Bank Size and the Severity of Financial Shocks

(MoneyWatch)  In his column today, Paul Krugman discusses the superior performance of Canada's banking system in recent years -- it did not experience the crisis our system just went through -- and what we can learn from Canada as we attempt to implement financial reform. In one part of the article, Krugman says:

Canada's experience also seems to refute the view, forcefully pushed by Paul Volcker ... that the roots of our crisis lay in the ... existence of banks that were "too big to fail." For in Canada essentially all the banks are too big to fail: just five banking groups dominate the financial scene.

On the other hand, Canada's experience does seem to support the views of people like Elizabeth Warren ... who place much of the blame for the crisis on failure to protect consumers from deceptive lending. Canada has an independent Financial Consumer Agency, and it has sharply restricted subprime-type lending.

I don't disagree with the main points in the article about following Canada's lead and limiting leverage, requiring originators to hold part of securitized loans, and creating a consumer finance protection agency. But I want to argue for limits on bank size as well. The political influence that large banks might wield is a reason and of itself to limit bank size, but this argument is based upon the increased vulnerability of the economy to financial shocks that comes with increased bank size.

While size/connectedness may not have been the cause of the crisis in the US -- Canada had large banks too but did not have a crisis -- bank size magnified the effects of the crisis once it began. Had Canada experienced the same shock, which was apparently prevented by its more stringent regulatory structure, bank size may well have been an issue there as well. Thus, even though bank size may not be the cause of the crisis, there may still be a reason to limit it. We cannot guarantee that regulation of the financial sector will prevent all shocks. Capping bank size can help to limit the extent to which a shock is magnified if one does occur, and thus limit its ability to do damage.

However, with that said, I've argued elsewhere that limiting leverage is the most important factor going forward, and once that is accomplished -- as it apparently was in Canada -- size may not be as much of an issue. But if there's little or no social benefit from allowing banks to grow beyond a certain size, and it's not clear that there is, and if there is a potential cost, why take a chance?