Financial Intermediaries in a Modern Economy: Risk Taking and Liquidity Provision

Stanford University, Stanford CA

Investigators

Abstract

The research team plans two projects that analyze two properties of modern banks. First, modern banks are exposed to risk not only through traditional ownership of risk assets (loans and securities) but also through their trading books, and in particular their positions from making markets in derivatives. The research team will develop a new approach to measure these risk exposures that uses regulatory data and the dynamics of interest rates. A second key property of modern banks is that they, along with other intermediaries such as financial clearinghouses, offer payment services to clients. The clients include households, firms, and institution investors such as broker-dealers and asset managers. Existing models of business cycles and asset prices do not typically include these payment services. The team will develop a model of the economy that includes such services and will use the model to quantify how inflation and asset prices respond to monetary policy and economic shocks. The research funded by this award will give policy makers and regulators new tools and models to predict the effects of economic policies. Existing approaches to the measurement of this risk using regulatory or stock market data have not been able to compare time variation in risk across positions. For example, the measures cannot be used to determine whether derivative positions hedge other business. In practice regulators therefore rely on stress tests. The new approach exploits the factor structure in interest rates to represent many bank positions (including derivatives) in terms of simple factor portfolios. A transparent algorithm delivers, for every bank position and every date, exposure numbers that are comparable across dates, banks, and positions. The second project builds a dynamic equilibrium model in which goods and services are paid for with inside money provided by banks or clearinghouses. The model jointly determines asset prices and inflation as well as the volume of payments. The model includes data on payments and bank positions and uses the data to make quantitative predictions.