CIT CDS selloff leaves correlation traders sitting pretty; Sallie Mae could be next

This article is provided to FT.com readers by Debtwire—the most informed news service available for financial professionals in fixed income markets across the world. www.debtwire.com

--------------------------------------------------------------------------------------------------------

When CIT’s flirtation with bankruptcy sent the credit default swaps market into panic two weeks ago, some traders sat and watched the mayhem with smug smiles. Correlation trading desks at a few firms, including Natixis and Goldman Sachs, have been buying up CIT protection on the cheap since January to hedge risk in the instruments they structure and trade.

Correlation desks fueled the explosion of CDS in 2004-2005 by selling bundled protection on dozens of credits at-a-time through securitizations called synthetic CDOs or “bespokes”. The dealers marketed part of the deals to investors but held chunky tranches on their own books and devised complex models to dictate how and when to buy additional protection to keep their exposure flat.

But as defaults proliferate, some traders are dropping quantitative modeling in favor of old-fashioned fundamental analysis to hedge certain credits – like CIT – that are particularly common in bespoke deals. The result is that those ostensibly risk-neutral correlation books stand to gain significantly if individual credits go under.

Student loan giant Sallie Mae is the next name in correlation traders’ crosshairs, two correlation traders, an analyst at a boutique brokerage and a sellside desk analyst told Debtwire. Five-year protection on Sallie Mae – another popular name in bespoke deals – trades at 20 points upfront, reflecting a relatively low risk of default in post-credit crunch terms.

Nevertheless, the company’s outsize representation in synthetic deals and mounting concerns about the impact of pending legislation is compelling some desks to act now.

A spokesman at Goldman and officials at Natixis declined to comment.

Risk Factory

While correlation desks don’t print much new business nowadays, five years ago they were one of the biggest growth drivers for the red hot CDS market. Dealers used the desks to securitize some of the risk they built up buying and selling single name protection for investors.

The market makers packaged their exposure into tranches that they sold to insurance companies, pension funds and hedge funds, many of them in Europe. The bespoke deals typically mixed 100 corporate names and sold slices with different degrees of risk, much like a standard collateralized debt obligation (CDO).

Most investors bought the mezzanine pieces of the structure – split into junior and senior – and the dealers held on to the equity and senior tranches. The junior mezz began to take losses if 10% of names in the bespoke defaulted and were wiped out if 15% went under, while the seniors took losses for defaults exceeding 15% and were wiped out at 30%.

In order to offset the risk of the equity and super senior tranches they kept, dealers used complicated models built by quantitative analysts. The models spit out directions for how much protection to buy when spreads on a portfolio name moved out on any given day and traders complied by positioning in the single name CDS market, index tranches and the CDX indices.

That strategy worked only as long as perceived default risk – and related CDS movement – changed gradually. After Lehman fell, perception of “jump to default” risk spiked, and, in 2008, the models broke down. Once a bespoke favorite hits trouble, it starts a negative feedback loop that hits other names in the structure, multiplying the deal’s exposure.

“If a name blows out from 20 to 40 points and you have negative convexity, you can go from being relatively flat to massively long,” explained one of the correlation traders. “It hits you twice.”

“CIT was easily in 8 out of 10 bespokes” said the trader.

Broken Models, Cheap Protection

As the cost of CDS blew out across the board, the quantitative correlation models dictated increasingly expensive hedges and traders abandoned them, said two analysts. The desks shifted focus to fundamental research instead to decide which credits to pick and pan, they said.

CIT stood out to some traders, both because of its exposure to the credit crunch and its ubiquitous placement in the bespokes, said the two correlation traders. Just six months ago, jump to default exposure to the name averaged roughly EUR 50m for many correlation desks, estimated two of the trader sources.

One correlation trader who runs a mid-sized bespoke book said that he bought USD 85m in short dated protection at an average price of 12 points upfront in January and throughout the spring. Five-year protection on the name was quoted at 34 on 9 July but ballooned out to the high 50s by 20 July when it became clear the government would not bail out the asset based lender.

Much of that price movement originated from other correlation desks rushing to hedge their jump risk, said the correlation trader and a trader on the buyside. The bid/asks on CIT’s five-year CDS has since leveled at 48/50 points as the company negotiates with its debt holders.

Though not as widely held in bespokes as CIT, nor as distressed, Sallie Mae caught attention in recent weeks as correlation desks braced for the company to be downgraded to junk, all the sources said. Standard & Poor’s last week placed the ratings on credit watch following a vote by the House of Representatives Education and Labor Committee to pass a bill that would eliminate the origination of federal student loans by private lenders after July 2010. Sallie Mae is rated BBB- by S&P and Ba1 by Moody’s

Offers to sell protection from dealers grew scarce this week and several dealers have faded on offers, in a sign that not all desks are covered for Sallie Mae jump risk, said three traders. Some bids did get picked up this week but the market is now 18/20 upfront on the contracts as opposed to 691bps running on 19 July, according to a broker and data from Markit.

Hedging jump to default risks puts correlation desks into positive jump risk territory on a mark-to-market basis but in a losing position if the credit in question avoids default and spreads eventually recover, noted a third trader. The decision is more art than science and ultimately requires traders to make a credit call, said the trader and an analyst.

--------------------------------------------------------------------------------------------------------

For more information or to inquire about a trial please email sales@debtwire.com or call Americas: +1 212-686-5374 Europe: +44 (0)20 7059 6113 Asia-Pacific: +852 2158 9731