The FSA’s voluminous banking liquidity “consultation paper” is available here.
It contains, among other things, a whole raft of new standards for UK banks to conform to. Central to those are what appear to be rigorous new rules over banks internal modelling of liquidity risk. There are literally pages of new modelling requirements. Plenty to keep quants busy for weeks, months if not years, as soon as compliance have got their heads around it all. Don’t the FSA know it:
Our proposals are far-reaching and robust; many institutions will need to significantly reshape their business model over the next few years as a result. Current agreements and practices will have to be reviewed and the status quo may no longer be acceptable.
The wrinkle though is that models are not enough. The FSA doesn’t want banks to be reliant on any one quantitative ratio (“as long as the ratio was met, liquidity risk was seen to be sufficiently mitigated”). Instead it wants a more dynamic approach. Chapter 4 gives the multivariate rundown on the new liquidity order.
The bottom line though is that extreme liquidity events are not extreme. And that banks need to have a grasp on exactly where all of their liabilities, err, lie.
While all of that poses a technical challenge for banks, the one part of the FSA’s paper that’s been getting the most press, requires something of a strategic shift. Chapter 6 – the shortest of them all – proposes an (increased) “liquid assets buffer”:
The assets in the buffer should be the most liquid by virtue of the significant depth and resilience in stressed conditions of the established markets in which they are traded. We consider these to be:
• highly liquid, high-quality government debt instruments as follows: gilts, plus
bonds rated at least Aa3 issued by the countries of the European Economic Area
(EEA), Canada, Japan, Switzerland and the United States; and
• reserves held with the Bank of England’s reserve scheme and with the central
banks of the US, the EEA, Switzerland, Canada and Japan.
The FSA’s new bank bondage laws. As Friday’s FT reports:
Britain’s banks face being forced to buy hundreds of billions of pounds in government bonds under proposed rules designed to make them less vulnerable to market shocks.
The FSA assumed banks would be required to hold 6-10 per cent of assets in government bonds. Among Britain’s 10 largest banks, the average is 5 per cent.
The FSA calculates the total cost of the move at £1.3bn-£5.3bn (€1.5bn-€6.1bn), implying that banks will be forced to switch £87bn to £353bn of their assets into government bonds. The regulator’s calculations assumed the shift would cost banks 150 basis points a year in lost revenues, because the government bonds they will be forced to hold will have lower yields than their current fixed-income instruments.
The conspiratorially-minded of you might see an attempt to bolster an already oversupplied market for gilts.
But we’re slightly sceptical for other reasons – as to the efficacy of the whole scheme as a liquidity buffer full stop.
* * *
The FSA’s proposal involved bolstering both bank reserves and bank holdings of gilts.
Reserves, of course, are reserves. Their levels can be altered through interbank lending. But holdings of liquid government bonds are less static. They are, of course, prime repo collateral. Which is certainly useful in a liquidity crisis, but just how useful. We reprise the below from a Bank of America report on systemic risks from a week ago or so. BoA’s lead credit analyst Jeffrey Rosenberg has compiled a table of some big US brokerage’s repo ops in relation to their financing (click to view full size):
As is clear, the banks were huge users of repos to meet their liquidity needs. Bear Stearns was repo-ing $74bn net – equivalent to 19 per cent of its liabilities. Simply increasing the number of easily repo-able securities on banks balance sheets from 5 per cent to 6 – or even 10 – per cent, is not always going to be enough.
To wit: Bear didn’t fall because investors were worried it didn’t have enough assets to repo (19 per cent!), it failed because of counterparty risk fears stoked by the shear amount Bear was repo-ing.
Banks could be forced to have 20 per cent of their assets in liquid government bonds, but if a liquidity crisis hits, that would simply transmute itself into a fear that the bank was 20 per cent dependent on short term repo financing.
The counter argument, we suppose, is that even with that being the case, better it be the repo market banks are dependent on, than the even more sensitive commercial paper market – over dependence on which was the downfall of such institutions as Northern Rock.
There is another – broader – potential issue with the scheme. Rather than increasing conservatism at banks, increasing holdings of gilts, and consequently putting greater squeeze on banks’ margins will surely only intensify the search for yield that is the underlying force behind this crisis in the first place. Assuming, of course, these banks remain shareholder-led institutions (shareholders in the non-HM government sense) then its hard to see how they wont be incentivised to put what remains of their capital to work in higher-yielding, riskier, ways.
If you also add to that, a “buffer” mentality of being newly emboldened with a superior liquidity provision (which is actually, as above, nothing of the sort) and surely you have a recipe, in the medium or long term at least, for hubris and failure all over again.