Factor Investing

In an AI-driven Taiwan stock market, mastering chip distribution (ownership structure) is the key to profitability. This study delves into the Share Distribution data from the TDCC, transforming 15 tiers of shareholding data into behavioral finance factors such as investor attention, opinion dispersion, and retail speculation. By utilizing Fama–MacBeth two-stage regression and the alphalens-tej quantitative tool, we precisely validate the predictive power of psychological biases on stock returns, providing investors with actionable Alpha strategies and robust risk management solutions.

Taiwan’s short-selling signals are often misleading because the market operates under a dual-track system: retail investors short stocks through margin accounts, while institutional investors use securities borrowing and lending (SBL). Only SBL-based short selling reflects informed institutional sentiment, while margin shorting introduces noise. This study isolates SBL to construct the Short Interest Ratio (SIR) and evaluates its ability to predict cross-sectional returns and reveal size-dependent patterns in informed short-selling behavior.

This study examines whether incorporating the Short Interest Ratio (SIR) can improve the performance of a 52-week high momentum strategy in Taiwan. By comparing a baseline momentum model with two SIR-enhanced versions—one using SIR as a filter and another integrating it into a composite score—we find consistent gains in returns, lower volatility, and reduced drawdowns. The results show that SIR strengthens momentum strategies by identifying stocks under institutional short-selling pressure.

This study employs TEJ’s Point-in-Time Audited Financial Database to construct a composite factor for stock selection in Taiwan’s equity market. By preserving financial data exactly as available at each historical moment, the framework avoids look-ahead bias and ensures empirical reliability. We find that higher-ranked portfolios deliver significant short-term excess returns, while predictive power weakens over longer horizons. The results highlight the practical value of Point-in-Time financial data for quantitative factor investing and underscore its role in building replicable, data-driven investment strategies.

Gabriel Watson  is a well-known American growth-momentum portfolio manager. In his early career, he worked at Morgan Stanley Investment Management and William O’Neil & Co., where he accumulated extensive market research experience. Since joining Black Rose Capital Management in 1998, Watson has broken away from the confines of traditional value investing and developed a systematic, rule-based stock-picking method centered on “revenue momentum and price strength.” He calls this framework “The Machine.” In an environment characterized by rapid information flow and swift capital rotation, this approach has enabled him to capture the powerful upward moves of leading stocks.

Factor investing targets quantifiable characteristics to improve returns. Explores common factors (macroeconomic and style types) and strategies in our guide.

In the previous study, we examined the Capital Gain Overhang (CGO) factor, designed to capture the behavioral bias known as the Disposition Effect. By measuring the gap between current market prices and investors’ average cost basis, CGO quantifies unrealized gains and losses at the market level. Empirical tests in Taiwan’s equity market confirmed that CGO is a meaningful predictor of future returns: high-CGO stocks consistently outperformed low-CGO stocks, generating significant positive alpha beyond standard Fama–French models, especially over medium- to long-term horizons.

The origins of the momentum anomaly have long been debated, with multiple competing explanations. Among them, one of the most influential behavioral interpretations attributes momentum to the Disposition Effect, a systematic bias in investor decision-making. This article focuses on the Capital Gain Overhang (CGO) factor, specifically designed to quantify this behavioral bias. Using the Taiwan equity market as a case study, we examine CGO’s predictive power as a stock selection indicator and evaluate its practical value through empirical analysis.

Building on the statistical foundation presented in Part 1, this article explores how Idiosyncratic Volatility (IVOL) can be effectively applied in investment strategy design. We present two categories of approaches: a single-factor sorting model and a set of filter-enhanced momentum strategies. Through robust backtesting across two decades of Taiwan stock market data, we demonstrate how IVOL can improve risk-adjusted performance when used as a portfolio filter—especially when combined with momentum or dividend-based signals.