Saturday, January 20, 2007

Mike is Back - This Time It's Derivatives

First a primer for those unfamiliar with Credit Default Swaps (CDS):

CDS are contracts that transfer default risk of a given borrower (usually based on a bond): The buyer of a credit swap receives credit protection, whereas the seller of the swap guarantees the credit worthiness of the bond. By doing this, the risk of default is transferred from the holder of the bond to the seller of the swap. Buyers usually pay an upfront lump sum and regular annual payments to keep this "insurance" in force. Prices are usually quoted as X thousand dollars per $10 million protected. For example, a swap buyer may pay $65,000 per year in order to receive $10 million in the event of default by ABC Inc., whose bonds the buyer may or may not possess. Apart from hedging, CDS are also used for speculation. If you believe that ABC's creditworthiness will deteriorate, you can buy a CDS with the expectation that its price will go up.

Let's consider a simple bond: it is a contract that stipulates that a particular borrower (the bond issuer) will pay interest on the money borrowed and return the principal at maturity. As such, the coupon rate of the bond itself when issued (i.e. the interest rate) should fully reflect the risk of default. High quality borrowers pay less, low quality pay more. Furthermore, the price variations of the bond itself after issuance (i.e. the secondary market) should also reflect all available information about its current ability to service its debt. Increasing risk will lower the price and decreasing risk will increase the price. Two observations:
  1. Creditworthiness is, by definition, included in and reflected by the price of the bond itself at all times. Theoretically, therefore, bond holders already own imputed default protection within the bond itself. By buying a CDS on a bond they already own, bondholders pay twice for the same thing - unless of course such credit-protected bonds are always mis-priced. Not a likely state of affairs in efficient markets, but possible.
  2. The issuer of CDS is operating as an insurance company selling life policies against corporate "death". It collects premiums from CDS sales, invests them and calculates it will get to keep more than it has to eventually pay out. Underwriting corporate health is extremely risky business; unlike human actuarial data, corporate default risk is very volatile and uncertain. Still, if you price the stuff properly you should end up with a wash, i.e. right around the risk-free rate of return. Add #1 above plus cost of doing business and this business can't possibly exist - at least on paper. But it does! So what gives?
There are only two - mutually exclusive - logical explanations for this state of affairs. That on average:

a) The CDS buyers are strictly un-savvy speculators constantly and wrongly punting on credit deterioration and who routinely overpay for their bets - resulting in handsome profits for the underwriters.
b) The CDS issuers are underpricing their policies to such a degree that both speculators and hedgers are buying them like hot cakes.

I don't know for a fact which is true, though I suspect (b) strictly from the astonishing growth of the amounts involved. Keep in mind that as long as the business cycle stays positive defaults are low and premiums are almost pure profit, i.e. the current return is almost infinity. But, of course, in order to properly price credit risk one has to take into account the other side of the business cycle, where all defaults normally pile up. It doesn't seem that CDS issuers are looking past the crest.

The last time an industry was born out of mis-pricing credit risk we got junk bonds: Mike Milken, Drexel Burnham, LBO's, The Predators Ball - and the collapse of S&L's plus the Crash of 1987. It looks to me like we are again mis-pricing risk today: all sorts of credit derivatives, sub-prime loans, the real estate bubble and - again - LBO's and "private equity" funds. And, ominously, a blast from the past:

Bloomberg reports that Michael Milken (long out of jail by now) will make a speech at a conference in London on high yield debt organised by BNP Paribas. He will say that by adding derivatives on loans, bonds and credit cards, poor nations could raise between $50 trillion and $100 trillion.

Can you imagine the fees involved in raising $50-100 trillion? Yum Yum Yum. Well, I don't know about you, but I lived through the Drexel era and the ding-ding-ding I'm hearing surely can't be the lunch wagon.


  1. Sorry for the bad link in the previous post, try this one

    Deutsche Bank Swap Makes Pennsylvania Taxpayers Lose

    Seems the school district got into a interest rate swap and lost about a quarter mill but they will pay that out of another swap they have. Thanks also for the link to definitions, I have bookmarked it.
    I agree that this state of affairs is caused by (b) but the above article even though it is not a CDS says (a) is also a factor along with the greed for fees IMHO.

  2. Thanks very much, great story. You are exactly right in your assessment about fees, greed, etc. Municipal finance has always been a breeding ground for either a) skinning the "dumb hicks" and/or b) corruption. In this case I think it was a desperate, poor school district throwing the dice on what they thought was an "80% chance", to get some extra cash for their schools. They were just easy prey for the "advisers". I feel sorry for them, I know first hand what it's like to run a school on a pitiful budget (I was the president of a school parents' assn. for a few years).

    Towards the end of the story comes the sad truth about highly complex structures for small clients. The hospital in Tennessee paid $3 million for advice on swaps and has lowered interest payments by $1 million...not a great bargain. The swap bank probably made another $3 mil in spreads. Total ratio 6/1. Typical.

    Those $25 mil bonuses have to come from someplace...