Obama's bank solution goes too far -- and not far enough
Last week and again Thursday, President Obama took aim at the central challenge in financial regulation: The top Wall Street behemoths have grown too big to fail. But despite two runs at the problem, Obama has yet to propose the correct solution. Last week he was too timid. Yesterday he went too far.
Obama is at least grappling with the right foe. Because the failure of a large bank can topple other ones, governments will rush to the rescue in a crisis; because the banks' creditors know this, they lend to banks too freely, creating the unstable towers of leverage that make crises likely. In effect, every large bank is like Fannie Mae and Freddie Mac, the late lamented housing lenders. Armed with an implicit government guarantee, they hoover up subsidized capital and gamble it in the markets. The profits go to the bankers and their shareholders. The losses are absorbed by taxpayers.
What would Obama do about this problem? On Thursday, he proposed a radical solution: Banks that enjoy subsidized capital should be banned from gambling. They should close down the "proprietary trading desks" that play the markets with the banks' own capital. They should not operate risk-taking subsidiaries such as hedge funds.
The hedge fund part is right. Hedge funds housed within banks tend to be riskier and worse-managed than freestanding ones because they escape the market's discipline. They raise capital on the strength of their parent bank's brand, rather than having to convince skeptical clients that they know what they are doing. They take crazy risks more readily, because they rightly suspect that if they get into trouble, the parent bank will bail them out. During the crisis, hedge fund subsidiaries of Bear Stearns and Goldman Sachs got into trouble while independent hedge funds survived relatively well.
But the proposal to shut down proprietary trading raises trickier questions. It's true that the government safety net makes it rational for banks to gamble recklessly; when you have car insurance, you drive faster. But a ban on proprietary trading would be impossible to enforce cleanly.
Banks need to take positions in the markets in the course of their normal businesses, and the line between these positions and their prop desks is impossibly fuzzy. For example, banks need to have cash on hand in case their depositors demand their money back; if they park that cash in the bond market, is that proprietary trading? Banks need to balance out the risks in their loan book; if a pharmaceutical client asks for a large loan, and the bank balances that by selling pharmaceutical stocks and buying the bonds of computer and car companies, is that dubious gambling or is it good practice? Revealingly, proprietary trading developed partly as an outgrowth of the traditional brokerage business. When clients offloaded a large block of Ford on a broker, the broker needed to protect itself against the danger that Ford would fall in value. So the broker took offsetting proprietary positions, for example by betting against the shares of Chrysler and GM.
Because proprietary trading is impossible to define clearly, an outright ban on it will spawn hundreds of compliance officers to police hopelessly vague lines. But the ban would also have a giant loophole, because it presumably would not extend to foreign banks. The foreigners would still gamble in U.S. markets, and the risk of destabilization and a job-destroying credit crunch would continue. In the good years, the foreigners would reap profits, and U.S. banks would lobby regulators to relax the definition of proprietary trading so that they could stay competitive. Moreover, when the next crisis arrived, it would not necessarily be foreign taxpayers that paid for the bailout. The risks taken by Iceland's banks might well end up being partially paid for by British taxpayers.
Given these uncertainties, the best response to the too-big-to-fail problem may be to tax size. Large banks, like polluters, create a cost that falls on the rest of society; society should charge for that. And a tax on size is precisely what the president proposed last week -- every bank with assets of more than $50 billion would pay a levy that rises in proportion to the subsidized capital raised in the markets. Unfortunately, Obama muddied his message by couching his proposal as a way to recoup the cost of the bailout. But he was effectively suggesting a too-big-to-fail tax.
Which was great, except that the proposed rate was far too modest and the recoup-the-bailout rhetoric obliged Obama to limit the levy to a few years. If Congress embraced the concept of the too-big-to-fail tax, multiplied its size and came up with a more subtle way to get at proprietary trading, financial reform would be on track.