In 2002 Warren Buffett wrote that in his view, credit derivatives were financial weapons of mass destruction ("FWMDs"). He made the comment while explaining to shareholders why he was unwinding a Berkshire Hathaway subsidiary dealing in them, and few outside of the value investor community paid much heed.
Eight years later we could be singing a different tune in both the media and Congress. After all, we've had a financial crisis, $23.7 trillion of potential government financial backstops and a "Great Recession" that were all exacerbated and partially caused by a FWMD explosion. With some notable exceptions however we haven't changed enough, and these contracts remain poorly understood even as the markets have evolved beyond the levels for which Mr. Buffett first expressed concern.
At the time of his statement, Buffett's reasoning was simply that credit derivatives were "time bombs" where parties would enter highly leveraged contracts with risks they often didn't understand or couldn't evaluate. This was effectively describing the problem AIG would later face (and he wanted Berkshire to avoid). In explaining how impossible to value some contracts were, Buffet gave the example that, if desired, he could easily purchase a speculative position on "the number of twins to be born in Nebraska in 2020." Those words should have raised red flags on how improper derivative contracts could easily simulate a massive underground gambling house.
This would not be the worst of it.
When Buffett wrote his letter, a subset of the market known as credit default swaps ("CDS") was still young. In 2000, the entire notional value of this market was just $900 billion, mostly reflecting bona fide "good faith" buyer hedging. For an individual, bona fide hedging is like buying insurance on your home; in return for a monthly premium (the contract's "spread"), you receive downside protection that pays in the event of catastrophic loss.
But in the early 2000s the CDS market started to change, and new investors whose primary purpose was to take speculative positions emerged in greater force. The reasons why are not hard to understand as potential payouts are enormous; buying $1 million of CDS protection can return as much as $200 million if the underlying company collapses. This 200X return is in contrast to the maximum 2X (ie, 100% profit) return an investor would receive using a generic equity short under the same scenario, because CDS by themselves are more inherently leveraged (ie, they simulate higher levels of debt).
By 2007 the CDS market had mushroomed to a notional value in excess of $45 trillion. At those levels, at least $20 trillion was primarily speculative -- we know this because the combined underlying markets they were based on (bonds and certain structured products) had a total value of less than $25 trillion. This would be like buying a home insurance plan worth $450,000 on the same day you acquired the house for $250,000.
That minimum of $20 trillion in contracts was therefore "naked," or at least mostly so. In finance, being naked generally means transacting in an asset without any ownership or correlated upside in that asset.
Put another way, buying a 5-year naked CDS is like buying life insurance on a stranger. The main point is that you would receive zero benefit if the stranger does well in life; in fact, you would stand to lose money (the premium) if they survived to the end of the contract. However, if they didn't do well, you would earn a fortune. This is both immoral and illegal for an individual but allowed to investors under current law.
If that were the end of it -- simple speculative insurance -- it would be problematic enough. In theory, we might even derive some benefits in the form of increased liquidity and better pricing for all insurance contracts.
However, this would only work in a world without avarice. In fact, the pure logic of buying life insurance on a stranger actually has a moral advantage to buying a naked CDS -- at least with individuals, the act itself of buying life insurance doesn't harm the stranger. It only creates financial incentive to do so.
In the case of naked CDS, buying insurance creates both incentive for harm and the circumstance for actual harm. This is similar to how a naked short works, except that naked shorts are quite rare and illegal under most instances. There are also two structural differences:
- CDS are more intrinsically leveraged (remember, 200X vs 2X).
- CDS relate to a different part of the capital structure, debt instead of equity.
In the case of a theoretical naked short manipulation, a speculator would sell large quantities of a company's equity in order to drive the price down, but do so with shares they either don't own or that don't exist ("phantom" shares). This creates a highly concentrated burst of selling activity in excess of what would be possible using only "real" shares, shares that could either be borrowed or owned outright.
For companies that don't access the capital markets regularly, CDS pressure can be frustrating but not debilitating. However, for leveraged financial companies -- who often rely on short-term, daily funding sources -- the results of a spiraling and unknown increase in their CDS would directly affect the core business and be much more dangerous. At minimum this would hurt their cost of capital, reducing profit margins in a healthy bank or causing a weak bank to lose money more quickly. In extreme scenarios, this would even create a confidence-liquidity trap and the modern-day equivalent of a bank run. The only difference here is that we would have institutions rather than consumers lining up for their money, and they would technically not be taking money out but refusing to put money in.
I've heard it speculated that this is exactly what happened in the cases of both Bear Stearns and Lehman Brothers. The evidence -- circumstantial at best -- is that one can see a series of sharp increases in CDS spreads for both companies prior to public awareness of any issue. Sadly, forensically determining what happened is next to impossible because the players leave few traces and are exempt from oversight since the Commodity Futures Modernization Act of 2000. Whether or not CDS pressure was exerted however, what remains likely is that had a similarly weak peer institution (say, Merrill Lynch) experienced the CDS run-up instead of Lehman, their situations and timing of collapse could easily have reversed.
More recently, I have read rumors that a dispute between a TARP-receiving investment bank and the fiscal authorities of Greece escalated into the use of CDS pressure by the bank and two hedge funds to increase the government's cost of debt. This is probably not true, however it is probably also plausible; George Soros famously "broke" the Bank of England using less aggressive instruments and with far less market power than the big banks have today.
- Credit default swaps are not complicated; they are simply corporate insurance. They should be treated like insurance by the authorities, especially after their misuse became a focal point in the US financial crisis.
- Naked CDS are the corporate equivalent of buying life insurance on a stranger. They should be treated no differently than the theoretical potential for naked equity shorting abuse.
I hope that readers and staff of the Huffington Post may join me in calling for regulators to propose a ban on abusive naked CDS that is at minimum similar to the one already in place that bans abusive naked shorts. In addition, I hope that together we can push more elected officials into doing the work needed to genuinely understand these issues beyond basic sound bytes. If they do, I am confident that more attention will be paid to smart proposals for dual-regime oversight such as the one put forward by CFTC Chairman Gary Gensler. In it, he establishes the basis for overseeing not just the markets (through exchanges, clearing houses, etc) but also the dealers themselves. Without a dual-regime approach, it is just a matter of time before smart, unregulated traders find new ways around the system because it is their financial incentive to do so.