Cost-Sensitive Online Window Size Selection for Portfolio Management

Battle Count: 0

Highly relevant as it directly addresses the critical problem of window size selection in rolling portfolio management while explicitly accounting for transaction costs, which is a major constraint in real-world trading.

Moderate complexity. Requires implementing convex optimization solvers (e.g., CVXPY) for the expert layer and online learning algorithms (Hedge/Fixed Share) for the aggregation layer. The theoretical tuning of parameters (learning rate, mixing parameter) based on cost rates adds some complexity.

The paper provides detailed mathematical formulations, algorithm steps (Fixed Share/Hedge), and specific experimental setups (synthetic data generation process, real data sources DJIA/S&P 500, transaction cost rates). However, no code repository is explicitly linked in the text provided.