Quantitative Trading Methods: Factors, Arbitrage, Execution, and Risk
Summary
These reading notes survey quantitative investing methods across stock selection, market timing, futures and options arbitrage, statistical arbitrage, algorithmic execution, and asset allocation. They outline factor selection, style and sector rotation, money-flow signals, momentum and reversal, trend following, and analyst expectations as ways to form equity views. They also describe spread-based strategies, including index futures basis trades, calendar spreads, related-asset pairs, and option combinations. The notes emphasize that statistical arbitrage depends on historical relationships continuing, and that even hedged trades retain risk until positions are closed.
The theory sections introduce machine learning, data mining, wavelets, support vector machines, fractal measures, and stochastic processes, with examples of their proposed use in classification, signal processing, and market prediction. Execution methods are grouped into passive, active, and hybrid approaches, with VWAP and TWAP presented as established passive examples. The notes also discuss portfolio construction and risk control, but much of the material is a high-level book summary rather than a reproducible strategy specification. It supplies few independent performance details, and its claims about predictive power or returns should not be treated as validated evidence.
Key ideas
- Multi-factor selection ranks securities using measurable characteristics linked to expected returns.
- Momentum and reversal strategies take opposing positions depending on whether strength is expected to persist or fade.
- Statistical arbitrage trades deviations from an estimated long-run relationship, while relying on that relationship to persist.
- Execution algorithms can follow a fixed schedule or adapt timing and order size to market conditions.
- Risk, trading costs, and position management affect whether a forecast can translate into a viable portfolio.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.