable, given Merrill's announcement of write-downs of an estimated US$4.5bn, net of hedges. This, the bank says, was related to incremental third-quarter market impact on the value of CDOs and sub-prime mortgages.

"These valuation adjustments reflect in part significant dislocations in the highest-rated tranches of these securities which were affected by an unprecedented move in credit spreads and a lack of market liquidity in these securities, which intensified during the third quarter. During the quarter, the company significantly reduced its overall exposure to these asset classes," Merrill says.

MP

BBA paper sets out credit crunch lessons

A round up of this week's structured credit news

BBA paper sets out credit crunch lessons
A worldwide targeted review of financial regulation will help the markets weather the credit crunch, according to a paper published by the British Bankers' Association (BBA).

'The Credit Crunch: Implications and Changes Required' is a position paper prepared for BBA members which calls for immediate reviews of regulation in the UK and internationally. It covers the regulatory treatment of:

• financial firms' cashflow needs and risk practices (so-called liquidity risk management);
• complex credit products and off-balance sheet vehicles (such as conduits and SIVs);
• accounting and valuation practices for complex financial products;
• credit rating agencies; and
• deposit protection arrangements.

BBA chief executive Angela Knight comments: "However uncomfortable it may be, we must tackle the question we are increasingly hearing from the international community about the credit crunch and the Northern Rock affair: 'Did it really have to be done like this?' No single factor was responsible for the customer panic that we saw last month. But the architects of the UK system of tripartite regulation have questions to answer and the international financial community has clear lessons it can learn from the mistakes that were made. This paper is a contribution to both of these processes."

ECB survey released
The European Central Bank has released its quarterly Euro Area Bank Lending Survey. The report notes that lending standards have tightened considerably for both businesses and households over the quarter, and that this is expected to continue for the remaining quarter of the year.

However, the findings indicate that the turmoil in the credit markets is only partly to blame for the tightening in lending practices. A significant majority of banks believe that the credit crunch has had no impact on credit standards, with only corporate lending deemed to have been affected, with almost four in 10 banks reporting that the credit crunch resulted in more conservative lending.

Regarding SME lending and lending for house purchases, only 21% and 10% of banks surveyed reported that the credit crisis has directly led to a tightening of lending standards. Analysts at Deutsche Bank note that these relatively low figures are likely related to the fact that most banks surveyed are not overly exposed to wholesale funding, given retail and other stable sources of financing.

Other factors were deemed to have played a more influential role in the tighter credit standards for mortgage borrowers, such as an easing of competitive pressures and expectations of a weaker housing market. The survey also indicated that margins on riskier mortgages have been raised, amid a weakening in demand for loans. Notably, 80% of euro-zone banks surveyed believe the turmoil has hampered their ability to use the securitisation market in funding mortgage loans.

BIS reports on change in the financial system
The financial system has been going through a historical phase of major structural change over the past three decades. A new paper released by the Bank for International Settlements traces the implications of this financial revolution for the dynamics of financial distress and for policy.

The report argues that, despite this revolution, some fundamental characteristics of the financial system have not changed and that these hold the key to the dynamics of financial instability. These characteristics relate to imperfect information in financial contracts, to risk perceptions and incentives, and to powerful feedback mechanisms operating both within the financial system and between that system and the macro-economy.

As a result, the primary cause of financial instability has always been, and will continue to be, overextension in risk-taking and balance-sheets. The challenge is to design a policy response that is firmly anchored to the more enduring features of financial instability while at the same time tailoring it to the evolving financial system. Using an analogy with road safety, policy has so far largely focused quite effectively on improving the state of the roads and on introducing buffers. More attention, however, could usefully be devoted to the design and implementation of speed limits.

Fitch downgrades KKR SLNs
Fitch Ratings has downgraded the short-term rating of the secured liquidity notes (SLNs) issued by KKR Pacific Funding Trust (KKR Pacific) to single-D. The downgrade reflects KKR Pacific's non-payment of the SLNs in accordance with the original terms of the notes.

Beginning on August 20 2007, parties to the programme executed a series of standstill agreements. The standstill agreements waived the overcollateralisation (OC) test, suspended the requirement for collateral liquidation, pushed back the expected maturity dates and the final maturity dates on all SLNs, and waived the obligations of KKR Financial and the affiliated depositor corporation under the repurchase agreements.

Fitch views the existing standstill agreements and the proposed restructuring of the existing SLNs as a distressed debt exchange under its criteria. The rating withdrawal reflects the effective replacement of the existing notes with amended notes with differing terms.

On August 15 2007 Fitch downgraded the SLN short-term ratings of KKR Pacific to single-B from F1+ and placed the ratings on rating watch evolving. On August 27 Fitch revised the rating watch to negative.

October 2 2007 would have been when the first SLN reached its final maturity date based on its original terms. The single-D rating reflects the non-payment of the SLNs based on the original terms.

KKR has indicated that a programme default, as defined in the documents, has been avoided through the standstill agreements and the potential restructuring. However, under its criteria, Fitch views the standstill and potential restructuring as a distressed debt exchange. Fitch's short-term ratings reflect the likelihood of full and timely repayment of an obligation in accordance with the original terms and does not incorporate the level of recovery prospects.

