Filter for Change Point Detection in Stock Prices for Investment Optimization
Financial models need specification of model coefficients in problems of portfolio optimization, pricing, hedging but these coefficients are not known fully as is the case with prices being given by equivalent market measures instead of real world probability measure. Stochastic filtering is used to infer estimates and in particular nonlinear filtering is needed for estimating purposes. The filtered estimates are crucial ingredient for decision making in investment problems. Partial Differential Equation based filtering is versatile in tackling non-linear, non-Gaussian models and due to past rich and extensive research on finite differences and finite element methods in stochastic calculus combined with parallel computing has made it very suitable for real time applications in financial industry. This paper applies these concepts with an implementation of change point detection problem for stock prices through Partial Differential Equation based filter for investment decision for an agent trying to smoothen his income flow over some time period.
Mathematical Finance Optimal Control Stochastic Filtering Portfolio Optimization
Deepak Kumar
Economics Indian Institute of Technology Kanpur India Kanpur, India
国际会议
上海
英文
612-616
2011-03-11（万方平台首次上网日期，不代表论文的发表时间）