We learnt this month that the US Federal Reserve had decided not to raise the countercyclical capital buffer required of banks above its current level of zero, even though the US economy is at a cyclical peak. It also removed “qualitative” grades from its stress tests for American banks, though not for foreign ones. Finally, the Financial Stability Oversight Council, led by Steven Mnuchin, US Treasury secretary, removed the last insurer from its list of “too big to fail” institutions.

These decisions may not endanger the stability of the financial system. But they show that financial regulation is procyclical: it is loosened when it should be tightened and tightened when it should be loosened. We do, in fact, learn from history — and then we forget.

Regulation of banks has tightened since the financial crises of 2007-12. Capital and liquidity requirements are stricter, the “stress test” regime is quite demanding, and efforts have been made to end “too big to fail” by developing the idea of orderly “resolution” of large and complex financial institutions. Daniel Tarullo, the Fed governor in charge of financial regulation until early 2017, recently noted that “the aggregate risk-weighted common equity ratio of the largest US banks increased from about 7 per cent in the years preceding the financial crisis to about 13 per cent as of the end of 2017”. (See charts.)