KKR Pacific is a special purpose, bankruptcy-remote Delaware limited liability company established to issue SLNs and use the proceeds to enter into repurchase agreements with KKR Financial Corporation or the affiliated depositor corporation. The repurchase agreements are collateralised by triple-A rated RMBS.

As of October 1 2007, KKR Pacific had approximately US$2.56bn of outstanding principal balance of SLNs. The asset balance reported (used for the purpose of the OC test) was approximately US$2.58bn.

SGAM launches latest CFO
Moody's has assigned the following provisional ratings to the notes expected to be issued on 27 November 2007 by Premium CFO II: Aaa to the €80m Class A notes due 2014; Aa2 to the €26m Class Bs; A2 to the €16m Class Cs; Baa2 to the €14m Class Ds; and Ba2 to the €14m Class Es.

This transaction is a collateralised fund obligation (CFO) that is backed by interests in a diversified pool of hedge funds. The transaction is arranged by SGAM Banque and managed by SGAM Alternative Investments.

According to Moody's, the ratings are primarily based on: the diversification of the underlying portfolio of hedge funds (individual and strategy concentrations); its estimation of the net asset value volatility for each strategy; the transaction's structural and legal protections; and the expertise of SGAM as manager of funds of hedge funds.

The issuer will invest the proceeds of the issuance in units of SGAM's Unlevered Premium Fund, which invests in a diversified portfolio of hedge funds, each following its own strategy.

The transaction's structure includes a frequent mark-to-market process for the collateral pool and a set of OC ratios that constitute sufficient credit enhancement for the transaction. Certain diversification criteria and portfolio limitations must also be maintained.

TriOptima completes first ABX index termination cycle
TriOptima, the post-trade service provider for OTC derivatives, has completed the first multilateral ABX index termination cycle. Ten dealers tore up over 8,000 transactions, which included trades linked to sub-prime mortgages, through the triReduce service.

TriOptima says that strong demand for a triReduce termination cycle was driven by the monthly payment cycles, frequent shortfall adjustments (especially in the wake of recent volatility in the sub-prime market), and the 30-year plus maturities associated with these transactions. Monthly payment cycles increase operational risk and cost exposures since the back office has to service the trades each month rather than quarterly as with other CDS trades, while 30-year maturities extend these risks over a long period of time.

Raf Pritchard, head of the firm's service management in New York, comments: "Dealers took this opportunity to net down their outstanding trades in advance of any potential events affecting notionals, such as factoring or rating downgrades. The benefits offered by our termination service - reduced credit and operational risks and costs - were particularly helpful to our clients, given the recent period of heavy market activity."

Starts super senior downgraded
Moody's has downgraded the Aaa rating of Starts (Cayman) Series 2006-1 US$60m leveraged super senior CLNs to Aa1. The transaction incorporates a trigger event that looks to the total loss of the underlying reference portfolio of corporate assets. If a trigger event occurs, investors may decide to incur the mark-to-market loss of the note up to the initial investment or to increase the size of their investment, which will reduce the leverage in the deal.

The rating action reflects the deterioration in the credit quality of the transaction's reference portfolio, consisting of corporate assets. There have been no losses associated with the portfolio to date. The rating addresses the expected loss to the investors relative to their initial investment and is based on an analysis of the credit risks in the transaction, as well as the legal structure.

Permacap news
Washington Square Investment Management's Carador reported that at the close of business on 31 August 2007, the unaudited net asset value per share was €0.8793. The vehicle's NAV decreased by 3.44% in August. This month's calculations include an estimated €577,364.28 worth of net cash flow interest received in the month (to be allocated between capital and income), which equates to €0.0115 per share.

The manager referred 13 investments to the pricing committee which it considered were incorrectly priced by market counterparties for the end-August net asset value calculation. The pricing committee proposed that, for these 13 investments, an arithmetic average between the price provided by market counterparties and the valuation obtained using Washington Square's internal models should be the preferred alternative to the market counterparty valuation. This proposal was accepted by the board on the 4 October 2007.

The following is the extract from the manager's monthly report to 31 August 2007: "Despite good corporate results and lack of defaults, technical factors continue to dominate CLOs. The volume of outstanding CP fell by 11.9% from July to August (the previous largest monthly contraction was -6.2% in September 1970). The lack of demand for triple-A structured credit paper has resulted in a significant slowdown in the new issue market for CLOs.

"From Carador's perspective, the situation is advantageous as the fund ramped up its CLO investment in 2006 when the cost of funding was at its lowest. The lack of liquidity in the CP market has pushed CLO triple-A spreads wider, making the arbitrage for new CLOs less attractive. Since the Fed announced its rate cut, credit markets have further improved and the new issue market has been active. Flight to quality seems to have ended and prices for loans are edging up."

