Online Portfolio Window Selection with Turnover Costs

Summary

This paper addresses portfolio management when market conditions change and the useful historical lookback window is not known in advance. It proposes a two-level process: construct portfolios using several candidate window sizes, then combine those portfolios online by treating each window choice as an expert. The aggregation weights are updated using losses that include portfolio turnover, so frequent changes carry an explicit cost.

The authors derive finite-horizon tracking-regret bounds that account for turnover in the combined portfolio; static regret is covered as a special case. With bounded losses and cost rates, a suitably tuned Fixed Share method achieves asymptotically no tracking regret when the switching budget grows sublinearly, while Hedge addresses the static case. The supplied description gives theoretical guarantees rather than empirical trading results, and does not specify the portfolio construction rule, market data, or practical cost calibration.

Key ideas

• Candidate historical window sizes are treated as experts whose portfolios are combined online.
• Aggregation weights respond to losses that include turnover costs.
• The analysis bounds tracking regret while accounting for the aggregated portfolio’s turnover.
• Fixed Share provides the stated guarantee for sublinear switching budgets under bounded losses and costs.
• Hedge covers the static-regret setting.

Tags

# Cost-Sensitive Online Window Size Selection for Portfolio Management


# Cost-Sensitive Online Window Size Selection for Portfolio Management









This paper investigates cost-sensitive online window size selection for portfolio management under changing market conditions. Specifically, we propose a two-level framework that constructs portfolios using candidate window sizes and dynamically aggregates them through online learning. By treating candidate window sizes as ``experts,'' we dynamically update their aggregation weights using turnover-inclusive losses. Moreover, we derive finite-horizon cost-sensitive tracking-regret bounds that account for turnover of the aggregated portfolio, with static regret as a special case. Under bounded losses and cost rates, suitably tuned Fixed Share achieves asymptotically no tracking regret for sublinear switching budgets, with Hedge covering the static case.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.