High-Frequency Trading: A Practical Guide to Algorithmic Strategies and Trading Systems

Chapter 14. Creating and Managing Portfolios of High-Frequency Strategies

The portfolio management process allocates trading capital among the best available trading strategies. These allocation decisions are made with a two-pronged goal in mind:

1. Maximize returns on total capital deployed in the trading operation.

2. Minimize the overall risk.

High-frequency portfolio management tasks can range from instantaneous decisions to allocate capital among individual trading strategies to weekly or monthly portfolio rebalancing among groups of trading strategies. The groups of trading strategies can be formed on the basis of the methodology deployed (e.g., event arbitrage), common underlying instruments (e.g., equity strategies), trading frequency (e.g., one hour), or other common strategy factors. One investment consultant estimates that most successful funds run close to 25 trading strategies at any given time; fewer strategies provide insufficient risk diversification, and managing a greater number of strategies becomes unwieldy. Each strategy can, in turn, simultaneously trade anywhere from one to several thousands of financial securities.

This chapter reviews modern academic and practitioner approaches to high-frequency portfolio optimization. As usual, effective management begins with careful measurement of underlying performance; distributions of returns of strategies composing the overall portfolio are the key inputs into the portfolio optimization. This chapter discusses the theoretical ...