Adventures in CDS reporting, GM edition
GM debt has been through a lot of late. In May 2009, car czar Steve Rattner made a bold and unexpected decision to nationalize the company rather than leave it with debt outstanding. That decision was followed by a CDS auction which valued GM’s defaulted debt at just 12.5 cents on the dollar — a valuation unthinkably low just a couple of years earlier. Clearly, when it comes to automaker debt, there’s a lot of uncertainty and volatility — and where there’s debt with uncertainty and volatility, there’s sure to be CDS trading.
The WSJ, however, has decided to take a golly-gee approach to the whole thing, larded with a good sprinkling of demonization. I’m surprised to see veteran bond-market reporter Matt Wirz — a genuine expert when it comes to such things — with a byline on this:
Fresh from Wall Street’s alchemy labs: Credit derivatives tied to General Motors Co. debt. The rub is, no such debt exists…
Investors who bought “naked CDS” to bet on the likelihood of default, rather than to hedge risk from other investments, are credited with worsening the liquidity crisis that gripped financial powerhouses, prompting calls for tighter regulation of the industry.
First of all, there’s no alchemy here. You might not like credit default swaps in general, but they’ve been around a long time, and there’s nothing new or innovative about CDS on GM. Sure, the amount of GM debt outstanding is low, but it’s a bit weird to say that “no such debt exists,” given that there’s still $4.6 billion in bank debt outstanding.
The assertion about the nonexistence of GM debt is backed up with a single extremely vague sentence:
Banks, some of which have made loans to the car maker, have been buying the CDS even though it is unclear whether the contracts would cover their debts, according to people familiar with the matter.
Nowhere is it explained what this is supposed to mean; I’m guessing that there’s a question as to whether a default on GM’s bank loans would trigger the CDS. And then there’s also the question of what would be auctioned and delivered in any CDS auction:
When a company files for bankruptcy or fails to meet its interest payments, the market stages an auction to determine the value of the defaulted debt, and how to compensate the CDS holders.
The value assigned to the CDS relies on investors being able to buy and sell bonds in the open market, so it is problematic for the newly revived GM not to have any bonds outstanding.
This isn’t really true. CDS auction prices are emphatically not a function of the open-market secondary-market price for individual bonds: that’s why there’s an auction in the first place. Would bank loans not be eligible to be tendered as cheapest-to-deliver debt securities? The article doesn’t say. But whenever any company has $4.6 billion in bank loans outstanding, there’s a secondary-market price for those loans, so in principle it should be possible to find them and deliver them. If the number of loans outstanding is small, then that just creates a familiar problem in the CDS market, when the amount of CDS written is larger than the amount of debt outstanding. The CDS market has dealt with that problem many times, and it’s not really an issue any more.
In any event, it has long been common practice for banks to hedge their loan exposure in the CDS market — that’s one of the generally-accepted “legitimate”, or non-naked, forms of CDS trading. There’s a reason why they’re called credit default swaps rather than bond default swaps.
But more to the point, the WSJ seems to be willfully naive about what’s going on here. Why would you sell credit protection on GM debt? Because it currently has very little debt outstanding, because you don’t think it’s going to reach a remotely dangerous level of debt in the next five years, and because you get to cash a steady flow of CDS premiums in the interim. Essentially, exactly the same reasons that you would buy GM bonds, if any existed — only selling protection is much cheaper, so you get a higher internal rate of return.
And why would you buy credit protection on GM debt, if there’s no such debt outstanding? Maybe you intend to buy bonds when GM issues them, and you want to lock in protection now, while it’s cheap. Maybe you are a GM supplier, or you have exposure to one, or in some other way you have GM counterparty risk which is easy and cheap to hedge at the moment. Maybe you’re just taking the opposite side of the GM-Ford relative-value trade featured in the WSJ, betting that over the long term GM is going to continue to struggle in the face of steadily declining US market share. Or maybe you just reckon the price of credit protection on GM debt is going to go up rather than down.
Whatever the dynamics of GM CDS trading, however, this kind of extrapolation is a reach too far:
If the cost of protection on GM continues to trade below Ford, for example, GM should be able to sell bonds at lower yields than Ford.
It’s bizarre to see this at the end of a whole article dedicated to the weirdness of the market in GM CDS, and the fact that the price is largely a function of the fact that GM does not have any bonds outstanding. At some point, GM is going to start issuing new bonds, and at that point various different investment banks will start talking to the carmaker about the level at which they might be priced. I very much doubt that any such bank would tell GM that it could issue through Ford just because of where the two companies’ credit default swaps were trading.
For the time being, GM CDS are trading at a tight level precisely because no one’s expecting a bond issue any time soon. If GM starts making noises about raising money in the bond markets, expect those CDS spreads to widen out significantly. It’s still possible that GM bonds could trade through Ford, of course — after all, Ford would still be much more highly leveraged than GM. But let’s not take today’s CDS market as much of an indication of anything. It might not be financial alchemy. But that still doesn’t make it a particularly useful guide to future bond pricing.