Empirical data analysis is a key component of financial economics. Many, if not most, of our knowledge in financial economics are based on our experiences and experiments with different data types related to financial activities we observe in the real world. Often distributional assumptions are made on the data without any valid economic reason for such assumptions. Using g-and-h family of distributions, an exploration and understanding of the data agreement with the assumptions is made. It will be shown with three different examples of data from the financial markets that the g-and-h distributions point us to different structure for the distributions from those that are currently used. Finally by back testing, it will be shown how different (and better) the risk management would have been if such assumptions, indicated by the g-and-h family, were made.
Keywords: Exploratory Data Analysis; g-and-h Distribution; Interest Rate Risk; Operational Risk
Biography: Kabir Dutta is a principal and senior consultant in the Finance practice of Charles River Associates. Dr. Dutta has more than 15 years of experience in financial risk management of market (equity, commodity, and foreign exchange), interest rate, and operational risk. He has experience across various industries such as telecommunication, energy, exploration and production, and financial services. Prior to joining CRA, Dr. Dutta was a senior economist at the Federal Reserve Bank of Boston. His research has been presented at many national and international conferences. Dr. Dutta taught at the Princeton University and holds a MBA and Ph.D. from the Wharton School of the University of Pennsylvania.