The company notes that the rating agencies' approach to CLOs is actually conservative as opposed to the aggressive assumptions used in CDOs of ABS. As an example, the recovery rate used by S&P on second lien loans in the context of CLOs is either 48% or 22% (48% for the first 15% of the CDO portfolio and 22% for the rest). Carador tends to be even more detailed in assigning recovery rates based on its knowledge of each individual issuer.

Meanwhile, Queen's Walk Investment Ltd has announced that at its extraordinary general meeting, shareholders approved the special resolution authorising the tender offer as defined in the company's circular to shareholders dated 13 September 2007.

The special resolution approving the tender offer was passed on a show of hands by those present at the meeting. Had a poll been held, the result would have been: 20,771,170 for; 1,681,588 against.

The total number of ordinary shares voted was 22,452,758, representing 56.66% of the share capital of the company in issue as at the time of the meeting. Accordingly, the tender offer has become unconditional and the company will repurchase 4,504,500 ordinary shares under the tender offer. The company expects that unsold ordinary shares will be released by the escrow agent (for uncertificated holders) or their share certificates will be returned (for certificated holders) by 15 October 2007.

Interwoven connects with DTCC
Interwoven has launched Interwoven DealConnect for DTCC, the capital market's first standalone solution to enable buy-side and broker dealers to connect with the Depository Trust & Clearing Corporation (DTCC) Deriv/SERV Trade Information Warehouse (TIW). Interwoven DealConnect for DTCC offers flexible connectivity to the TIW no matter what systems are employed in-house, and advances the industry's efforts to increase post-trade efficiency in the OTC derivatives market.

Interwoven DealConnect for DTCC provides firms with access to Deriv/SERV's full range of post-trade processing tools, including TIW integration, infrastructure to support processes such as payments, notional adjustments, trade matching and workflow updates for a variety of OTC instruments, with minimum implementation required. The Interwoven DealConnect solution for DTCC is built on a service-oriented-architecture and integrates with any existing in-house system and offers the flexibility to construct and send trade messages from any format.

SuperDerivatives new release
SuperDerivatives has launched a new release of its online credit derivatives platform, SD-CD. The new version represents a substantial increase in the level of functionality, price transparency and analytic power compared to that offered by currently available solutions - addressing challenges posed by the recent dramatic developments in structured credit markets.

The enhanced version introduces a unique dynamic pricing model for bespoke synthetic tranches, facilitating efficient and transparent hedging and risk management of CDOs. The SD-CD platform provides intraday pricing relying on tradable two-way data from multiple sources (in contrast with competitive products that use previous-day consensus data, which does not reflect actual tradable prices) and portfolio utilities which support name-specific default and market risk metrics.

Kamakura upgrades
Kamakura Corporation has released version 2.0 of its Kamakura Risk Information Services KRIS-CDO web-based collateralised debt valuation tool. In addition, the firm is making available a new research report entitled 'Implications of Alternative CDO and Credit Portfolio Modeling Techniques'.

KRIS-CDO, officially launched in April 2007, allows users of the KRIS default probability service to analyse synthetic CDOs under a wide range of valuation techniques. Valuations run on a series of servers, including an 8-processor IBM blade server, based on a secure facility shared with the US Department of Defense and several major multinational corporations.

The upgrade to KRIS-CDO version 2.0 includes an unlimited number of scenarios with a series of default choices from 100 to 500,000 scenarios per run. KRIS-CDO version 2.0 displays concentrations by industry and rating, and it allows notional principal to vary among the reference names in the portfolio.

CS

Trading ideas - last buying opportunity?

John Hunt, research analyst at Credit Derivatives Research, looks at a negative basis trade referencing Sallie Mae

As the posturing and arguing over Sallie Mae's pending-but-in-doubt LBO may be coming to head, we recommend a negative basis trade on the Sallie Mae 4.75s of March 2014. Like all negative basis trades, this trade fundamentally seeks to take advantage of mispricing of the bond relative to the price implied by the CDS market. In addition, this trade stands to benefit from certain scenarios that are specific to Sallie Mae and its currently uncertain LBO.

First, if the LBO goes off as planned, without protection to bondholders, we would expect Sallie Mae's credit spreads to widen out quickly upon announcement that the transaction is going through. That scenario tends to benefit negative-basis positions because, under those circumstances, CDS spreads tend to widen before bond spreads do – just take a look at how Sallie's bond-CDS basis spiked briefly in April when the deal was first announced.

We would expect to exit the trade with a quick profit shortly after such an announcement. Even if we didn't see this basis spike, we still would expect an eventual profit on the trade based on bond-CDS convergence. The "credit deteriorates" scenario of Exhibit 4 illustrates our estimate of the performance in the no-spike, deal-consummated case.

Second, there is a possibility that the LBO would be abandoned – perhaps with the payment of a breakup fee (set at US$900m under the contract) to Sallie. Given that the firm's credit spreads are still far wider than they were before announcement of the LBO, we would expect spreads to tighten considerably in this case – especially with US$900m in free money coming in the door. The "credit improves" scenario of Exhibit 4 illustrates performance under this scenario.

