This project continues to pursue a very productive line of theoretical and empirical research on the role of banks in the economy. This work is extremely important, as it is relevant for monetary and central banking policy, for banking and financial regulation, and for work that attempts to maintain the stability of the financial system. The project takes up the question what fundamentally makes a bank a bank. The traditional definition is that a commercial bank is an institution that simultaneously engages in two distinct activities: deposit-taking and lending. But why are these two activities necessarily conducted under the roof of a single entity? What is the underlying economic synergy? The answer is rooted in the same sort of credit-market imperfections that play a key role in the earlier work. In particular, banks are all about providing liquidity to agents on demand. This is true of both their customers on the liability side-their depositors-as well as their customers on the asset side-who are often borrowing via credit lines that can also be taken down on demand. In a world where external finance is not easily raised on short notice, the capacity to service these kinds of customers will turn on the bank's ability to hold a buffer stock of highly liquid assets in a cost-effective fashion. Our basic argument is that by engaging in both deposit-taking and commitment-based lending at the same time, a bank can spread any deadweight costs associated with its securities holdings over multiple businesses, and can thus be a more efficient provider of liquidity.
Another project that involves a similar theme asks why liquidity premia in the Treasury bond market-defined, e.g., as the difference in price or yield between "on-the-run" and "off-the-run" bonds with the same cashflows-vary quite dramatically over time. To understand variations in the market price of liquidity, one has to look to the demand side of the equation, which is embodied by those financial intermediaries who are most concerned with holding very marketable securities. Banks are one obvious example of such a class of intermediary, for the reasons outlined above, but so are open-ended mutual funds, who face large and unpredictable redemptions and hence must also hold liquid assets for similar reasons. The project's long-run goal is to build and empirically test corporate-finance-based models of such intermediaries that can speak to the whole issue of time-variation in the general equilibrium price of liquidity.
