Factor Investing in Practice: Building and Testing a Quantitative Equity Strategy

Overview

An individual quantitative finance project implementing and evaluating a factor-based equity trading strategy, built entirely in Python around one of the market's most well-documented stock return predictors (size, book-to-market, profitability, or asset growth).

What I did

• •Discussed the theoretical rationale behind cross-sectional return predictability, weighing market-inefficiency explanations against risk-based explanations
• •Ran Fama-MacBeth cross-sectional regressions to test the statistical significance of the chosen characteristic in predicting stock returns, controlling for other firm characteristics individually and jointly
• •Constructed equally-weighted quintile portfolios sorted on the chosen characteristic and built a monthly long-short trading strategy exploiting its predictive power, evaluating average portfolio returns and the long-short spread's t-statistic
• •Attributed the strategy's returns to the Fama-French 3-factor and 5-factor models to test for abnormal (alpha) performance beyond standard risk exposures
• •Plotted the cumulative performance of the strategy over time and assessed the consistency of returns
• •Rebuilt the entire analysis using value-weighted (market-cap-weighted) portfolios and compared the results against the equal-weighted version
• •Extended the analysis with a dependent double sort on a second characteristic, constructing 25 portfolios and a refined long-short strategy based on the top and bottom five, comparing performance against the single-sort strategy

Skills & tools