Posts Tagged ‘moral hazard’

adverse efsf

October 24, 2011

There is a widespread view that the EU will attempt to lever up the EFSF capital to achieve an insurance capacity of EUR2-3tr, by turning it into the equity tranche of a sovereign CDO. Others put forward a TARP-like structure which will buy back bad sovereign debt.

A Special Purpose Vehicle (SPV) will be the financial entity that will implement this endeavour, where whatever is left of the initial EUR440bn EFSF money will serve as the capital. This is certaily a big number, which is bound to catch the headlines and trigger a stock market rally. But is there any meat on this bone?

How do you value that crap?

Although yet unknown, the bond buy-back structure will be probably similar to the TARP model which was implemented in the US to help banks offload their ‘misunderstood’ assets. The idea behind TARP was that the market for CDOs and other morgage backed instruments just dissappeared overnight. For that reason, if banks were to sell these assets they would receive very little, even though foreclosures at the time were not justifying such a severe price cut.

In order to implement the TARP, a price discovery mechanism was needed to gauge what a fair value would have been. An ‘inverse auction’ was put into place, where the holders of the assets posted the prices at which they were prepared to sell, and the lowest prices were filled first until capacity was reached. Therefore selling institutions had an incentive to post a low price to get rid of their assets, and losses were written down. To fill in the holes these sales punched in their capital, banks issued preferred shares that were bought by the same programme. A total of USD700bn were earmarked for this programme, of which about two-thirds were actually used.

How would that work in the context of the EFSF? Here Greek, Portugese and Irish bonds appear to be eligible at the moment, while Spanish and Italian bonds could become eligible in the future (if eligibility is defined as being bailed out). Banks that hold periphery bonds will post a selling price, and the cheapest ones would be filled first. One does not know yet the exact mechanics, but it is safe to assume that Greek bonds will be the first ones to go. Obviously, the current market price serves as a lower bound, but if enough banks try to get rid of the bonds, then the price might not be too far away.

The main driver behind the TARP was that households were not defaulting on their mortgages at a rate that would justify the market prices, and that if these ‘misunderstood’ securities were kept for a while then the market would resume and reasonable price levels would be restored. The current situation is markedly different: the problems facing Greece (and other periphery countries) are not due to the lack of liquidity but lack of solvency. The low price of Greek debt is not unreasonable if one considers the fundamentls, and is in fact supported by the implicit or expected haircuts described in the various PSIs. If these PSIs were gone, the price of Greek debt would be arguable even lower. Pumping liquidity in the secondary market would do little good to the banks involved, and no good to the Greek state.

The related recapitalisation would be welcome though. Banks gain some immunity against sovereign default if the bonds are off their balance sheets. This also takes the burden to recapitalise the banking sector off the sovereign, but causes a moral hazard issue which invalidates the whole idea.

Survival of the fastest

As it stands the EUR440bn are not sufficient to provide a decent backstop. Already nearly half of the money has been spent, but let’s say that there is another EUR250bn left in the fund, and the liabilities of the larger periphery countries easily exceed that. An argument is to use the magic of leverage to turn these EUR250bn into EUR2tr or even more. I think that this argument is flawed, if one considers the incentives of the countries involved.

The size of the Greek default is constantly being debated in the financial press over the lat few weeks. The initial PSI of 21% has grown into a PSI+ of 50-60%, with some commentators advocating levels of 90-100%. There is a valid argument that the costs of recapitalising the Greek banking sector do not go away. Every Euro that a Greek bank loses through a haircut is a Euro that Greece has to borrow in order to recapitalise the bank. This means that the true impact of a 50% haircut might be actually less than the 21% that was originally proposed.

In the TARP world the default of a household was disconnected from the profit or loss that the MBS realised. But in the EFSF case this does not hold: if Portugal defaults then the Portugese state will need funds to recapitalise Portugese banks; if the EUR250bn have run dry because Greece and Ireland have already tapped the fund, then Portugal will be left exposed. It matters little that EUR2tr were insured, if the equity tranche is wiped out. The incentive for Portugal is to default before everyone else, just to make sure that her banks take advantage of the EFSF rescue funds.

The size of the default is now subject to adverse selection too. If Portugal need not worry about the impact of a default on her own banking system, then she might opt for a much larger haircut. As her banks are recapitalised externally, why not give them a helping hand?

Obviously moral hazard becomes an even more important issue for Spain and Italy, where the money involved would be stretched in the better of scenarios. If they want to have any chance of protecting their financial system, then they have to rush to the EFSF gates early in the day.

What this leverage will create, as the crisis deteriorates, is the equivalent of a good old bank run. A leveraged institution might not have enough capital, and everyone eligible rushes to get the most out of it. It also resembles the good old CDO: if CDOs underestimated tail dependence and systemic correlations, European politicians are making sure that a cluster of sovereign defaults will happen in the EU sooner rather than later.

As if there is not enough contagion as it is…