The Baseline Scenario

Why Is Finance So Big?

By James Kwak

This is a chart from “The Quiet Coup,” an article that we wrote for The Atlantic three years ago next month. Many people have noted that the financial sector has been getting bigger over the past thirty years, whether you look at its share of GDP or of profits.

The common defense of the financial sector is that this is a good thing: if finance is becoming a larger part of the economy, that’s because the rest of the economy is demanding financial services, and hence growth in finance helps overall economic growth. But is that true?

Thomas Philippon is trying to figure this out. In an earlier paper, he looked at demand for financial services from the corporate sector and concluded that growth in the financial sector from 2001 could not be attributed to increasing demand. In a recent working paper, “Has the U.S. Finance Industry Become Less Efficient?“* he now measures the finance share of GDP against the total production of financial intermediation services by the sector. For the latter, he uses time series of corporate debt, corporate equity, household borrowing, deposits, and government debt. The conclusion is that the per-unit cost of financial intermediation has been going up for the past few decades: that is, the financial sector is becoming less efficient rather than more, and that accounts for two percentage points of finance’s share of the economy.

Theoretically, finance could be providing other benefits that aren’t captured in the output series, such as improved pricing or better risk management. Philippon does provide some evidence to be skeptical of those claims.

The main reason why finance’s share of GDP has outstripped its production of intermediation services, according to Philippon, is a huge increase in trading volumes in recent years. Trading, of course, generates fees for financial institutions, with limited marginal social benefits. Yes, we need some trading to have price discovery. But if I sell you a share of Apple on top of the other 33 million shares that were traded today, is that really helping determine what the price of Apple should be? The more that financial institutions can convince us to trade securities, the larger their share of the economy, whether or not that activity improves financial intermediation.

There are still a lot of open issues, but Philippon’s approach—measuring what the industry actually does—seems to be a useful one.

* The paper first debuted in the blogosphere back in December. Sorry, I’m behind.