The proposal to floor the outputs of internal rating based approaches to bank capital in the latest Basle implementation standards has attracted a lot of controversy. But if industry loses the argument, what are the implications?
The use of internal risk models by banks to apportion capital is, like securitisation, a great idea that has been abused in some cases usually due to poor supervision. Like securitisation, the attempts to address the abuse have resulted in a massive and ultimately counter-productive regulatory response which has emasculated a very valuable tool. The latest proposal – that there should be a limit on the possible difference between capital according to the standard model and capital according to internal models – has naturally attracted a lot of criticism from banks who have invested heavily in the internal models. Many of the arguments reappraise those that we heard about the leverage ratio – another way to remove risk-based factors from capital optimisation.
So, if the capital reward for being prudent with your loan book is less than it was, the obvious thing to do is to take more risks. With asset values everywhere feeling all ‘toppy’, that strategy will have the added advantage of letting the banking sector say ‘I told you so’ to the Basle committee when the inevitable crash happens. Your shareholders might not be so enthusiastic about this approach.
But – facetiousness aside – there will certainly be a reappraisal of asset risk allocations as a result of the change. But what else?
It is always good to turn a necessity into a virtue. The regulator might reduce your ability to rely on economic models of capital but the rating agencies will continue to use them. If you need to hold more capital, yes it is expensive, but at least you should get a higher credit rating as a result. Sadly, the current distortion of credit spreads means that there is little economic benefit to being AA rated, rather than A rated in the bond market. Although there is some to be had elsewhere; lower collateral requirements at clearing houses, for example.
Securitisations may have been de-clawed by the regulators – in particular the requirement to retain the first 5% of loss in a portfolio – but the reason that they were invented in the first place was to arbitrage the difference between regulatory and economic capital. I’ve heard it argued that one of the reasons for the lack of a recovery in that market was the increased risk sensitivity of regulatory capital. If we take a step back from that by flooring the output of economic models, there is more incentive to reconsider securitisation as a capital management tool again.
Then there is pillar 3. A welcome development of the grumpiness about the output floor could be greater disclosure. Don’t just say that the rules are dumb; quantify exactly how dumb they are and let the market decide for itself. I suspect that no bank will stop calculating capital ratios without the application of the floor – if nothing else it makes an impressive page on an investor relations presentation (“you think we’ve got a lot of capital, just see how much we’d have if we didn’t have these dumb rules…”).
Another, admittedly rather random, implication of the output floor is that the cost of switching from an internal capital model to a standard capital model will be less. So what, I hear you say? There are bank mergers that haven’t worked in the past because the potential acquiree would have had to switch to the standard approach to capital allocation used by the acquiror, with a vast capital hit as an implication. With less of a capital hit, more acquisitions might make sense.
A final implication of the proposal worth highlighting – as regular readers will know, it always comes back to covered bonds – is the relative benefits of the covered bond and national mortgage agency models. Political conspiracy theorists amongst us will have realised that the output floor proposal is a far greater problem in Europe than America, due to the much higher proportion of very low risk assets that American banks transfer to the government sponsored agencies there – Fannie Mae and her friends will not have to allocate more capital as a result of the proposal. The delicate balance between covered bonds and agencies could be upset by higher capital demands for the assets if they are kept in covered bond portfolios.
By Richard Kemmish