In this case, we would expect the trade to perform like a plain-vanilla basis trade, with its performance driven by convergence between bonds and CDS. Because we construct the trade to be as close as possible to credit-spread-neutral at a six-month time horizon, the simple fact that credit spreads improve should not affect performance too much.

Finally, there is a possibility that the deal will go through after the parties renegotiate it to include a tender offer for Sallie's bonds – a feature that did not appear in the original structure, according to Gimme Credit's Financials expert, Kathleen Shanley. If the bonds are taken out at or around par, we would expect a very significant profit on the bond leg of the trade, as the bond here is trading in the mid-80s currently. We would also expect a profit on the CDS leg of the trade because we'd expect Sallie's credit to deteriorate.

The firm's credit spreads have tightened significantly since consummation of the LBO came into question, indicating that current levels reflect some consensus likelihood of a "good" outcome (deal falls apart, spreads tighten) and a "bad" outcome (deal is consummated, spreads widen out again).

Although we've heard the term "orphan CDS", apparently referring to a swap that has no deliverable obligations and is therefore worthless, we call investors' attention to Section 2.30 of the 2003 ISDA Credit Derivatives Definitions. That seems to provide for the designation of new reference obligations upon tender of the old ones, and to render newly issued obligations deliverable if they are not subordinated to the new reference obligations.

While we are not rendering any legal opinions here, a lay reader could conclude that any new bonds Sallie issues in conjunction with the transaction would be deliverable obligations under the CDS contract, and that CDS would be "orphaned" only if no new bonds were issued. We don't illustrate this scenario in Exhibit 4 because bond and CDS performance is decoupled almost by definition in this case, but as explained, this could be the best outcome of all for the trade.

We don't take a position on how likely it is that the LBO will go through – and prefer not to delve into the intricacies of the "Material Adverse Event" clause in Sallie Mae's contract and whether the clause was or was not triggered by recent credit market changes or proposals in Congress to cut student loan subsidies. We do see that legal commentators have noted that courts have set a fairly high hurdle for buyers attempting to walk away in analogous cases – including a requirement that any claims-adverse event create a sustained negative effect on the jilted target's business.

To the extent we tilt slightly toward believing that the deal is more likely than not to be consummated, this trade is consistent with the steepener we recommended on Sallie Mae on 28 September (see SCI issue 58). Because the CDS is longer-dated than the bond, the position we recommend here will benefit from any steepening of Sallie Mae's credit curve.

Basis trade basics
The underlying idea of the negative basis trade is that credit risk is overpriced in the bond market relative to the CDS market. The investor buys a risky bond – and thus is paid to take credit risk on the issuer – while paying for credit risk in the CDS market by buying protection on the issuer.

Eventually, the prices for credit risk in the two markets should converge, resulting in an arbitrage-like profit. In the interim, the investor should earn positive carry because the credit spread that is collected in the credit market is greater than the spread that is paid in the CDS market.

The credit risk implied by market prices in the CDS market can be inferred in a fairly straightforward fashion from the CDS spread curve, as discussed in the Trading Techniques section of the CDR Website. The price of credit risk in the bond market is sometimes measured by the "z-spread" – the amount that must be added to the risk-free curve to cause the discounted present value of the bond's promised cashflows to equal the bond's current price.

A "quick and dirty" comparison of how the bond and CDS markets are pricing a bond's credit risk can be done by simply looking at a bond's z-spread and the CDS spread for an instrument of matching maturity. In fact, the "basis" is defined for our purposes as the CDS spread minus the bond z-spread.

While the z-spread often is a good measure of how the market prices a bond's credit risk, the z-spread approach implicitly makes the incorrect assumption that all the bond's promised cashflows definitely will be received, and received on time. The z-spread is not a good measure of credit risk price when bonds are trading at or near distressed levels and/or are trading away from par.

For that reason, we prefer to extract the probabilities of default implicit in CDS spreads and use those probabilities to arrive at a CDS-implied bond price. If the bond trades in the market below the CDS-implied price, we say it is "cheap" compared to its CDS-implied market value. If the bond is trading above CDS-implied price, we say it is "rich."

Because bond and CDS maturities usually do not match exactly, interpolation is usually required to perform the comparison. The Trading Techniques section of the CDR website gives details of this computation. We then compare the market price of the bond to the CDS-implied bond price to determine whether the bond is trading rich or cheap to the CDS-implied level.

Constructing the negative-basis trade position
We generally construct negative basis trades so that they are default-neutral – that is, so that the CDS fully hedges against bond losses in the event of default. This ensures that we are selling the same amount of default risk in the bond market as we are buying in the CDS market.

In doing so, we take account of the fact that the CDS payoff in default is defined relative to the notional amount, while the loss on a bond in default depends on its pre-default price. Thus, a CDS and a bond with the same notional amount will pay off different amounts in default if the bond is trading away from par.

