Another point that stands out [in a review of the empirical evidence] is the high degree of stability of free banking systems. Overissues of currency were corrected rapidly by the banks’ clearing system operating on the basis of the commodity standard (White, 1984; Selgin, 1988). There is evidence that interest rates were more stable under free banking, and the less regulated banking systems of Scotland and Canada were apparently better able to weather shocks than their more regulated counterparts in England and the USA. There is also evidence that shocks tended to originate in the more regulated systems and spread to the less regulated ones, rather than the other way round. Banks typically observed high capital ratios by modern standards and, in the absence of ‘official’ regulation, bank management was monitored by the market. The evidence seems to confirm that banks that were not considered sufficiently sound would lose their market share, and competition for market share would force banks to maintain the margins of safety and soundness that their customers desired (see, for example, Kaufman, 1988). But perhaps the most striking evidence is that, in the absence of any ‘official’ lender of last resort or deposit insurance, banks only rarely faced runs, and the runs that did occur were typically confined to small numbers of banks, and often only a single bank, which had suffered some shock (such as the revelation of loan losses) which led their customers to doubt their soundness.