Megan McArdle asks, "Do We Hate Credit Default Swaps for The Wrong Reasons?" As Megan notes, blaming credit default swaps for all kinds of things is quite fashionable these days, since simply uttering the term makes commentators feel sophisticated. While this is itself a topic worthy of discussion, the more interesting point in Megan's article concerns how credit defaults swaps affect the incentives of bondholders in the context of restructurings.
The basic argument is as follows: Suppose that ABC Co. is on the verge of bankruptcy, but wants to avoid bankruptcy by restructuring its debt (renegotiating interest rates, maturity, etc.) with its bondholders. Further, assume that bondholder B has fully hedged his ABC bonds using CDS, or over-hedged to the point where B would profit from ABC's bankruptcy. The problem seems obvious: B either doesn't care if ABC does or actually wants ABC to file for bankruptcy, and so he will do anything he can to stop ABC from restructuring and force ABC into ruin.
At first blush, this looks like a serious loophole and a nice way to make some fast cash. Sadly, there are several reasons why this is not the case. The key factor to understanding why we shouldn't expect this to be a major problem is to appreciate that there is no CDS vending machine. You cannot go to the market and demand credit protection on all of your bonds at your whim. You have to find someone willing to take the exact opposite position that you are taking. That is, if you bet heads they bet tails, by definition. As a result, if everyone knows the next toss is coming up heads, you probably won't find someone to take the opposite side of that bet.
As discussed above, when you buy protection, you (the protection buyer) buy it from someone else (the protection seller) who will end up paying out if a bankruptcy does indeed occur. These protection sellers are very interested in making money, and so, as the probability of default increases, the price of protection or "spread" widens, making it more expensive to purchase protection. So, as firms get closer to a restructuring or bankruptcy, the cost of buying CDS protection on soon-to-be-junk bonds skyrockets. And not only does the cost of protection go up, liquidity, or your ability to enter into CDS trades, on distressed entities dries up. There's a fine reason for this too. As the probability of default edges closer to certainty, fewer people are willing to take the other side of the trade. They're just as convinced as you are that ABC will fail, and they'll tell you to go sell your bridge to someone else.
This means that in order to take advantage of the restructuring-sabotage-strategy, you have to either (i) guess which companies are doomed for failure well in advance of any real trouble; or (ii) wait for trouble and then lay out a ton of cash and find someone stupid enough to take the obviously wrong side of a bet with you. Neither scenario seems likely to occur often, since (i) requires some fairly remarkable foresight and (ii) requires remarkably stupid counterparties. Moreover, in the case of (i), if you're truly convinced that ABC is headed for restructuring or bankruptcy, you can buy protection with "Restructuring" as a credit event, which means that if ABC does restructure, you'll get paid. So, in that case, you don't have to sabotage anything. You can just sit back and wait for an ABC restructuring or ABC bankruptcy, since you'll get paid in either case.
Moreover, rather than waste all that time and effort trying to sabotage a restructuring, you can cash in before a bankruptcy ever occurs. As the spread widens beyond the point at which you bought in, your end of the trade is "in the money," and so it already has intrinsic value that you can realize in a variety of ways. For example, assume that when you bought protection on ABC, the spread was 150 bps. When rumors abound that ABC is entering talks with its bondholders, you can be sure that the spread will be well above 150 bps. Let's say that the spread widened to 1000 bps. As a protection buyer, your side of the trade has economic value that you can realize by entering into another trade in which you sell protection to someone else. (The CDS market has recently begun changing the way CDS spreads are paid, but we'll assume we're operating under the old system where the protection buyer pays the spread in quarterly installments). That is, you sell protection at 1000 bps, pay for protection at 150 bps, and keep the remaining 850 bps for yourself. Sure, you could go for the gold and sabotage a restructuring, but that's a lot more involved than simply entering into an offsetting trade and pocketing the juice.
In addition to the market based reasons above, there are corporate governance reasons why we shouldn't coddle these kinds of claims. When a company issues bonds, it includes terms that it and its bondholders must live up to. That is, each bondholder could be asked to swear on a stack of bibles that, "I will not go out and buy CDS protection to the hilt and ruin you." If a company were truly concerned about the risk of restructuring-sabotage, it would include such terms.
*And* they have to actually own the bonds to be able to queer the restructuring. Many CDS buyers didn't/don't hold any bonds.
