Basel’s trading book theory struggles with the practice

What is happening?

At the end of October 2013, the Basel Committee on Banking Supervision (BCBS) published its second consultation paper on the fundamental review of the trading book, fleshing out an earlier, less detailed first draft in May 2012. Industry representatives met with BCBS members to discuss the proposals in December 2013. The consultation period ends on January 31, 2014. The BCBS hopes to finalise the rule in time for approval by the G20 summit in November 2014.

What do the regulators say?

In the words of the BCBS, during the financial crisis of 2008, “the level of capital required against trading book exposures proved insufficient to absorb losses”. This led to a revision of capital requirements under the interim agreement dubbed Basel 2.5, but the committee wanted a more comprehensive overhaul as part of Basel III. The second paper also constitutes an “increased focus on achieving a regulatory framework that can be implemented consistently by supervisors across jurisdictions”.

What are the main innovations?

Value at risk, a market risk methodology discredited during the financial crisis because of its failure to incorporate the losses from extreme ‘tail-risk’ events, is replaced with the expected shortfall measure, which includes the tail risks. The second major innovation is the introduction of clearer boundaries between the trading book and the banking book. This is to prevent banks from moving more volatile trading assets onto the banking book, where they would have less impact on the profit-and-loss accounts.

The first consultation paper proposed an evidence-based boundary, requiring banks to prove that assets belonged in the chosen book. The industry lobbied for a boundary based on stated intent and the BCBS settled on a compromise between the two.

Finally, the BCBS is pursuing the idea of liquidity horizons, where banks will have to model larger shocks for less liquid assets that would take longer to sell or hedge during an episode of market stress. But this innovation could conflict with the actual risk implications of hedging or correlation trades.

“A bank might hedge the interest rate risk on a less liquid bond that has a three-month liquidity horizon with a liquid interest rate swap that has a one-month horizon, leading to a gap in the model outputs, even if the hedge is a good one in risk terms. And the process of measuring correlation risks between assets with different liquidity horizons could also lose economic meaning,” says Lars Popken, head of market risk methodology at Deutsche Bank.

What do the banks say?

Fears over liquidity horizons reflect a more general concern that the proposals are increasingly divorced from the reality of running a trading desk. “The paper introduces new requirements that are based on theory and not on practice. The ‘use test’ of whether these are models the banks would use to run a business day to day has been replaced by an approach of laying down normative regulatory capital rules that focus on severe stress,” says Mark White, senior vice-president for capital management and optimisation at the Bank of Montreal. Mr White was a chair of the BCBS risk management and modelling group during his time as Canada’s assistant superintendent for financial institutions until 2011.

There is particular dismay about the suggestion that BCBS might use its revised standardised approach as a floor for market risk calculations produced by banks’ own internal models. The new standardised model is based on decomposing the present values of trading book cash flows. Jouni Aaltonen, a director of prudential regulation at the Association for Financial Markets in Europe, says this technique is ill-suited to traded instruments.

“The standardised approach penalises matched-book activities. This is because the market risk capital attributed to a portfolio of perfectly offsetting transactions will not be zero and it will increase with the size of the portfolio, even though its market risk is zero,” he says. He warns that this drawback makes the standardised approach inappropriate as a floor for internal models because “it does not provide the right incentives for continuous improvement of risk models”.

What is the alternative?

Mr Popken says there are other means to improve the comparability of internal model results between banks without requiring a full new capital regime such as the fundamental review. For example, regulators could set the same time period for calculating the volatility assumptions used in models.

“The advantage of keeping risk and capital management connected is that the model is constantly improved by the discussion between capital managers and traders whose risk limits are set using the same model. There is an argument for keeping that incentive to refine the model,” says Mr Popken.