In response to my last post, Zvi Bodie writes to suggest that I may be simply rephrasing a point that he and Robert Merton made in a series of papers some years ago (the seminal paper is here). I don’t think so, but the suggestion brings into the foreground the question of the relationship between the money view and the finance view. For some financial intermediaries (insurance companies) the finance view is arguably most of the story most of the time, but for others (banks) the money view is often the main part of the story.
Here is how Zvi Bodie summarized his point in an email to me:
“One has to distinguish between the customers and the investors of such an institution. MM [the Modigliani Miller theorem] applies only to the investors’ capital, not to the total liabilities on the RHS [right hand side] of the balance sheet. While it is true that depositors are providing funds to a bank, they are not investing in it. Policyholders of an insurance company are providing funds to the insurance company, but they are not investors. The required capital is the amount of total investor capital (debt + equity) needed to make the customer claims secure. It could be 0 for the case of a matched book narrow bank, or for the case of a mutual fund where the customers are also the investors.”
Bodie is thinking about the financial intermediary as a firm holding a portfolio of assets with a given risk-return profile, and issuing liabilities that slice that risk-return profile into various tranches. “Customers” hold liabilities that are sheltered from the solvency risk of the firm itself; conceptually, that risk is all shifted to the liabilities held by “Investors”.
In this way of thinking, problems potentially arise whenever the government guarantees the liabilities held by Customers, since this guarantee takes on the risk exposure that would otherwise be borne by the Investors. In effect the government becomes an Investor, but unlike other Investors it is not necessarily insisting on fair market compensation for the risk it is bearing. Indeed, when the guarantee is implicit rather than explicit, it might not be insisting on any compensation at all. The consequent mispricing of risk is a distortion of efficiency and also of incentives.
For the insurance industry, I think this line of analysis has a lot to offer. The question is how well it works for banks.
At first glance it seems like a straightforward extension. Bank deposits are liabilities held by Customers, while bank bonds and equity are liabilities held by Investors. Most bank deposits are insured by the government, so the government is also an Investor, and the policy question is whether the government is charging the right price, and whether its guarantee provides incentive for bank owners and managers to increase the riskiness of their asset portfolio by levering up in one way or another.
But that is not the only way to think about banks. Bank deposits are not merely a secure store of wealth, sheltered from possible insolvency of the bank. They are also a liquid means of payment, sheltered from possible illiquidity of the bank.
When and why is this shift of focus important?
When I buy a house, I do so by borrowing from a bank. When my counterparty sells the house, she does so by accepting as payment a bank deposit, which is to say by lending to a bank. From the bank’s point of view, there are two risks involved in standing between me and her. First, I may default on my loan; this is solvency risk. Second, she may spend her deposit or otherwise transfer it to another bank; this is liquidity risk.
The point is this. Even supposing that the bank is completely successful in shifting solvency risk away from Customers onto Investors, that leaves liquidity risk that cannot be shifted to Investors. Let’s be clear about this. By assumption, investors hold claims that are not payable on demand, so they are bearing liquidity risk and presumably being compensated for doing so. (Concretely, if Investors need to make a payment, they have first to sell their bank bonds and equity for the going price, whatever that may be.) But this bearing of liquidity risk by Investors does precisely nothing to ensure that Customer claims are payable on demand.
This is where the money view comes into its own. The way that banks ensure that Customer claims are payable on demand is by holding some of their assets in liquid form (market liquidity), and by maintaining borrowing and lending relations with other banks in the wholesale money market (funding liquidity). The important backstop for both is not the FDIC (deposit insurance) but the Fed (discount window).
Banks, and other dealers, make money by making markets, which involves bearing liquidity risk. That is what they are selling, and what their Customers are buying. Abstracting from liquidity, as is standard in the corporate finance literature that stems from Modigliani-Miller, including Merton-Bodie, means in effect abstracting from banks, or at least from the dimension of their business that the money view is all about. As we think through the problems of financial reform, the finance view is not going to be enough. New economic thinking is also needed.
Merton, Robert C., and Z. Bodie. “On the Management of Financial Guarantees.” Financial Management 21 (winter 1992): 87-109. http://www.jstor.org/stable/3665843