Any Greek Restructuring Should Be Designed To Trigger A Credit Event
As talk about an actual restructuring of Greek debt increases, the EU continues to think that avoiding a CDS Credit Event is a good thing. More and more stories and leaks indicate that a real restructuring of Greek debt is on the table, with write-offs of as much as 50%. Whether it will be real, permanent reductions in principal this time, or some other form of principal protected rollover with a subjective NPV calculation like the 21% haircut, remains to be seen. In any case, the EU continues to head down the path of bending over backwards to avoid triggering a CDS Credit Event.
They are wrong to be avoiding a Credit Event on the Hellenic Republic. If they are really pushing for a true restructuring where banks and insurance companies are for all intents and purposes forced to accept a big haircut, they should want to trigger a CDS Credit Event. Allegedly, the EU is avoiding a credit event because it “could unleash a cascade of losses” according to a Bloomberg article. That reason just makes no sense, as we’ll discuss below. It seems also that pride plays a role as the EU doesn’t want to have the stigma of a default and tries to avoid calling even a 50% write-off “a default”. Finally, the EU also seems to want to punish speculators. This is where they really have it wrong, because very few hedge funds are likely short Greece via CDS at this time, but the problems created by avoiding a “CDS Credit Event” while realizing a meaningful restructuring of the bonds would directly impact bank risk management desks and will have long-term negative consequences for sovereign debt demand.
It will be interesting to see what form the latest “restructuring” takes. What bells, whistles, and accounting gimmicks are built in, but maybe for once it will be real. Triggering a Credit Event and surviving it, would inspire far more confidence than the desperate attempt to avoid it. The other issue that is hard to understand, is what will happen to bonds that aren’t restructured? Non Restructured bonds and CDS will have a very curious relationship going forward.
Bank Credit Risk Management Back to the Dark Ages
JP Morgan is one of the few financial institutions that have come through the financial crisis with an enhanced reputation. Their skill for managing through the problems has been clear and the respect for the firm, and Mr. Dimon in particular, has grown. It is their risk management that has been the key to their success. On their earnings call they were able to point to low “net” exposure to Europe. They had “Portfolio hedging” for their European exposures of $5.2 billion, 80% of which was related to sovereigns. So it is possible that JPM has bought about $4.2 billion of sovereign CDS (short the credit) to manage their overall exposure to Europe. If the regulators can pressure banks into taking big write-downs on their positions and make their CDS ineffective simultaneously, how long will it be before Mr. Dimon realizes he has to do something different to manage his exposure.
If buying CDS is unlikely to provide the relief it should, the only way to reduce economic exposure is to sell assets. JPM would likely be ahead of the curve and sell their CDS while it still had value, and sell bonds/loans at the same time. Other banks and investors will eventually realize that they cannot rely on “net” exposures, when the regulators corrupt the product. Investors will start to focus on gross exposures because they will doubt the ability of banks to ever monetize their hedges. All the big banks, still likely to be risk averse, even after some multi-trillion euro EFSF announcement, will want to maintain low economic exposure to European sovereign debt. If they don’t believe their hedges will protect them because they would once again be forced to write down assets and not collect on their hedge, the only prudent risk management decision is to reduce assets.
By eliminating a tool for banks to hedge their bond exposure by blatantly working around it, the EU will reduce future demand for bonds. Not exactly what they are trying to accomplish. The whole point of the EFSF and other programs is to stimulate demand for bonds, and they will have achieved the opposite as big banks will have to reduce bond holdings since they will realize they cannot rely on their hedges.
At the other extreme, some weak banks, the ones who likely wrote CDS rather than buying bonds because of the leverage (to create long credit positions), will want to write even more CDS. Why would they ever want to buy bonds when the EU just taught them that selling CDS is “free money.” The weaker the institution, the more appealing that trade will be. So in the future, the unregulated, difficult to track CDS risk, will all be the hands of the weakest institutions – again, a result that does nothing good in the long run.
