Credit Default Swaps

Sunder Ram Korivi NISM What makes a CDS unique?

 An insurance-like product used by credit institutions, traded like a capital market instrument  It is the connecting thread between 3 markets: Insurance, Credit and Capital  OTC versus Exchange traded instrument  Regulatory Arbitrage Risk-Mitigation Devices: When Markets are Down…

 Put Options  Short-selling of bonds and bond indices  CAT Bonds (insurance-reinsurance alt)catastrophe bonds or alternate risk transfer(ART)these are 1 yr bonds  Contingent Convertible (CoCo)  Credit Insurance  Bank Guarantee  Letter of Credit  Deposit insurance and moral hazard  Credit Default Swaps Credit Default Swaps

 Invented at /  JP Morgan was an early adopter  Became popular at Morgan Stanley  John Paulson, Michael Burry used it on CDO  Standardization of CDS and ISDA  GS, DB: sellers-turned-buyers  AIG became the biggest seller  RBI – cautious beginning CDS Application Areas

Underlying assets:  Corporate Bonds  Sovereign Bonds  Mortgage Backed Securities  Asset Backed Securities  CDO

Business models:  One-off deals and  Pooled risks Statistical Context

 Under-writing considerations  Law of Large Numbers  Tail Risk  Gambler‟s Ruin  Default Correlations CDS evolution: Corp Bonds. Bankers Trust and DB

 CDS: an insurance policy against default on a bond  In 1998, Deutsche Bank wanted to become a powerhouse in derivatives. DB wanted to take over Bankers Trust, a based investment banks packed with quant experts who thrived on designing complex securities  A new derivative instrument, called Credit Linked Notes, later became more commonly known as Credit Default Swaps, created in the early 1990s  Initially, there were only a few trades every day, a sleepy arcane, illiquid business  CDS really took off after the math wizards at JP Morgan got into this field  Ronald Tanemura, a former Salomon Bros employee, was the trailblazer from Deutsche Bank who trained Boaz Weinstein CDS evolution: Concept, Pricing, the Traded CDS

 A debt of $ 10 mn to GM, if insured for a premium of $ 1 mn, implied a 10% chance of default  The lending bank is comfortable in paying $ 1 mn for the protection bought. Firms like DB sold protection. Trades were, in those days, bespoke (custom-tailored between two parties)  DB needn‟t wait until default or maturity. It could buy protection from someone for $ 2 mn, if a worsening credit of GM was perceived.  In theory, it was just like betting on defaults levels, like on stock prices or price levels. So the swap trades could be several times the actual loan outstanding. OTC, no central counter-party.  In practice, swaps are written on large baskets of bonds and loans  No regulatory oversight over the CDS market  By late 2000, the value of CDS traded exceeded $ 1 tn, rising to 4.5 tn by 2004, to 42.6 tn by 2007, as per BIS  Buying a CDS while shorting the underlying bond (or index) could actually trigger the payoff from the CDS.

CDS evolution: Deutsche Bank gambit

Arbitrage: Short the stock and sell CDS.  Weinstein of DB tried it on the bond of GM in May 2005. Rationale: Stock price falls, capture the gain. Whereas bondholders have safety through the debt covenants. Default is not triggered. The CDS premium collected is easy money.  In another set of events, someone made a block-deal bid for GM shares. While S&P downed the GM bonds to junk. Both sides of the trade soured for Weinstein. They decided to double down.  Fortunately, in end-2005, GM shares lost considerable value and the bonds were upgraded.  Lesson: sometimes, things spin out of control  Thinking ahead, one needn‟t actually hold the loan to trade in CDS. Selling and buying CDS was all based on a perception of default  This left the world of KYC and originate-to-hold far behind  Bear Stern was also brought down by shorting the stock and selling CDS on its debt, in a classic Cap Stru Arb CDS evolution: Enter JP Morgan

 Until now, CDS were applied in the corporate bond market  The surreal world of CDS met the world of securitization. That‟s when the idea of CDS got married to the idea of CDOs  In 1997, JP Morgan insured its 300 corporate loans worth $ 9.7 bn through CDS bought from investors, transferred these slices into an SPV and got capital relief, retaining the high-grade tranche. retained. This synthetic CDO was called the Broad Index Secured Trust Offering (BISTRO).  From there on, the CDO-CDS market took off, in the same manner as had Morgan Stanley‟s Gerry Bamberger‟s stat-arb (Ed Thorp, pioneer of stat arb and convertible arbitrage) through block trades  A robust secondary market for CDS trading sprang up.  In 1998, Brooksley Born, head of Chicago Futures Trading Commission (CFTC) was discouraged by Warren Buffet and Alan Greenspan from bring CDS under exchange surveillance. CDS evolution: Default Correlations

 BISTRO was an example of over-the-top quantitative creativity  What if more than the expected number of loans blew up? (the rotten apple analogy). Are defaults correlated across loans, across pools? This bothered Aaron Brown of Citigroup and Boaz Weinstein of DB.  David X Li, a Chinese-born, Canadian-educated actuary, studied the default correlation problem. It was based on survival models of spouses. The same was applied to corporate bonds/loans. The paper first appeared in 1997 through his employer, JP Morgan and later, 2000, in the Journal of Fixed Income.  CDS prices were assumed to reflect all default risk. This obviated the need for micro-granularity of borrower-level data. Data examined was CDS on each of the underlying CDO tranches, and correlations were found to be low. This was imported into the modeling of the risk premium in CDO pricing. It was a brilliant*and neat „redux‟ of default estimation.  The correlations, gathered from CDS price information were low, all over the place, and assumed a normal Gaussian distribution, with a few outliers at the tails.