I had read that companies were buying CDS as a way to increase the rating on bonds. In other words, if something was crap with a BBB rating, you buy CDS on the company to improve the rating to AAA. So your i) case above would happen a lot more if people were routinely insuring their bond purchases to improve their rating.
I thought this was how AIG got into so much trouble -- it was selling CDS to improve ratings on mortgage-based products. It sold at very low prices (because everyone knows real estate never goes down everywhere at once) and got burned.
Correct me if I'm wrong. I only know what I read on financial blogs!
mgoodfel,
In general, yes. But there are a host of scenarios where banks and others used CDS. The situation you're referring to comes in two general flavors: one, to keep assets you can't sell on your books without taking a capital charge; and two, to make bonds more appealing to investors by having someone (usually a monoline or AIG) "wrap" the deal, that is provide insurance on the bonds, usually through a CDS (which is then guaranteed by the monoline).
In the first case, if banks have some assets that they can't sell, that would normally require them to set aside capital to cover them. Alternatively, they could enter into a CDS on those assets, which will afford them some capital relief. In the second case, certain investors are only willing to purchase highly rated bonds, and so the wrap provides the bump-up in ratings that investors are looking for, since if the bonds don't pay out, the monoline or AIG will (hopefully).
1) Would Bondholder B ever want to see restructuring go forth more as a result of holding a CDS position in ABC? Possible if he's grabbed the arb, but 850bp is nothing compared to the CDS payout. I'm trying to think of a way that this regulatory feature of CDS would stabilize companies and the greater economy by encouraging restructuring, but I don't see one.
2) How much do you worry that the CDS vendor cost is really just transferred back to the ABC Corporation in the form on higher financing costs? Even if Bondholder B was paying nearly what the CDS contract pays out, how much pressure will that increase in trades put on the ABC corporation?
All in all a clear explanation of how things work if everyone involved is both well-informed and clever, i.e. it supports the theory that financial markets do not need regulation because self-interest will make sure that everyone wins...
Unfortunately the past 24 months have shown that this is not always the case. I would be interested in hearing if the CDS market should be regulated or not. If firms like AIG are not to be punished for their obvious stupidity we have a serious problem.
Yeah, sure; but up to the credit crunch biting "really stupid counterparties" seem to have been in ample supply. I don't recall derivatives markets back beyond the 1980s; but whenever these markets have been really active since, the counterparty who is either stupid or too clever for his/her own good seems to have always been there and eager to take the suckers' bet.
CDS are a great idea but lets trade them through exchanges. That would contain the counterprty risk; and even the mugs would have a fair chance of understanding their positions.
As my opinion and mentioned above in the article, credit default swaps are not a issue for me but insurance policies given a different name so as to avoid retaining the reserves required by regulations of the insurance company.Annuity Ratings
Kinda sorta. Your explanation not only implies that there was a orderly secondary market for CDS but some degree of visibility into the assets underlying the securities. CDS "protection sellers" were essentially flying blind hoping that profits from selling contracts would more than outweigh the cost of any defaults. There was very little understanding of the underlying credit quality and little ability to remarket these contracts.
In reality there was never any compelling reason to arbitrage. You could get 'free' money by becoming a protection seller and worrying about default if/when it happened.
Xavier Van Zandt,
There is no assumption about a "secondary market" for CDS. You just need to enter into an offsetting trade. You are not novating or "selling" your current trade, but entering into a new one, the net effect of which is to cash in.
There is also no arbitrage. Before you came into the money, you took on risk. Namely, the risk that your CDS protection would end up being a giant waste of money.
If by "CDS protection sellers flying blind" you mean the bond insurers, yes, they clearly did not appreciate the risks they were taking on. But I don't see how that's relevant to this particular issue.
The above article discusses arbitrage. Without a secondary market with some degree of liquidity, arbitrage is problematic. It was also largely unnecessary in the case of CDS insurers.
I'm not clear on what points you're trying to make here.
Xavier Van Zandt,
As author of the above article, I assure you it does not discuss arbitrage, but control rights.
You seem convinced that the protection buyer is selling his position. That is not the case and so there is no "secondary market" trading. He is entering into another, NEW, offsetting position, which allows him to profit. You would be correct in saying that the CDS market itself must be liquid in order for this to happen. As a general matter, the CDS market is fairly liquid.
Secondly, that is not arbitrage from the perspective of the initial trade. At the time the trade was first entered into, the spreads could have moved against the protection buyer, causing him to be on the losing end of a trade. Therefore, he took on risk. Since there is risk, the trade is not "arbitrage."