Is Ireland’s problem a Basel problem?

By Felix Salmon
November 24, 2010
Simon Nixon thinks so: the fact that investors won't lend to Bank of Ireland, he says, "highlights a major weakness of the Basel capital rules that European banks operate under".

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

Does the Ireland crisis bespeak a major weakness in the Basel capital-adequacy regime? Simon Nixon thinks so: the fact that investors won’t lend to Bank of Ireland, he says, “highlights a major weakness of the Basel capital rules that European banks operate under.”

This is an interesting idea: Ireland’s problem is a banking problem, banking problems are Basel problems, and therefore it stands to reason that Ireland’s problem might be a Basel problem. But if you look more closely at Nixon’s reasoning, his thesis ends up falling apart.

Nixon lays the blame at the feet of Basel’s well-known weakness: the fact that it concentrates on risk-weighted assets rather than total assets. And he implies that there might be large national differences when it comes to the ratio of risk-weighted assets to total assets, while conceding that thesis ” is impossible to prove from regulatory disclosures.”

But there are three huge things missing from Nixon’s piece. First, he takes just one bank from each of four different countries (Santander in Spain, BNP Paribas in France, Barclays in the UK, and Deutsche Bank in Germany) to illustrate national differences. He would be much more interesting, and much more compelling, if he presented a couple more datapoints in each country, to help give readers a better idea of whether French banks in general tend to have a higher ratio of risk-weighted assets to total assets than German banks in general, or whether we’re just looking at idiosyncratic differences between BNP Paribas and Deutsche Bank.

Second, Nixon thinks that the credibility problem in Ireland is a function of the way that the banks’ risk-weighted assets are calculated. He’s right that the market is skeptical that Bank of Ireland really has a core Tier 1 ratio of 8%. And he’s right that risk-weighted assets are a key part of that calculation, and can throw it off. Essentially, the ratio is (A-L)/R, where A is total assets, L is total liabilities, and R is risk-weighted assets. If R is artificially low, then that makes the ratio artificially high.

But the fact is that it’s not the denominator here that the markets are worried about. Instead, it’s the numerator. The key problematic number is A, Bank of Ireland’s total assets. Many of those assets are Irish commercial real-estate loans for which there’s essentially no buyer right now except for the Irish government. And a huge proportion of the rest are Irish residential mortgages, which might be performing for the time being but which would surely sell for much less than par if the bank tried to sell them on the open market.

Any mark-to-market valuation of BoI’s assets, then, would almost certainly show the bank to be insolvent. (This is not news: it’s true of all banks in all crises.) And the reason that the market won’t lend to BoI is that it fears the bank is insolvent. And that has nothing to do with its risk weightings at all: it doesn’t matter what the denominator is, if the numerator is negative.

Finally, Nixon nowhere mentions any Irish ratios of risk-weighted assets to total assets! The very heart of his thesis would seem to be that Basel understated the riskiness of Irish banks by coming up with an unreasonably low number for their risk-weighted assets. Yet Nixon doesn’t tell us what Bank of Ireland’s ratios were, in comparison to those other European banks, and he doesn’t give ratios for any other Irish banks, either.

I suspect this is more than just an oversight. In general, banks reduce their ratio of risk-weighted assets to total assets in one of three ways. They can be heavily involved in investment banking, where loans tend to be taken out in the repo market and are fully collateralized; they can be heavily invested in structured products with triple-A credit ratings; or they can be heavily exposed to OECD sovereign credits. Ireland’s banks, by contrast, were more old-fashioned than that: they just loaded up on property loans, which tend to carry a full risk weighting. There are clearly lots of things wrong with Ireland’s banks, but I doubt that artificially reduced risk weighting was one of them. Certainly Nixon adduces no evidence that it was.

Is Ireland’s problem a Basel problem, then? I don’t think so—or if it is, then we’d need to see a lot more numbers first before Nixon came close to making his case. I understand that the Heard column has space constraints and specializes in short, punchy analysis, but this piece is so short as to be pretty much useless. At the very least, Heard should allow its writers to put extra material online, showing their work, as it were, to back up the conclusions in the printed paper.

6 comments

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 http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/

Great post. To put it more succinctly, total assets are risk-weighted assets where everything on the balance sheet gets a 100% weight and everything else gets a 0% weight (or a variation with some weight for off-balance-sheet exposures). How can that possibly be better than a more risk-sensitive weighting system, properly implemented and supervised?

The “properly implemented and supervised” proviso pertains to both. We have plenty of past experience with failed leverage ratio systems.

Posted by Eric_H | Report as abusive

“… property loans, which tend to carry a full risk weighting”

Well, their *current*, manifest difficulties are due to commercial property loans. Para 74 of Basel II says these should be weighted at 100% … but the footnote allows this to be reduced to 50%, essentially at the discretion of the national regulator. So the strength of this “tendency” is determined by the CEBS.

And residential mortgages represent the bulk of future, unknown losses that are influencing market expectations. These are weighted at 35% – para 72.

Posted by Greycap | Report as abusive

If only Basel II were so simple! Some loans for the Irish banks would use the standardised Basel approach, while others would use internal (IRB) models.

On the standardised approach, the Irish authorities took some more prudent measures on the introduction of Basel II, given the property market was already quite hot in the lead up to Basel II implementation. Residential mortgages would be allowed a 35% risk weighting (in line with the Basel II standard), but the Irish regulator indicated (http://www.financialregulator.ie/indust ry-sectors/credit-institutions/superviso ry-disclosures/Documents/Implementation% 20of%20the%20CRD.pdf) that for loans with a loan-to-value greater than 75%, the portion over 75% LTV would need to be risk weighted at 75% or 100% (depending on conditions met).

Residential investor mortgages or second homes are risk weighted at 75% or 100% (again depending on conditions). Commercial property and property development loans were 100% risk weighted However, loans for land acquisition or if a planned development did not have 50% pre-lets by value were designated as “specialist lending” under the Basel II rules and were 150% risk weighted. If the borrower could provide alternative security or access to other cashflows to repay the loan it could keep a 100% risk weight.

Where the Irish banks had IRB models, as far as I understand they were not as procyclical as those of their closest peers (some aspect of the “through-the-cycle built in).

Also on Bank of Ireland, it has a significant mortgage and business lending book in the UK as well as an international corporate lending book.

Posted by WaxLyrical | Report as abusive

Thanks for that information, WaxLyrical.

Posted by Greycap | Report as abusive

yes, its better to have stringent norms and controls.

Arvind Pereira
http://www.ArvindLeoPereira.co.nr

Posted by pereiraarvindin | Report as abusive

“In general, banks reduce their ratio of risk-weighted assets to total assets in one of three ways. They can be heavily involved in investment banking, where loans tend to be taken out in the repo market and are fully collateralized; they can be heavily invested in structured products with triple-A credit ratings; or they can be heavily exposed to OECD sovereign credits.”

What. On. Earth. Are. You. Talking. About?

You’ve never worked in a bank, I presume?

Posted by drewiepe | Report as abusive