European banks saw their risk weighted assets leap at the turn of the year, as new trading book rules collided with the European Banking Authority's call to achieve a 9% capital minimum.
HSBC was the only institution of the 10 to detail Basel 2.5’s impact in terms of its various components - the stressed VAR figure caused the biggest jump, adding $14.4 billion in RWAs.
Given the scope for modelling choice built into the trading book regime, it is no surprise that its impact differs from firm to firm - subject to the consent of their regulator, banks can choose the 12-month period used for stressed VAR, the look-back period for historical VAR, and how they deal with a lack of data for any particular instrument or portfolio.
As an example, Credit Suisse uses two years of historical data when modelling VAR, whereas UBS has decided to use five years of data. According to Darryll Hendricks, head of strategy for the UBS Investment Bank in New York, this could have a significant impact.
“Because the five-year approach includes 2008 and 2009, for many portfolios today it would give a bigger RWA number for normal VAR than the two-year approach. Depending on the portfolio, especially if it was substantially long credit exposure, I would not be surprised if the difference was 20% or possibly higher,” he says.
But Hendricks is comfortable with this disparity. “I don’t think there is any right or wrong answer on whether to use two years or five. There is a legitimate desire at many banks to use a measure that is more responsive to recent volatility. Regulators have to balance the desire for consistency and comparability with the desire for VAR to be linked to how the firm itself wants to measure its own risk internally, which is not a new issue in this context,” he says.