The Facts Do Not Show a Risk of “Cascading Losses” From a CDS Credit Event on Greece
According to DTCC, the net Hellenic Republic exposure in the entire system is €2.7 billion. Yes, the open longs (or open shorts) are a total of €2.7 billion. That is trivial compared to the €330 billion or so of Greek bonds outstanding. It isn’t even 1% of the exposure the system has via the bond market. With a 50% haircut, bond investors will lose about €165 billion, and in the CDS market, a total of €1.3 billion would find its way from the net sellers of protection to the net buyers of protection. If the regulators and EU are sure the system can handle the bond write-offs, the write-offs for CDS are a rounding error, at best.
There is a lot of confusion over DTCC in Europe and how much the DTCC captures. For trades done over the past few years, any single name CDS would have gone through DTCC in Europe, same as in the US. Just like in the US, many legacy trades would have been transferred over, but there are likely legacy CDS trades still out there. As a recurring theme, can you guess where those CDS reside? Who wrote that protection? If it is out there at all, it is not at the hedge funds, so it is at banks and insurance companies. Big, active banks would have shifted to DTCC and moved their legacy exposure into the system as well. In the end, any CDS not in this number was likely sold by weak banks and some insurance companies. The number should be small, the regulators and EU should know this number, but we cannot know how big or large it is since the regulators failed. Greece could be in some synthetic CDO’s, but since it was only ever rated as high as A1 and didn’t trade particularly cheap for its rating, I doubt it was included in many synthetic CDO’s. Since banks could write protection on a single name basis and get 0 risk weighting, they would also have less incentive to include in CDO structures. In the end, I believe the DTCC number is reasonable and that legacy trades are rounding error relative to the CDS exposure let alone the Bond/Loan exposure, but cannot confirm it.
What about the “transfer mechanism”? Isn’t there some way the chain of payments could break down? There gross notional for Hellenic Republic CDS is €54 billion. These represent a combination of things, but primarily dealer to dealer trades where one dealer has an ultimate seller, and the other dealer has an ultimate buyer, and the trades run through them since they either don’t have that client or weren’t the “axe” at the time the client was putting on the trade. There are also curve trades. Curve trades will collapse down. The street will run “tri-optima” or some equivalent to net the risks down. Banks are well prepared for the settlement of CDS. The settlement of Lehman went smoothly in spite of concern at the time. There is no reason to expect it not to go smoothly this time. Once again, JPM’s quarterly Earnings Presentation has some useful insights. They have $8.2 billion of Trading Exposure to Europe, which is “predominantly client-driven derivatives exposure of $14.2 billion, offset by collateral of $6.7 billion (95%+ held in cash).” I’m willing to assume that JPM is doing a good job on their counterparty risk management. Other big banks are going to be very similar.
But let’s look at a worst case. Assume one bank (“Dumb Bank”) has written the entire €2.7 billion of net outstanding Greek CDS. That bank then owes €1.3 billion. If the bank doesn’t have the money to pay, they would not pay the money to whomever they sold the protection to. They would have sold it to one of the “Dealer” banks, one of the 20-30 biggest banks that make markets in CDS as part of their core fixed income platforms. If they had trades on with several bank, then each of those banks would take a loss. It would not change their obligation to pay on their contracts? Does anyone really believe that one of the big banks couldn’t afford that €1.3 billion loss? It would be painful, but they would absorb it, and the rest of the payments would flow through the system. They would honor their obligation to whoever they sold CDS to, in spite of not receiving the money. That is how the system works. The extreme example where one bank provided that much counterparty exposure to one institution that couldn’t pay is unrealistic, but at least from the CDS chain of events, the losses would end at the big bank(s) that made that decision. No Contagion.
The Dealer Bank would then proceed against Dumb Bank to collect its claim. Dumb Bank would enter into bankruptcy in some form or another. Bondholders would have a loss, and Dumb Bank’s other counterparties would all have to scramble to replace risk. So this does have the potential to create contagion, but is the market really so stupid that no one would have noticed how bad Dumb Bank’s finances were? The debt wouldn’t be trading anywhere close to par if the bank had such big exposures and was in that much trouble. The loss to any single bank from triggering CDS is not likely to be enough to force them into bankruptcy. Unless a bank has been able to hide massive exposures from the market neither the share price nor the debt of these banks should be significantly affected by monetizing a mark to market loss already priced in, and in many cases, already collateralized.
