Ancora sugli stress test bancari

David Reilly sul Wall Street Journal si aggiunge al coro di quanti raccomandano un po' di prudenza: i risultati degli stress test sono molto meno rassicuranti se invece di guardare al Tier 1 ratio si calcola la leva finanziaria utilizzata dalle banche statunitensi:

Capital ratios at the heart of the tests aren't always as solid as regulators think.

One way to compensate for that is to look at a measure of leverage. This is important because capital ratios rely on risk-weighted measures of assets, while the leverage ratio uses total assets. And risk-weightings are sometimes wildly off base:(...)

When it comes to leverage, the biggest U.S. banks didn't look as robust in the stress tests. Strong banks are required to have a minimum leverage ratio of 3%. Since the crisis, the six biggest financial firms have boosted that ratio to more than 6%. The higher the ratio, the lower the level of assets to equity—and, theoretically, the risk.

But under the Federal Reserve's most stressed scenario, and taking into account capital-return requests, the firms look less robust. Citigroup, whose capital-return request was rejected, was at 2.9%. Even without a capital return, its ratio would have been just 3.2%.

Morgan Stanley, which didn't ask to return capital, was at 3.4%, while Goldman Sachs and J.P. Morgan Chase, both given green lights for capital returns, were at 3.8%. Looked at another way, those measures equate to leverage of 29 times and 26 times capital, respectively. That is eerily reminiscent of levels seen at investment banks before the crisis.

Granted, the tests were theoretical. Even so, the Fed shouldn't forget its leverage lessons.