Statistics Seminar（2016-13）

Topic:Optimal Credit Investment with Borrowing Costs

Speaker:Lijun Bo, University of Science and Technology of China

Time:Wednesday, 8 June, 14:00-15:00

Place:Room 115, Guanghua Building 1

Abstract:

We consider the portfolio decision problem of a risky investor. The investor borrows at a rate higher than his lending rate, and invests in a risky bond whose market price is correlated with the credit quality of the investor. By viewing the concave drift of the wealth process as a continuous function of the admissible control, we characterize the optimal strategy in terms of a relation between a critical borrowing threshold and solutions of a system of first order conditions. We analyze the nonlinear dynamic programming equation and prove singular growth of its coecients. Using a truncation technique relying on the locally Lipschitz-continuity of the optimal strategy, we remove the singularity and show existence and uniqueness of a global regular solution. Our explicit characterization of the strategy has direct financial implications: it indicates that the investor switches from overinvesting in the bond when his borrowing costs are low and the bond sufficiently safe to underinvesting or short-selling it when his financing costs are high or the bond very risky.

Introduction:

Dr. Lijun Bo is a Professor in the Department of Probability and Statistics at School of Mathematical Sciences, University of Science and Technology of China (USTC). His research focuses on theory of stochastic analysis and application to mathematical finance. For the stochastic analysis, he mainly worked on reflected stochastic differential equations and stochastic partial differentiable equations arising from physics and chemistry (e.g. Cahn-Hilliard SPDE) since he was pursing my master degree of probability in Nankai University. Now he is working on mathemtical finance, in particular it includes credit risk modeling and valuation, systemic risk and continuous-time portfolio optimization.

//staff.ustc.edu.cn/~lijunbo/

Your participation is warmly welcomed!