报告题目：基于多步凸优化的高维稀疏指数追踪

报告时间：2024-06-21   10:00-11:30

报  告 人：舒连杰  教授 (澳门大学)

报告地点：腾讯会议 991-493-307

报告摘要: For financial index tracking, a sparse tracking portfolio with only a small number of assets is often desirable in practice to avoid small and illiquid positions and large transaction costs. Owing to its computational efficiency and variable selection properties, the regularization technique based on the least absolute shrinkage and selection operator (LASSO) has been widely discussed for high-dimensional sparse index tracking. However, a relatively large regularization parameter needs to be selected in LASSO to generate a very sparse solution. This could lead to relatively large biases in the estimates of the tracking portfolio weights, which would in turn deteriorate the out-of-sample tracking performance. Although non-convex penalties could be used to improve the bias issue of LASSO penalty, the resulting problem is non-convex optimization, which is often difficult and computationally intensive. Aimed at countervailing bias while preserving computational efficiency, this paper proposes a multi-step convex optimization approach based on the multi-step weighted LASSO (MSW-LASSO) for sparse index tracking. The proposed approach is very general and nests the traditional LASSO and its variations as special cases. Empirical results show that the proposed method can achieve smaller out-of-sample tracking errors than the classical approaches based on LASSO regularization.