Federal agencies increasingly voice a preference for using the leverage ratio and stress testing as an alternative to Basel risk-weighted capital” by Philip Alexander. You will read that I am quoted in the article as “banking industry veteran: John Perry who was global head of independent model risk review at HSBC until 2013”. Here are my comments that are included in the article"
Variations in risk weight outputs are not surprising once the underlying model is understood.
The major challenge is in the process of scaling up the bank’s expected loss calculation from PD and LGD inputs, to form a risk weight that is designed to be enough to cover unexpected losses as well.
The risk weight on an individual credit does not reflect the amount of capital needed to cover potential losses from that particular name, but is rather part of an entire portfolio calculation based on factors such as each bank’s individual loss experience on that type of asset
PDs are typically classified on a master scale from one to 20. Most credits will fall within the first six subscales of that, which means each subscale encompasses a difference in PDs of as little as 100th of a basis point from minimum (of a subscale) to maximum (of the subscale below) That means it is relatively easy for the marginal risk weight contribution on a single name to differ by as much as 50% purely due to different default histories at the two banks,” says Mr Perry.
However, he does not see this as an argument to scrap risk-based capital ratios altogether.
Mr Perry warns that while a high leverage ratio reduces the risk of bank collapses or bail-outs, it could exacerbate other systemic risks. Each bank would have less flexibility to manage its capital due to rigid leverage ratios.
At the first sign of losses, he believes, banks would be forced to deleverage sharply, accentuating the cyclical impact of the financial sector on the real economy.
Nor is Mr Perry averse to the idea of national discretion in risk weights. In his eyes, this makes more sense than using the Gordy formula that was essentially based on the experience of US and Western European bank balance sheet performance in the 1990s.
He suggests that countries should be able to calibrate their own risk weights alongside a global Basel standard, examining their own economic cycles and market trends including portfolio performances during the most recent financial crisis.
This would also provide a genuine risk-based macro-prudential tool with a stronger analytical basis than simply hiking risk weight floors.
“Regulators fear losing global uniformity, but at the heart of improving risk models is the concept of benchmarking. It is impossible to compare the efficacy of Basel models if there is only one of them,” says Mr Perry.