For example, consider a bond and a CDS each with US$100 notional. If the bond is trading at US$110 and recovery value in default is US$40, the bondholder will lose US$70 in default and the CDS protection buyer will gain only US$60, so the investor will have to buy more than US$100 in CDS protection to hedge against default. If the bond is trading at US$90, then the bondholder will lose only US$50 and the CDS protection buyer will gain the same US$60, so less than US$100 notional of CDS protection would hedge against default. Given that bonds pull to par over their life if they do not default, we typically construct our positions so that the CDS hedges a bond price halfway between the current market price and par – although we may adjust the hedge amount upward or downward if we maintain a bearish or bullish fundamental view on the credit.

Because we do not expect to hold the trade for a long period of time and because there is a relatively high likelihood of a large move in one direction or the other in Sallie Mae's case, we select the CDS weight relative to the bond so that the trade is as close as possible to credit-spread-neutral across a range of scenarios based on a six-month time horizon.

To do that, we selected the CDS weight relative to the bond that minimised the effect of absolute spread movements across a range of scenarios. This weight is quite close to the DV01-neutral weight, so the trade should gain or lose little from small parallel credit curve movements in the short term.

Another consideration in constructing the position is that a liquid CDS usually will not be available in an interpolated tenor that matches the maturity of the corporate bond. We construct the trade using a CDS of the closest available liquid tenor.

Because CDS premiums vary with tenor (generally, longer-dated CDS have higher spreads), the mismatch between CDS and bond maturity will affect trade characteristics such as carry and rolldown. We generally recommend only trades that have positive carry.

We also present a simple duration-matched position in a government bond to hedge interest rate risk. This risk could also be hedged with an interest rate swap, but we understand that government bonds are the most useful hedging instrument for most investors. We understand that most investors will prefer to hedge interest-rate risk at the portfolio level in any event.

Our strategy for both the CDS and the Treasury hedge is strongly influenced by a desire for simplicity. Perfect hedging would require adjustment of the CDS and Treasury positions over the life of the trade. We are happy to discuss such strategies with clients, and we also provide a set of sensitivities to help clients implement more sophisticated hedging strategies.

Trade specifics

Bond cheapness
Based on our valuation approach, the SLMA 4.75 of March 2014 bond is trading cheap to fair value. Exhibit 1 presents the price-based term structure of the company's bonds with maturities of two to twelve years and indicates that the March 2014 bond is trading cheap to fair value, and appears cheaper to fair value than most of the issuer's other bonds.

Exhibit 1

Exhibit 2 compares the bond z-spreads with the CDS term structure and fair bond z-spreads, and shows that the recommended bond's market z-spread is wide of the closest-maturity CDS and its fair z-spread.

Exhibit 2

Exhibit 3 illustrates the basis between the matching maturity on-the-run CDS and the bond over the past year. The graph has several things we like to see: the basis currently stands at a much more negative level than where it has been over most of the year; we see a lot of volatility, so we have the possibility of quick profit from reversion toward zero basis; and the basis has spiked toward the positive on bad news about the company's credit – specifically, the LBO announcement in April. The last point suggests that an announcement that the LBO will go through as planned will cause the basis to move in a positive direction again.

Exhibit 3

Exhibit 4 illustrates the projected performance of the trade as a function of bond-CDS convergence over a six-month time horizon under assumptions of no change in credit, credit deterioration of 100bp across the curve, and credit improvement of 200bp across the curve. The deteriorating and improving credit scenarios correspond to our estimates of credit performance in the event that the LBO goes through or is abandoned, respectively. Based on our assumptions about trading costs (seven-year CDS bid-ask spread equal to 20bp, bond bid-ask spread equal to US$0.50), the raw basis needs to converge by only about 1/4 of its current value over the six-month period to cover trading costs.

Exhibit 4

Risk
The position is designed to minimise the effect of changes in Sallie Mae's credit spreads over a number of scenarios with a six-month time horizon. This weighting is very close to a DV01-neutral weighting today, so parallel changes in credit spreads in the short term should not affect the value of the position by much at all.

In the recommended weights, the CDS overhedges the bond, so the position will gain in the event of jump to default.

We provide a suggested simple, duration-matched government bond position to hedge interest rate risk. Exhibit 5 presents bond sensitivities in case investors are interested in pursuing more complex hedging strategies.

Exhibit 5

The trade does present currency risk for dollar investors. We expect that investors would hedge this risk at the portfolio level, but are happy to discuss currency hedging strategies upon request.

As with any bond-CDS basis trade, investors should ensure that the bond is a deliverable obligation for the CDS. It also may be advisable to confirm the counterparty's understanding that the CDS would not be orphaned in the event that Sallie Mae tenders for its existing bonds.

The trade has positive carry given current levels, and this carry should cushion the investor from short-term mark-to-market losses.

Execution risk is a factor in any trade; this risk is discussed in more detail in the "Liquidity" section below.

Liquidity
Liquidity – i.e., the ability to transact effectively across the bid-offer spread – is a major driver of any longer-dated transaction. 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 the trade and the bid-offer spread costs.

The recommended CDS is in the seven-year tenor, and we continue to see steady quote flow in this maturity as Sallie Mae remains in the news. We have been seeing decent quote flow on the Sallie Mae CDS in this maturity.

The euro-denominated bond appears to be available in size at the recommended levels.

