Treating the fallout from financial-system fragility by buying more trust, at the cost of…more fragility?
February 5, 2013 4 Comments
Streetwise Professor took up the mantle against the push to centralize (Large) capital calculations here, and I endorse those comments.
Perfectly correlated errors are a recipe for systemic risk. All banks will crowd the risks that the model underprices and avoid the risks that it overprices. If an adverse shock hits the crowded investment, every bank is screwed simultaneously. All will need to liquidate their positions in that asset, or in other assets. Every bank will have fire sales at the same time.
Deus Ex Macchiato has responded in a way that surprised (if no one else) me:
Will this central calculation be wrong? Yes, clearly so. Will it be differently wrong for different banks? No, it will be wrong in the same way for everyone. But does that introduce an incentive for all banks to behave the same way? No, it doesn’t. This is partly because many banks don’t care about market risk RWAs (think Wells Fargo or Lloyds) and partly because even the ones that do don’t base their behaviour simply on what is cheapest in capital terms.
Perhaps I’m naive. Ignorant. Both. In fact, on second thought, scratch that ‘perhaps’. No perhaps about it. In particular, I certainly don’t claim to know what all banks in the world ‘care about’. But as firms, and (especially) as financial firms, surely somewhere in their thought process is making money. In particular, return-on-equity, or something very much like that. If a bank ‘doesn’t care’ about that, then I’m totally confused about everything and probably need to revisit the very foundations of my knowledge about everything else. (Numbers, for example. Colors. Shapes.)
Now, one way to improve ROE is higher R. That’s (I think) where the thing called ‘profit’ comes into play. And surely no one denies that firms want to obtain this ‘profit’ thing we’ve all heard so much about; it would seem to help make R bigger, all else equal. But another way to help ROE, probably, is to keep E small. And what dictates the size of E? Well, in the current regime, for the firms in question, it’s (among other things, perhaps) this “RWA” thing. As DEM says,
All I am asking is that they all, in their diverse ways, meet a simple standard: that centrally calculated market risk RWAs are less than the amount of capital they set aside for market risk.
That way of saying it seems backwards to me so let me turn it around: ‘all’ he asks is that banks set aside more capital than a centrally-calculated floor, calculated externally on their portfolio (or more precisely, and importantly: on what the government/regulator thinks is their portfolio, based on the datafeed it received and modulo its ability to process/understand that data) via a Large Calculation. Surely it follows that the higher their floor, the more capital – the more “E” – a bank will minimally have to ‘set aside’. And unless I’ve really screwed up the math, the more E set aside, the less ROE, all else equal.
And then of course there is the less-tangible but still-important (after all, it’s presumably why we’re doing this) fact that having more capital vs. your floor would get you perceived as ‘healthier’ than having less, which helps you borrow and helps your stock and all that. Either way, how on earth does that not lead to an incentive to take actions to minimize one’s floor? Which – in this case – everyone agrees means to (consciously or via trial-and-error) reverse-engineer a centrally-calculated model that is ‘wrong in the same way for everyone’ – in particular, crowding into its underpriced risks and avoiding its overpriced risks? I suppose he could be trying to pin a lot of meaning onto the word ‘strong’ – it creates an incentive, just not a ‘strong’ one? Okay, it creates an incentive. But that’s all it takes for SWP’s comments to fully apply.
If I do try to picture a bank ‘not caring’ about RWAs I can only come up with a few possibilities.
- They don’t care because they are so (legitimately) well-capitalized as a matter of policy, culture, health or whatever else that they just never threaten to come anywhere near the (proposed/anticipated) floor. (A subcategory of this is a bank that primarily sees the floor as a useful anti-competitive regulation because it mostly affects and culls weaker banks.)
- They do not perceive RWA-attracting activities per se as forming the core of their business (e.g. Wells?), hence in practice RWA does not meaningfully constrain, nor do proposals on the table (so they think, rightly or wrongly) threaten to constrain, their revenue generation.
- They are cheating on, fudging, or screwing up their internal RWA calculation with ‘clever’ and/or inept modeling, thus keeping their RWA artificially low. ‘We don’t care about RWA, we have plenty of buffer, just look at these BS numbers!’ (Of course, when this is intentional, it would be more accurate to say such a bank cares very much about their RWA.)
Assuming they’re nonempty, Categories 1 and 2, I submit, do not need their capital centrally-calculated. Nothing is gained from doing so and there is no problem that needs fixing anyway as far as they are concerned. If I am meant to be thinking about banks that (we are sure) are in Categories 1 and 2, and then someone comes up to me and suggests centralizing RWA calcs, my reaction is: Why bother? If we are not sure that a bank is in Category 1 or 2, then a centralized capital calculation might at most help tell us that it really properly belongs to Category 3 - but then of course any such banks would quickly stop ‘not caring’.
So of these, Category 3 would seem to be the category we’re interested in. And sure, by assumption, any banks in that category ‘don’t care about RWA’ right now. But of course the very point of centrally-calculating their capital would be to (try to) disallow their BS calculation keeping their RWA artificially-low. Again, I submit that this would, pretty quickly, flip them from the ‘don’t care’ to the ‘care a lot about RWA’ column. Any bank in Category 3, currently relying on fraud/incompetence to meet its well-capitalized label, would have every reason to suddenly become very interested in the Centralized Capital-Calculation Model: what makes it tick, what attracts RWA vs. what doesn’t – and to position itself accordingly.
So it’s just as SWP said then: a centralized capital model that is ‘wrong in the same way for everyone’ creates an artificial (non-economic) incentive for banks to pour into/avoid the same risks. I can’t say whether it’s a ‘strong’ incentive, by whoever’s lights, but it’s surely an incentive.
And I just can’t agree that the financial system needs any more incentives than it already has - weak, let alone strong – to behave in correlated ways and take only centrally-calculated and -endorsed risk. The purported ‘standards that create investor trust’ benefits of doing so do not outweigh the drawbacks in my mind. ABS CDOs were nothing if not a creature of centrally-endorsed risk and ‘investor trust’. There will always be such things; that cannot be avoided. What can be avoided – and wasn’t, when all the things that we’re ostensibly trying to fix went wrong - is hooking the entire fate of the financial system in lockstep onto all the same asymmetrically-understood risks. If reduced ‘investor trust’ is the price of avoiding that going forward, that seems a small price to me, if not to others. In fact, what am I saying: it’s a benefit, not a cost.