Grants

Cornell University

To develop, test, and propagate innovative Bayesian methods for estimating and regulating risks taken on by financial institutions

  • Amount $183,809
  • City Ithaca, NY
  • Investigator Nicholas Kiefer
  • Year 2009
  • Program Economics
  • Sub-program Economic Institutions, Behavior, & Performance

Nicholas Kiefer of Cornell University is developing new analytical techniques directly motivated by and applicable to the needs of effective international bank regulation. Specifically, the Basel Accords set standards for determining how much capital a bank should be required to keep in reserve to guard against financial and operating risks. There are two main approaches that statisticians use to estimate probabilities. Frequentists think of the probability of an event (say, flipping heads with a coin) as the long run proportion of repeated trials when it occurs. By contrast, Bayesian statisticians approach an estimation problem with subjective beliefs about the "prior probability" and then concentrate on systematically using whatever data becomes available to update their estimates. Through further consultations with practitioners, more research publications, continued work with graduate students, and the development of a course on Bayesian risk estimation in finance, this project stands to improve the understanding, management, and regulation of financial risk.

Back to grants database