Parsing banks’ exposure to Greece

By Felix Salmon
June 17, 2011
The Guardian has UBS data on the exposure that European banks have to Greek sovereign debt; the grand total, of €93 billion, seems low to me, especially when you back out the €46 billion owned by Greek banks.  

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

The Guardian has UBS data on the exposure that European banks have to Greek sovereign debt; the grand total, of €93 billion, seems low to me, especially when you back out the €46 billion owned by Greek banks. Add up all the French banks combined and you get to €9.3 billion; Germany’s even lower, at €7.9 billion. All of these sums are entirely manageable and imply that the impact of a Greek default on European bank solvency would be de minimis.

But those aren’t the only numbers out there. Kash has found data from the BIS which shows much larger exposures: $65 billion in France, $40 billion in Germany. (And another $40 billion in the US, which I’ll come to in a minute.)

But it’s worth looking at the raw data here — which is found on pages 102 and 103 of the PDF. The French $65 billion is made up of $57 billion in direct exposure, and $8 billion in potential derivatives exposure. And of that $57 billion, just $2 billion is held by banks: most is held by the public sector and the non-bank private sector. In Germany, too, direct exposure of banks to Greece is a mere $2 billion — and total derivatives exposure is even lower than in France.

The main conclusion I draw from all this is that no one really knows what the effects of a Greek default would be — but that non-Greek banks are unlikely to be the main vector of any contagion. And while Kash is worried about US banks’ derivatives exposure, I’m pretty sanguine on that front, too.

For one thing, $40 billion is not huge in terms of derivatives exposure — especially when we don’t know how much is held by banks and how much by deep-pocketed insurance companies and unleveraged fixed-income investors.

More conceptually, crises occur when banks suddenly find that debt they thought very safe is in fact very risky. That’s what happened during the financial crisis with structured products carrying triple-A credit ratings, and that’s what could happen again with European sovereign debt which carried a zero risk weighting under many bank-capital regimes. It’s emphatically not what’s happening with US investors who are writing credit protection on Greece, in full knowledge that a default is a real possibility.

That credit protection is an expensive hedge for European investors who got exposed to sovereign debt when it seemed much less risky than it does now. For the people selling the protection, it’s a speculative play that there won’t be a formal event of default — and when banks make speculative plays, they can generally cope fine if and when those bets go bad.

None of which is to say that a Greek default would be easily manageable. There’s those Greek banks, for one, which could act as contagion vectors — and of course there’s Spanish and Portugese debt, too, and a whole slew of other highly-correlated assets we can only guess at ex ante. What we do know, though, is that Greece has been a very, very slow trainwreck — even more than Argentina was. And when debt crises come slowly, they’re generally more manageable than when they come fast (think Russia in 1998).

So although no one is going to enjoy a Greek default, I suspect that in and of itself it won’t prove catastrophic. But I could be very, very wrong about that. And that’s a risk no one in Europe really wants to take.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

You’re getting closer. In fact, your conclusion is denotatively correct. If the Irish government hadn’t made a CF of everything, a Greek default should be managed and manageable.

But they did, and it’s very difficult to argue–from a philosophical perspective, let alone a political one–that the countries that did much more correctly in the run-up to the “Little Depression” (as Dr. DeLong is now calling it) should suffer more than a country that was actually mismanaged.

Any haircut that Greek bondholders suffer is a starting point in the negotiations with the Irish and even the Spanish–both of whose RE bubbles were inflated in no small part by French and German banks whose knowledge of proper risk management practices appears to be on a par with Countrywide’s.

As I said a week or so ago, the correct move (for everyone, in the long term) when the ECB and EU started playing games with Greece would have been for Ireland to announce a haircut to its bondholders and let the market decide.

Sequence counts. Giving Greece a haircut sets a minimum level for what Spain and/or Ireland would expect. But it doesn’t set a maximum.

As I said, I blame the Irish. Had they not immediately Geithnerized their banks, we would be in wind-down phase right now. But given that they did, the negotiations should have started with the Most Solvent Country, not the one least able to deal with “austerity demands.”

There’s a reason FT headlines follow the same pattern: Day 1: Greece restructuring imminent. Day 2: Investors in Spain fear contagion effect of Greek restructuring. Day 3: Oops, They Did It Again.

Meanwhile, there are several French and German banks whose asset markings would make Vikram Pandit proud.

Posted by klhoughton | Report as abusive

“$40 billion is not huge in terms of derivatives exposure”

That is true if you are confusing notional and exposure. It may even be true if you are confusing current exposure and potential future exposure. Not otherwise.

Posted by Greycap | Report as abusive

For some reason I am reminded of the markets’ 1992 ejection of the UK from the ERM under Norman Lamont… as long as Lamont promised the pound would be supported, it was worth taking a bet against the currency. Seems like so long as everyone promises to support Greece, the same dynamic could apply.

The funny thing about debt is that when expressed as a percentage, you can halve your debt by doubling the size of your economy; the danger is that higher rates and austerity measures halve the size of the economy but leave the debt alone – which really means it increases as a percentage.

Posted by FifthDecade | Report as abusive

My technical analysis indicators continue to warn of significant EURUSD / S&P500 weakness and USD strength.

In short, the EURUSD chart is very bearish.

Posted by GrandSupercycle | Report as abusive

Give this data 394 I would say that UBS is completely wrong.

Total exposure is around $468bn, they apparently left off a zero….

Posted by ChrisMaresca | Report as abusive

Apples and Pears. One adds Bank and Private debt together to look at the size of Government exposure; the other adds Private and Government debt together to separate out the Bank debt.

Posted by FifthDecade | Report as abusive

I wonder how much of the ECB/IMF response has deliberately been to turn this into a “very, very slow trainwreck”, just to allow people to get prepared for it. I agree that that will make problems less — as you imply, most derivatives are short-term enough that a lot of those currently outstanding have been written since the Greek government first admitted to having been misstating numbers quite badly — though I imagine there are a few corporate entities that have been insolvent for three years and clinging to Greek debt exposure because they couldn’t afford, from an accounting and regulatory standpoint, to write it down. I hope there aren’t too many and they aren’t too big.

Posted by dWj | Report as abusive

I don’t view these two data sets as inconsistent. (Chris Maresca- I think you are wrong)

The UBS number is the exposure that European banks have to Greek sovereign bonds.
The BIS data is the total exposure to the Greek public sector, banking sector and private sector.
So if you take the $57bn of direct exposure from France to Greece- (the way I read it) is that French institutions (of which some but not all will be banks) own $15bn of public sector debt, have $2bn exposure to banks and have $39bn of bank loans to the private sector in Greece.
Which ties in with 1. what french banks have publically declared as their govt bond holdings and 2. what we know about the size of the french owned greek banks balance sheets.

The big question is that when (if) Greece defaults, what happens to the private sector exposures. Given that these are being provisioned against over time, and a large amount is likely to be loans to companies with non greece revenue sources (shipping companies etc), I wouldn’t view these exposures themselves as being the problem, but rather the contagion effects.

Felix- I think you are reading the data wrong- ie the groupings- banks/ public sector/ private sector is Frances exposure to that sector in Greece NOT French banks/public sector exposure to Greece.

Posted by haraldo | Report as abusive