image missing on blob storage!!!

Ring fencing

23 February 2018
Richard Kemmish

Is the ringfencing of UK banks a good thing for covered bond creditors? Yes, but with some fairly significant ‘buts’.

UK banks are to be divided by a big ring fence between the risky stuff and the cosy stuff. Its an age old idea – going back to the Glass-Steagall Act in 1933 in America. Why sully the low risk, important retail banking operations with the high risks of the investment banking casino? Its one which we are bound to see more of in more jurisdictions soon.  In a somewhat unsurprising commentary piece Moody’s have said that the covered bond programmes will be in the cosy bit (along with all of the residential mortgages) and that this is A Good Thing for the credit of UK covered bonds. Not really much of a surprise that, we all think it.

Is there a counter-argument? 
A traditional argument would be that supervisory rules, in particular capital, will mitigate differences in risk between financial institutions: riskier business model, more risk capital…safe assets:low capital. So the riskiness of financial institutions tends to equalise if supervisors are doing their jobs right.

This idea is slightly undermined by both leverage ratios and the proposed floors on asset risk weights under Basle 4 both of which loosen the relationship between capital and credit risk, both of which imply that more capital needs to be held against low risk assets like residential mortgages than a strict understanding of their credit risk would imply.

Safe banks go wrong too
The fact that some of the highest profile failures of the financial crisis were banks with low average risk weights tends to support the new regulations – low risk assets do not necessarily mean low risk banks. Bradford and Bingley and Northern Rock (to take two British examples in tune with the British theme but yes, plenty of banks in plenty of other countries had a similar story) both had mainly safe therefore lowly risk weighted residential mortgages on their balance sheets. The failure of both organisations was not a function of those mortgages going wrong – their risk weights turned out to be totally justified by the subsequent quite good credit performance of the assets – but of other risk factors being more important than asset quality.

If capital ratios tend to equate asset credit risk, what in the new regulatory tool box equates other risks in a bank?  And do they work when back-tested through the financial crisis?

It’s all about liquidity 
The main non-credit risk, the death-knell for the many banks with low risk assets was of course liquidity. The flipside of having residential mortgages on the asset side of your balance sheet is, more often than not, having more volatile retail deposits on the liability side.

I would argue, and have argued here before, that the new rules such as the requirements for minimum liquidity buffers and for net stable funding are fine in everything except their assumptions about the reliability of deposit funding. If anything retail deposits are getting even less sticky thanks to both technology (swipe right to start a run on the bank) and to precedent – we can’t unremember queues of depositors taking their money out of a failing bank. Ring-fenced banks with safe assets might not have quite such safe liabilities.

Another counter-argument to the ‘ring-fence is safe’ idea is – and apologies for getting a bit structured credit-y here – that there is a greater correlation between the covered bond holder’s two sources of recourse. If a safe bank is mainly funding residential mortgages then its probability of downgrade is a function of the housing market. As is the quality of the cover pool. A diverse stream of revenues from investment banking may be more volatile but it is also more independent.

Of course Moody’s are right and banks inside the ringfence are safer (they are far more likely to be bailed out by tax payers for example), but the credit considerations are a bit more nuanced than they seem at first.

But by far the biggest implication of the ring-fencing of covered bond issuers is not on credit but on issuance volumes. Retail banks are better rated and better at gathering deposits. Their need for covered bonds is far less than that of their riskier cousins.

Perhaps the biggest impact on covered bonds from the ring fencing proposals is that there will be less of them.

Event, Articles and Videos that might interest you


22 May 2018 | Richard Kemmish

Are windmills eligible assets in cover pools? It depends which country. Should they be? That’s more difficult.

Tags: Covered Bond

Another complaint about deposits

17 May 2018 | Richard Kemmish

A recent study from the Bank of Italy on social media and depositor behaviour renews my scepticism about deposits as a source of stable bank funding.

Tags: Covered Bond