Last week, the Basel Committee published the final version of its new rules for bank trading books, laying out which assets stay in the trading book, how they will be hedged and, crucially, how much extra capital banks have to assume.
For the banks, the rules were widely seen as a good thing. Instead of increasing trade book capital needs by more than 70%, the revised rules increase capital by 'only' 40%. Securitizations and credit will still be expensive to trade but less so than under previous plans.
Vigorous discussion has already broken out over how the Committee arrived at these numbers. The banks, and the industry at large, think Basel has underestimated just how onerous its rules are. But whether the right figure is 40%, 50% or 30%, is a finger-in-the-air guess. The rules don’t officially come in until 2019 and could well be pushed back. By then the big bank restructurings under way in 2016 will have adapted in response to the rules.
But the really big change, the one which can’t be unwound, exited, or side-stepped, is the move to regulate banks trading desk by trading desk. This rule would apply to banks that want to use internal models.
That means each desk submits its own risk models for regulatory validation, and is directly responsible for its regulatory capital consumption.
If a desk head screws up a model, and loses regulatory approval for it, it directly hurts the P&L of that desk, because the regulatory capital consumed by that desk will shoot up.
When P&L wasn't P&L
Banks have made big strides in this direction already — P&L now often actually means profit and loss, rather than gross revenue, and balance sheet constraints have forced desk bosses to carefully husband their regulatory capital. Senior managers in trading-related areas complain that regulation is taking more time than ever.
But having internally allocated capital is a very different proposition to being one of the first lines of contact between regulators and banks. The desk-by-desk approach effectively separates the trading divisions of investment banks into a series of small trading units, and places far more weight on the desk heads to be responsible managers, not just senior traders.
It also runs contrary to broader trends, which seemed set to make the banks safer.
Since the crisis, banks have set up entirely new trading desks to centralise some risk functions. Counterparty credit risk became more expensive, and so did collateral, under pressure from margin requirements and the shift to secured lending.
So banks set up centralised desks to trade counterparty risk and to optimise collateral. They would trade with external clients, but a big part of their business would be making sure banks used their own resources efficiently. This meant more derivatives could be cleared more cheaply, more clients could access high grade collateral, and banks hedged more of their risk. Regulators have got to like that, right?
Banks have also been centralising their execution, risk, and reporting systems, and cutting back on the latitude which individual desks have to develop their own tech. The waves of IT spending are largely going on unpicking the idiosyncratic systems developed pre-crisis, and replacing them with integrated central ones under the watchful eyes of management. Investment bank chief executives these days literally brag to investors about how few trading systems they have.
Firms are also trying to manage trading businesses as one cohesive whole. At BNP Paribas, equities and fixed income now report through the same management. Rates and credit frequently share salespeople. The same click-to-trade portals now, in theory, allow investors to access a range of product offerings.
So the timing could not be worse. Banks are finally starting to get a handle on how to manage their trading businesses, and making all their traders pull in the same direction. It is a long, expensive and difficult road, but it’s finally happening.
Now, unfortunately, regulators are pushing them back the other way.