Richard Murphy says that the Bank of England's statement (pdf) that banks create money by lending means there is a case for credit controls. I fear he's being a little hasty.

It is of course true that banks create money by lending; I'm surprised that anyone should be surprised by the Bank's statement of the obvious*. But that is not the only way in which banks affect the money stock. They also do so by issuing shares or bonds. If they issue shares, the money stock falls and if they buy shares it rises; this process is measured by net non-deposit liabilities in table A3.2 here.

And here's the thing. The problem we had in the run-up to the crisis was not simply that banks were creating loans and therefore money. It's that this money creation was not matched by issues of capital. The ratio of banks' total assets (mostly loans) to their equity rose. This suited bankers because the higher the ratio of assets to equity, the higher is return on equity in good times and so bank bosses can claim higher salaries; Anat Admati and Martin Hellwig are good on this.

It was this high asset-to-equity ratio that got banks into trouble. RBS and Northern Rock didn't fail because lots of their loans turned bad. RBS failed because it didn't have a sufficient equity cushion to absorb losses at ABN Amro. And Northern Rock failed because it had funded loans by borrowing in wholesale markets, and those markets seized up in 2007.

Quantitative credit controls wouldn't have prevented all this. At a time when banks wanted to leverage up, such controls might instead have caused them to buy back equity. This would have left them vulnerable to losses on bad takeovers or to a loss of interbank liquidity.

Instead, what we needed was better prudential regulation (or public ownership!) to ensure banks had an adequate equity base; quantitative credit limits, in themselves, do not achieve this.

I stress this point for a reason. I fear that we're in danger of misdescribing the financial crisis. It was not so much a macroeconomic failure, caused by high debt, as an organizational failure; bad management led to over-extended balance sheets, and shareholders failed to stop this. In this sense, the crisis was a failure of hierarchical capitalism and not something that could have been prevented through better macroeconomic management.

In saying I don't mean to deny that there might be a case for quantitative credit controls. If there is such a case, it lies in stories of adverse selection and asymmetric information; high interest rates mean that only reckless or confident borrowers apply for loans. In such a world, quantitative limits might be superior to interest rates.

However, such limits don't follow automatically from the Bank's description of the money creation process, and they aren't sufficient in themselves to prevent banking crises.

* Maybe because I've tended to regard textbooks as a help in passing economics exams rather than as a means of understanding the economy.