The system is just not that fragile, and the possible payments from triggering CDS are negligible relative to the losses that will be experienced from the bond market write-downs. If the EU believes the financial system can handle writing down the €330 billion of bonds (and I believe it can), then it is highly unlikely that the additional losses on €2.7 billion of CDS will be the straw that breaks the camel’s back. It is just not plausible as the number is small, and the counterparty risk management is actually pretty good, and this would require gross negligence in virtually each and every bank to trigger the contagion risk the EU seems to fear.
Pride and Punishment
An actual restructuring where financial institutions permanently write-off 50% of the principal owed is a default by any other name. Pretending it isn’t a default so you can say you have never defaulted is just bizarre. The loss can be called anything you want, but the end result is the same. Worrying about the semantics of having had a CDS Credit Event is just absurd. You can say that you “are slightly above ideal weight” but people will still know you are fat.
And who is getting punished? Reading between the lines, the EU seems to be licking their chops at punishing all the hedge fund speculators who are short Greek risk having bought CDS. Well, guess what? They are NOT short Greek risk anymore. The hedge funds are now generally flat or even long Greek risk by owning bonds or having sold CDS. All you need to do is think about it for a moment. Greece trades at 62 points up front. So on a $10 million trade, you pay or receive $6.2 million. If you felt governments weren’t going to manipulate the situation, where would you think the CDS would trade after a Credit Event? What is the recovery rate then? I think assuming anything lower than 20% is extremely aggressive. So you are risking 62 to make 18? (paying 62 points with the hope of receiving 80) That would require a high degree of certainty, or an even lower recovery assumption. It would also require you not to have read a newspaper or turned on the TV for the past month. The G-20, the IMF, the EU, the ECB, are all lined up to try to prevent a default, and even more importantly, continue to state that they want to avoid triggering a CDS Credit Event. You are making a bet against their ability to circumvent the rules. From a risk/reward standpoint, I would not be short Greek risk by having bought CDS. If anything I would have sold Greece CDS here (especially with talk of a 50% settlement). I would much rather be short French or Belgium risk by having purchased CDS. They have a lot more opportunity to widen, with a limited ability to tighten.
But if hedge funds aren’t short Greek risk through CDS, who is? Bank hedging desks! The banks have bought CDS on Greece in their hedging books to cover some of their bond and loan exposures. Banks do not want to take off their hedges because they don’t want to report larger net exposures. They aren’t taking profits on these because optically they cannot report to shareholders increased exposure to Greece. They have done the same analysis as hedge funds and would like to cut their shorts/hedges, but this isn’t about making money for the banks anymore, this is about presenting low exposure numbers. The smart, hedged banks will be the ones punished. The EU wants to punish the hedge funds, but all they will do is punish banks that have been most prudent.
And who is rewarded? Good old Dumb Bank. They sold some CDS because they could get more leverage, and here they are the ones being rewarded by the EU. The efforts to punish are likely to punish the wrong people and further reward the weakest institutions. At one time hedge funds were short Greece via CDS, but at one time I was young and athletic – things change over time. The EU should get over their anger and think responsibly. That is the only way to truly start correcting the core of the problems. Lashing out by manipulating markets and rules will do more harm than good.
Since the program is “voluntary”, not all bonds will be restructured. Some bonds will remain outstanding. All of the CDS will remain outstanding. Yes, the EU can avoid triggering a Credit Event if they want, but the CDS does not go away. The CDS remains outstanding based on its original Scheduled Termination Date. As bonds that weren’t restructured come due, what will the EU do? Will they allow those bonds to be paid at par in order to avoid a Credit Event? Given their fear of CDS and belief that they will have solved all Greek debt problems, any bonds that don’t get restructured will likely be paid back at par. So if you hold a bond that if you agree to voluntary restructuring you potential lose 50% of the value, but if you don’t agree, you will receive par, you have a strong incentive not to participate. In a more normal, less manipulated market, you could expect that the authorities would attempt to default on non-restructured obligations in an effort to punish those who didn’t play the game. They would want to send the message that you are better off participating in their programs. Yet, their fear of CDS may cause them to reward those institutions that restructure the fewest if any bonds. Avoiding a Credit Event now does not eliminate the CDS and unless 100% of existing bonds are restructured, there will be some interesting games that get played.
Copyright © 2011 · Peter Tchir
* * *