Fundamentals
As explained, our negative basis trades generally are based on the assumption that the bond is mispriced relative to the CDS, and are not premised on an expectation of general curve movements. As explained above, we do believe that this trade should perform well in the event of a sudden deterioration of Sallie Mae's credit, which is what we'd expect to see if the LBO goes through. We review the company's credit fundamentals here.

Kathleen Shanley, Gimme Credit's Financials expert, maintains a "Deteriorating" credit score on Sallie Mae, largely based on the pending LBO. She notes that existing bondholders would not be equally and ratably secured with the acquisition financing under the present LBO agreement, and that the final form of the LBO – if it is completed at all – is up in the air. Apart from the LBO, Kathleen notes that SLM is the dominant student-loan firm in the country, and boasts a loan portfolio that is largely federally guaranteed.

Summary and trade recommendation
Sallie Mae's bonds and CDS have been both volatile and liquid over the past months as the company has been at the center of attention with a pending LBO, potential adverse changes in Congress, and the questions that have surrounded all securitisers since the beginning of the credit crunch. The euro-denominated 4.75 of March 2014 bond is trading cheap to CDS-implied fair value, and we propose a basis trade that takes advantage of that cheapness and also should perform well under the most likely LBO-related scenarios we see: completion of the LBO without a bond tender offer, as planned; abandonment of the LBO; and completion of the LBO in conjunction with a tender offer for the bonds.

Although we don't have a strong view on the likelihood of the deal's going through (and certainly are not offering any legal opinions), we understand that generally courts have looked askance on efforts to wriggle out of contracts based on the clause on which SLM apparently is relying here. Because the trade is based on mispricing between the bond and CDS markets, it ought to perform well regardless of what happens with the deal, but we note that the trade does benefit from credit curve steepening because the CDS is slightly longer-dated than the bond. As explained in our SLM steepener trade recommendation of 28 September, we would expect the 5s-10s portion of the curve to steepen if the LBO parties do consummate the transaction.

Buy €8.9m notional SLM Corp. 7-year CDS protection at 273bp.

Buy €10m notional (€8.48m cost) SLM Corp 4.75s of March 2014 at 84.82 (z-spread of 327.8bp) to gain 83.6bp of positive carry.

Sell €10.1m notional DBR at a price of 97.71 (€9.87m proceeds) to hedge bond interest rate exposure.

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

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

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

High-frequency CDS index trading

High frequency trading strategies on the CDS indices are discussed by Ulf Erlandsson and Amit Bhattacharyya of Barclays Capital's quantitative credit strategy team

Credit is at a juncture where the asset class offers a deep, liquid, fast-moving market via the standardised indices, but where the decision-making and analysis is more often than not on a monthly, weekly or daily horizon. For example, much credit index analysis will still be based on close of business (COB) spreads.

In this article we argue that this leads to imprecise analysis in many cases, but more importantly to missed opportunities. By entering the realms of high-frequency data, we believe the problems can be alleviated and opportunities unleashed.

In other asset classes, such as equities and FX, trading strategies based on high-frequency switching in and out of individual assets or indices have become a substantial part of the trading strategy spectrum. First and foremost, the capacity to sustain high-frequency trading strategies in these markets comes from diminishing transaction costs. The cost efficiency of being able to switch in and out of a strategy on a minute-by-minute basis is essential to capture short-lived marked fluctuations.

Credit is not yet at the extreme of tic-by-tic trading. Transaction costs are still too high – although they have been shrinking significantly over the last few years, especially in the indices. But slightly lower frequency strategies can still work.

Over the last few months, it has become apparent how broad market sentiments drive indices, and how these sentiments have had an effect outside the geographically-decided trading hours. Exhibit 1 highlights this sentiment by plotting the intraday spreads of iTraxx Main and CDX.IG.

Exhibit 1

What is apparent when overlaying US and European trading hours is how the tightening momentum after the mid-day peak persisted throughout the whole global trading session, even after the European close. We show how the global trading day can be utilised for sophisticated risk management, as well as improving high-frequency trading strategies.

30 July 2007 stands out as a prime example of how intraday analytics are likely to be crucial to investors in credit. It is also testament to the pitfalls of end-of-day studies.

In the European hours, iTraxx Main skyrocketed from 60bp at the open to 76bp at mid-day, only to put up a tremendous rally back to 64bp at the close. CDX.IG prolonged the rally, trading at 84bp when Europe closed (it traded as wide as 101bp during the day) and tightening another 12bp to close at 72bp. Realised volatility, as the basis for Value-at-Risk analysis and swaption pricing, is suddenly a complicated beast.

A key analytical difficulty of intraday data, aside from the difficulty in obtaining and storing it, is the unequal distribution in the time of quotes. In standard day closing data, we assume a fixed time interval between quotes, which is the basis for calculations of differences, returns, realised volatility etc. This is obviously not the case for intraday data, which is also why a slightly different set of tools is needed to conduct meaningful analysis.

