Why do Financial Intermediaries Exist?

Economic models assume them away

The phenomenon of financial intermediation is another puzzle for economists. Obviously, corporations in general, banks in particular, and financial innovation are all very important. But the natural inclination of economists is to assume them away.

Economists often think in terms of simple models. Treating individuals as all having the same information makes a model easier to work with.

For example, take the paper published in 1958 by Franco Modigliani and Merton Miller, which earned them a Nobel Prize. It said that the capital structure of a firm will not affect its valuation. A firm that issues a lot of debt (called financial leverage) will not raise or lower its stock price relative to issuing little or no debt.

One way to give the intuition behind M-M is that personal leverage substitutes for firm leverage. Suppose that the firm’s project is to build a data center at a cost of $2 million, and it funds this by issuing $1 million in stock and $1 million in debt. My personal leverage depends on how much debt I carry. If I want more leverage from owning some of the stock, then I can go deeper into debt to borrow money to finance part of the stock purchase. If I want less leverage from owning the stock, I can reduce my personal debt.

Another way to give the intuition is to say that investors “see through” the firm. That is, they see that the firm is building a data center, and the risks and rewards of that data center are present regardless of the ratio of debt/equity used to finance the data center. The firm’s capital structure does not matter.

In a world where M-M holds, what determines the firm’s capital structure? One answer to which economists gravitated is that there is a trade-off between tax costs and bankruptcy costs. Under corporate tax laws in the United States, debt finance is tax-advantaged. But heavy reliance on debt finance increases the risk of bankruptcy, and the bankruptcy process imposes legal costs. The optimal strategy is to take on debt to the point where the marginal benefit of reducing tax liability is equal to the marginal cost of increased risk of bankruptcy.

But banks and other financial intermediaries have existed at times and places that did not have a corporate income tax. So this standard explanation for capital structure cannot be the answer.

A better perspective is to question other assumptions about the M-M world. Because in an M-M world, there is no particular reason for financial intermediaries, including firms and banks, to exist. If investors can “see through” to the underlying projects (such as data centers), then they can invest directly in those projects without requiring any intermediation. I think that the reason that we have firms and banks is that most people cannot see through to the underlying projects but still want to earn a return on their savings.

What if individuals do not all have visibility into the risks of investment projects? In that case, I argue that firms and banks can add value as financial intermediaries. Individuals want to hold short-term, riskless assets. A firm that funds a data center can issue bonds that will pay off in most circumstances. This concentrates the risks of the data center project in the hands of shareholders, allowing some of the investment in the project to involve lower-risk bonds.

A bank can buy the bonds of the data center firm as well as debt from other firms. By owning debt with different maturities, the bank can issue short-term deposits to individuals and be able to handle occasional withdrawals by depositors.

The way that I like to put it is that households want to hold short-term, riskless assets. Borrowers, like the data center builder, want to issue long-term, risky liabilities. Financial intermediaries accommodate this by taking on long-term risky assets and issuing short-term, riskless liabilities. Intermediaries achieve this by being very selective in choosing their asset portfolios, by managing assets carefully, and by diversifying their assets and liabilities. A risky, long-term project like a data center ends up financed in part with riskless, short-term bank deposits.

The financial system that has emerged in the United States is multi-layered and complex. Government intervention is widespread, including deposit insurance, implicit guarantees of various sorts, and many forms of regulation. Some financial instruments exist to shift risks to actors who are best able to manage them. Other instruments exist to shift risks to taxpayers. The latter can be termed “regulatory arbitrage,” meaning taking advantage of anomalies in government regulation.

It is difficult to build models in which economic actors differ in their access to information. But economists have to resist the temptation to simplify their models by assuming away such differences. Such assumptions lead to strong M-M conclusions that make financial intermediation superfluous. But financial intermediaries do exist, and they play key roles in the economy.