*Flaw: Heisenberg Principle, as enunciated by George Soros

David Li’s Model in Transition: Corporate to Individual loans

 However, from 2004, this model, hitherto used for CDOs on corporate debt, was applied to CDOs on retail mortgage debt. Default correlations were assumed to be low…this was the flaw in the model, as CDOs later proved to be substantially over-valued.  More and more substandard loans (sub-prime, Alt-A, NINJA, liar loans) were being stuffed into the CDOs, as the entire mortgage loan sector became commission driven. Default correlations were bound to be high  However, the overhang of Gaussian normality prevailed due to convenience and ignorance.  To make matters worse, CDS prices considered high house prices (or borrowers) as insurance against bad debts. This went undetected for months, as profitable foreclosures in 2005 and 2006 masked defaults, and were labeled as Prepayment (and not Foreclosure)  [Later, ISDA specified that the credit event is the non-payment of installments, to be recognized in CDS and CDO pricing]. Three types of CDS contracts- A quick review.

 Physically settled CDS  Cash settled CDS and Digital  Traded CDS

 Single name CDS  Basket CDS CDS and the Sub-prime Meldtown

 The idea of CDS was born out of bets on more prepayment v. less prepayment impacting CDOs. Then things changed in 2005.  Made popular at Morgan Stanley 2003, by Howie Hubler & Mike Edman. Originally to protect proprietary sub-prime CDO portfolio. Was a one-off, Non-standard, illiquid, opaque and arcane contract  MS needed protection for portfolio. Found someone stupid enough to sell CDS to MS, someone with a bullish view on mortgages. Found German investors in 2005, who misread fine print and trusted ratings  Greg Lippmann (Deutsche), Mike Edman (MS) and others at GS hammered out a standard CDS, blessed by ISDA. By 2006, the mortgage market machine was roaring.  Faults: VaR, coined as RiskMetrics by JP Morgan, was a variant of Bankers Trust‟s RAROC; was based on past (better) originations. AAA disguised the rot. Also, loan default correlations were assumed low, due to geographical diversification. Moody‟s assumed 30% correlations for BBB bonds. Moody‟s stamped 80% of loans as AAA. Retail loans defaulted en masse, compared to corporate bonds. Raters missed it.

CDS and Sub-prime

 Deutsche Bank celebrated their success at betting against the sub-prime meltdown in the second half of 2007. Their strategy of buying CDS on sub-prime CDOs paid off.  AIG FP, based in London, side-stepped regulations and used its hitherto AAA ratings to raise cheap funds. They sold $ 400 bn of CDS on sub-prime debt.  Banks could buy CDS and obtain capital relief on the underlying loans  Unregulated Hedge Funds, CDS markets, cheap money from carry-trade and the originate-to-distribute models created mayhem in the financial system. Alan Greenspan‟s reliance on Adam Smith‟s notion of self-correcting markets, and the brilliance of quantitative finance proved to be the undoing of the financial system.  Quant models don‟t work when panic sets in. There was lack of faith in the CDS seller‟s balance sheet to honour the deal. Sub-prime Turmoil and Thereafter

 DB Bought corporate bonds of distressed (but fundamentally good companies) and sold CDS on them. He gained initially, when the bonds gained prices  The subsequent panics, led to shocks, causing capital losses and CDS claims  With a large inventory, it was not profitable to short the market and gain from falling prices  Jim Simons of Renaissance Capital, said he hadn‟t been affected by the sub-prime meltdown as he didn‟t deal in CDOs and CDSs.  Trading in CDSs dried up. Other instruments like ETFs came in.  Hedge funds came under regulatory scrutiny  Dark Pools, High-speed, Algorithmic and Prop-desk trading also came under regulatory scrutiny CDS – Consequences of Misapplication

 Stress tests were conducted with scenarios of only 6% default.  Thus, models produced higher values.  MS sold CDS to DB. DB made a claim at 70, whereas MS models valued at 95. DB agreed to make market 70/77, trapping MS. CDS – Those who Profited

 2005 witnessed a spike in use of CDS  In 2005, CDS contracts were standardized by ISDA  Cost of CDS rose from 1% to 3%  managers John Paulson and Michael Burry were prominent buyers of CDS, to monetize their bearish view on CDOs