We make the distinction here between the "high-frequency" data that we have collected (quotes that can be several hours in between) and what traditionally is referred to as high-frequency data. Usually, this will be either tic data (recordings of each level at which the asset has traded) or data sampled at a 1-10 minute interval. As an over-the-counter market, credit will not have the same sort of quoting frequency, primarily because small moves at high frequency appear irrelevant as a percentage of the bid-ask that is involved in trading at such high frequencies.

One popular device for intraday analysis is the candlestick graph, which depicts the daily trading range, open and close spreads, and direction of the trading session. We show the trading ranges derived from intraday data of the iTraxx and CDX indices in Exhibits 2 to 5.

Exhibit 2 - iTraxx Main

First, the candlestick graphs confirm general market correlation between different markets in both HY and IG. The broad movements are well reflected.

Exhibit 3 - CDX.IG

Second, we note a distinct difference in how the indices have traded intraday. Particularly in the more volatile days, the size and the colouring of the candles often look very different when contrasting European and US indices.

This may seem a trivial observation – the US trades at different hours to Europe and consequently has a different information flow. But it highlights the inaccuracy of comparing COB data between the US and Europe. Such data will inevitably capture the dislocation shown in the difference in the candlestick graphs between geographical areas.

Exhibit 4 - iTraxx Xover

Third, we note that for a large number of days the indices have covered a substantially wider spread than what is reflected in the open-to-close or close-to-close range. In the figures, this is reflected as the (vertical) length of the line compared to the height of the box.

When the line is long relative to the height of the box, we have seen large intraday movement relative to the open-to-close range. We discuss this further below, but note for the timebeing that this gives rise to several different approaches to estimating daily volatility.

Exhibit 5 - CDX.HY

Estimating index volatility
A first application of intraday data is to understand the volatility of index spreads better. This estimation is essential for understanding P&L volatility and consequently relevant for risk management purposes. It can add important information in the pricing of, and relative value between, options.

For an example of how intraday data give a much clearer picture of index dynamics, let's look at the events of 30 July, when the iTraxx Xover breached the 500bp barrier for the first time. The index opened at 462bp, up from 434bp at the close on 27 July. The index quickly sold off up to the 510bp region, then surfed on a wave of protection selling, reaching 472bp at the close.

Below, we present a number of potential volatility measures:

• Interday-based, with a metric based on close-to-close. This measure includes the overnight gap and the moves on market information between the 27-30 July European closes.
30 July move: 38bp
• Open-to-close. This measure requires opening spreads that are not readily available. It allows for greater granularity in terms of extracting the overnight gap and the actual total move during the trading day. In that sense, it allows for a specific view on the actual volatility in a single day. It does not take into account the intra-day moves. In a candlestick graph, this is depicted as the "box".
30 July move: 10bp
• High-to-low. This is the most common measure of intra-day volatility, as it is fairly simple to compute and also reflects the trading range within the day. In a candlestick graph, this is shown as the "line".
30 July move: >46.5bp
• Full intraday dynamics, model-based and intermediate spreads. This requires a full statistical model introducing – not only spread changes, but also dispersion (in time) of quotes.

The list is in ascending order of the complexity of the approaches, as well as in theoretical accuracy. As we can see, there is a distinct difference in the conclusions one can draw about how volatile 30 July really was.

Although the close-to-close measure of 38bp is high, it only ranks as the sixth most volatile day in July through August. And of this volatility, most appears to have been driven by overnight gapping rather than intraday moves.

Our open-to-close measure records only a mediocre volatility of 10bp. The measure that tells the most intuitively accurate picture is the high-to-low metric, which records a 46.5bp move.

The summer of 2007 – evolution of index spreads on an intraday basis
Just a cursory inspection of intra-day data yields some interesting points. In Exhibit 6, we plot quote-by-quote data for the iTraxx indices over the past four months.

Exhibit 6

Each individual section of the graphs represents a trading day, with the dark blue lines outlining the actual index quotes during European trading hours, and light blue lines showing after-hours market movements in the US market on an iTraxx-equivalent basis. We can see that the first patch of volatility in mid-July appeared quite substantial compared to previous history. But it was nowhere near the magnitudes of spread changes that the market underwent later in July.

On the basis of these graphs, we can also observe a marked decrease in volatility from the final weeks of August and onwards. This summary inspection of the data lends credibility to spread volatility being driven by regimes, rather than smooth increases/decreases. The four weeks from 25 July definitely appear to be fully disconnected from what happened both before and after.

Another important observation concerns the sources of total spread volatility. We can see in this graph that movements between close and open – ie, the overnight gaps – constitute a large proportion of total volatility. When looking at the series only covering European trading hours, we expect a fair amount of spread volatility from information flows while the US market was still open. As the next section shows, there are possibilities to actually reduce that portion of the overnight gap risk.

Volatility measures – implication for VaR
Following on from the realisation that volatility can be measured in several alternative ways and that intraday-based measures can capture volatility dynamics that COB data is not able to spot, there could be several implications for risk management. For example, there are two important areas in which intraday dynamics will make Value-at-Risk (VaR) systems more precise.

