PhD Seminar • Scientific Computation | Quantitative Finance • Optimal Asset Allocation during High Inflation

Please note: This PhD seminar will take place online.

Chendi Ni, PhD candidate
David R. Cheriton School of Computer Science

Supervisors: Professors Yuying Li, Peter Forsyth

In this seminar, we discuss the optimal multi-period asset allocation problem during high inflation periods. We first establish theoretical results for the stochastic optimal control problem under synthetic market data. We then take a machine learning approach by modeling the optimal controls as outputs of a neural network function of features, which avoids the curse-of-dimensionality problem in common dynamic programming approaches. The optimal strategy learned from bootstrap resampled data shows superior performance over the benchmark fixed-mix strategy over the entire investment horizon.