Fabulous Fab and his ‘Ab’ ABACUS 2007-AC1

 Synthetic CDOs were a mind-blowing concept; gather a quantity of sub prime CDOs, held together by a CDS  Faster than warehousing, less riskier. Took advantage of distress of New Century, Fremont etc  ABACUS: a platform for those wanting to go short on the mortgage markets, a $ 2 billion synthetic CDO.  Suited Paulson (he also bought CDS on all other products: e.g. Bear Stern debts). CDS which cost 0.18% in Dec 06, cost 7.5% in March 2008 …and other dubious deals

 In Mar 06, put together a CDO2, ABACUS HGS1, exclusively for the hedge fund of Bear Sterns  Anderson Mezzanine Funding, another deal put together by Goldman, based on New Century originations, with a 50% equity. Rabobank and Smith Breeden, investors, backed out  Timberwolf, another CDO series, was successfully sold in Mar 07. Collapsed 85%, after GS sold it at par to Bear Sterns! And GS bought CDS from AIG on this!  Then IKB and some Belgian investors overcame inhibitions and bought ABACUS  ABN Amro sold CDS to GS, which were sold to Paulson  Goldman began to axe the remaining mortgages on its books Enter AIG

 Traditionally, used its B/S strength and AAA rating to borrow cheap and deal in Interest Rate Swaps  Its FP subsidiary, based in London, chose the lax French regulation from a choice of Euro  Led by Howard Sosin, a hardcore quant from Drexel  Learnt about JP Morgan‟s BISTRO. Quants at AIG dissected it, it seemed like free money. AIG entered CDS deals in 2001  AIG wrote up $ 500 bn of business, of which $ 61 bn was for European banks that wanted to lay off some risk for capital relief. To them, it was a piece of clever financial engineering  AIG boasted of its ability to hold devalued instruments until recovery  Ratcheted up on CDS deals from 2005, including sub prime, at behest of GS perhaps up to $ 600 bn  GS finally, came up with the nastiest of claims, that were triggered by its own market machinations  In the absence of a private sector solution, AIG was bailed out by the government, as it was a lynchpin counterparty to the biggest of the firms. Alan Greenspan in 2005…

“ In recent decades, a vast risk management and pricing system has evolved, combining the best insights of mathematicians and financial experts, supported by major advances in computer and communications technology. In recent times, a Nobel Prize was awarded for the discovery of the pricing model that underpins much of the advance in the derivatives markets. The modern risk management paradigm held sway for decades…derivatives have contributed to the stability of the banking system by allowing banks, especially the largest, systemically important banks, to measure and manage their risks more effectively. In essence, prudential regulation is by the market through counterparty evaluation rather than monitoring by authorities. We regulators are often perceived as constraining excessive risk-taking more effectively than is demonstrably possible in practice. Private regulation has proved far better at constraining excessive risk-taking than has government regulation” “There‟s nothing involved in Federal regulation that makes it superior to market regulation. There appears to be no need for government regulation of off- exchange derivative transactions. We do not believe a public policy case exists to justify this government intervention” ..Alan Greenspan in the 2008 aftermath

“ I made a mistake in assuming that the self-interests of organizations, banks and others, were such that they were best capable of protecting their own shareholders and their equity in the firms…The problem here is, something that looked like a very solid edifice, and indeed a critical pillar to market competition and free markets, did break down. And I think that shocked me. I still do not understand why it happened, and, to the extent that I figure out what happened and why, I will change my views.” “To exist, you need an ideology. The question is, whether it is accurate or not. Yes, I have found a flaw. I do not know or how significant of permanent it is. But I have been very distressed by the fact. A flaw in the model that I perceived is the critical functioning structure that defines how the world works, so to speak. The whole intellectual edifice collapsed in the summer of last year. I made a mistake. I was shocked. Because I had been going for forty years, with considerable evidence that it was working exceptionally well ”

Sobering thoughts…

 Issuance of CDS seems like a low risk, profitable business, so firms will load up on them to maximize profits. Bonuses are annual. No worry about long- term risk, and things blow up eventually.  CDS markets involve counter-party risk, left un- monitored in OTC markets, privately traded, including Bear Sterns  AIG had $ 400 bn in CDS, and was a critical counterparty to significant portions of the financial system  CDS – a $ 60 trillion market, grew without regulation Counter-Point

 Benoit Mandelbrot: a case against EMT  Warren Buffet: the anomaly  Nicholas Nassem Taleb: Hume, Bacon and Popper  Paul Wilmott: Not entirely David Li‟s fault. Past data have limitations in modeling

 ..there IS something called Model Risk (George Soros on Heisenberg and Reflexivity) CFA India Investment Conference

 Pranay Gupta: Fundamental and Quantitative approaches  Brian Springer: What it takes to get around regulation Select References

 Rene Stulz: Credit Default Swaps  David Li: On Default Correlation: Copula Function  Christopher Culp: Structured Finance  Michael Lewis: The Big Short  William Cohan: Money and Power  Greg Zuckerman: The Greatest Trade Ever  Scott Peterson:  Andrew Ross Sorkin: Too Big to Fail  Satyajit Das: Traders, Guns and Money  Satyajit Das: Extreme Money