First, deriving a realised distribution of index dynamics will truly capture the full potential of index movements. This will potentially remedy some of the problems in many realised volatility-based VaR systems, where one is not able to achieve fat enough tails in spread/return distributions. If volatility is based on only interday measures, then some of the actual moves that we have seen on an intraday basis are many standard deviations out, making their appearance very improbable.

Second, being able to track exposures on an intraday basis can bring substantial value for VaR systems. If the VaR systems set risk limits, for example, the limits will start to adjust as soon as spread moves are a reality, rather than at the close of the business day. This area is probably more a systems development issue rather than an analytical one, but we believe that if trading strategies can be designed to work on an intraday basis, then risk management systems could as well.

The time difference bias– implication for empirical analyses
Non-tradeable/aggregated levels can differ substantially from actual tradeable levels and the timing of the actual quote makes a noticeable difference for spread level in volatile times like these. Hence, for empirical analysis purposes, we believe that some COB prices may actually be misleading. For example, COB spread ratios between iTraxx and CDX will be substantially different in this case from the "true" simultaneously-traded spread ratio.

High-frequency trading strategies
Predicting future market performance by looking at historical data has been likened to driving a car by looking in the rear-view mirror. Past performance is no guarantee of future returns, but with precious few instruments that allow us to look into the future, an understanding of the historical market environment plays a big role – especially when "historical" can refer to the dynamics of the market within the last few days, hours, minutes and seconds.

Strategies based on high-frequency momentum effects have been extensively researched and implemented in traditionally more liquid markets and in the equity markets in particular. It therefore seems natural to begin our investigations of high-frequency credit strategies in this space as well.

The growing liquidity and small relative transaction costs now make such strategies a possibility in credit. Additionally, technical and market factors make the existence of momentum in spreads a possibility.

Although intraday trading may not be for the fainthearted, the timing of entry and exit points in the indices can potentially be achieved more efficiently with a momentum strategy, which should be very important to the broader investor base. The method will never call the peak or the trough of spreads exactly and relies on a relatively rapid reaction of inflection points.

This is why higher-frequency data becomes essential for such strategies. Not only does it allow for a quicker reaction to changes in trends, but it is also better at determining the trend itself.

Designing a momentum strategy
A simple type of momentum strategy uses averages of spreads over particular horizons, and potentially with different weighting schemes to compute the average. For example, we can design a rule so that when Xover spreads are above their moving average we have a "Buy protection" signal. Conversely, if the current spread level is below the moving average, we have a "Sell protection" signal.

In this way, in a trending widening market, the current spread will tend to be above the moving average, and we have a "Buy protection" signal. The drawback with the signal is that it will be fairly slow to capture the turn of the trend. The current spread will remain above the moving average for some time after the peak of the index itself.

This sluggishness is a function of which horizon the moving average is calculated over. The longer horizon, the more sluggish the signal reacts to trend changes.

A natural solution is to decrease the horizon over which the average is computed, in which case the signal will become more responsive. However, the signal also becomes more and more sensitive to short run fluctuations as the horizon decreases. With too short an horizon, the signal will tend to switch between "Buy" and "Sell" signals at a frequency that is very punitive in terms of transaction costs.

One way to improve the sluggishness with a minimal loss in stability is to adjust the calculation of the average to put more weight on recent observations rather than older ones.

Momentum in high-frequency data
As hinted at in the previous section, a key characteristic of a momentum strategy is to capture the turning points of a trend as soon as possible. The foremost risk to the strategy is large gaps.

In close-to-close index data, the size of these gaps on average exceeds how much of the trend you can capture via the momentum strategy. Switching into high-frequency data is a natural progression to start reducing gap size and is also extremely useful for implementing momentum strategies. The only caveat with high-frequency data is that it is even more prone than closing data to non-systematic fluctuations.

In addition to the problems of designing a not-too-sluggish, not-too-wobbly moving average in a high-frequency setting, one should also take into account the dispersion in timing of quotes. If there is a cluster of quotes at a particular time or lack of quotes at another, periods and spreads where there are many quotes will be overrepresented in the momentum calculation.

We do not account for this in the strategy presented. A sketch on a solution would include a secondary weighting function, which is based on some dispersion measure for each quote (eg, the number of quotes within the previous 60 minutes). The exact parameterisation remains the topic of future Quantitative Credit Strategy research.

Some final thoughts
We reiterate the point that these strategies are used to illustrate the potential for high-frequency trading in credit indices. Our parsimonious, almost simplistic, models can achieve a decent performance even without layering in more sophisticated rules and data. For example, at times, the tick-by-tick correlation between credit and equity (future) indices has been high, which could be used in designing the momentum rules.

Quote clustering should be properly modelled, and ways to obtain even higher-frequency data than we have in the current sample are certainly open. We also see potential to forecast volatility to turn on or off the strategy. In that way, one could switch the strategy into a carry (long-risk) mode when movements are too small to motivate active switching in and out of the indices.

© 2007 Barclays Capital. This Research Note is an excerpt from 'High-frequency CDS index trading', which was first published by Barclays Capital on 3 October 2007.