Sunday, July 24, 2011

Greek Games

I haven't made up my mind yet about the wisdom of the latest plan to secure the financial viability of Greece within the eurozone, but as a piece of financial engineering it has some very intriguing features.

Current bondholders have the option of exchanging their assets for new issues that promise less and deliver later, but are considerably more secure. There are four new issues to choose from, varying with respect to maturity, interest rate, and the proportion of principal that is guaranteed (by highly rated zero coupon bonds or funds held in an escrow account). But these options are designed to be roughly equivalent in present value terms, and it is expected that they will be selected in approximately equal measure by those who choose to participate in the exchange.

Participation is voluntary, so current bondholders can simply choose to do nothing. For this reason, the financing offer does not trigger payouts on credit default swaps.

What makes the mechanism strategically interesting is that the payoffs from participation are highly sensitive to the overall participation rate. The higher the participation rate, the greater will be the ability of Greece to meet its financial obligations not only on the new issues but also on the outstanding ones. Participation by some raises the value of the assets held by the remainder. If the target participation rate of 90% is met, then those who decline to participate will find themselves holding bonds that are much less likely to default than is currently the case. In anticipation of this effect, yields on Greek bonds (and the cost of insuring them with credit derivatives) fell sharply following the announcement.

It's interesting to think about who gains and who loses from this. Contrary to most accounts in the media, current bondholders benefit from the existence of the financing offer, regardless of whether or not they choose to participate. Those who decline to participate experience a capital gain on their assets (relative to the status quo without the offer). And those who participate are choosing to forgo this capital gain and must therefore be even better off. Of course, there will be many bondholders who purchased their assets at times when Greek default was considered highly unlikely, and they will experience a loss on their original investment. But this loss has already been inflicted on them: the financing offer just gives them an opportunity to capitalize it in a manner that eases Greece's debt burden, as an alternative to selling their bonds in the open market. 

The fact that credit default swaps are not triggered by the offer, coupled with the lowered likelihood of default on current bonds, benefits sellers of protection on Greek debt. In fact, such sellers have strong incentives to buy up Greek bonds and participate in the exchange, thus lowering the probability that a credit event will arise in the near future. I would not be surprised if some of the buying that raised prices on the heels of the announcement came from such sources.

So who loses as a result of the financing offer? First and foremost, those who bought naked credit default swaps, thus making a directional bet on a credit event that is now less likely to occur. It is quite conceivable that the plan was designed to have precisely this effect. Speculators betting on sovereign default have come in for a fair amount of public criticism by political leaders in Europe, and stand accused of raising the cost of borrowing and the likelihood of default. (I have argued in joint work with Yeon-Koo Che that there is some theoretical basis for this claim.)

Costs will also be imposed on the countries of the eurozone core, who are providing the collateral to guarantee principal on the new issues (Greece remains solely responsible for all interest payments). But these countries are motivated by the belief that a formal default by Greece would have contagion effects across the periphery, leading to a chaotic collapse of the currency union. The biggest risk entailed in the current initiative is that it may not, in the end, be enough to prevent this.


  1. Good clear post Rajiv! I hadn't seen this explained before, and it's important.

    Two minor comments:

    1. Presumably there will be an interior equilibrium. The more people who participate, the lower the risk of default for those who do not participate, and so the smaller the gains to participation.

    2. There seems to be a (positive) externality here. Each person who participates confers an unpriced benefit on those who do not participate. Hmmm. One would have thought that some body (like those countries guaranteeing the new bonds) could have captured those external benefits. Isn't that how the IMF can make a profit?

  2. Hi Nick,

    Yes, there must be an interior equilibrium. If almost all participate the residual bonds will be virtually safe, so all will prefer not to participate. If none participate then Greek defaults and all would prefer to have taken the safer option.

    But the equilibrium participation rate depends on details of the mechanism, especially simultaneous versus sequential choice, disclosure of participation rates in real time, short versus long window to decide.

    My guess is that the best way to internalize the externality and push participation rates up would be to have no disclosure and a short window, effectively approximating simultaneous choice.