Using High-Frequency Factor Short Signals to Improve Long-Only Stock Selection
Summary
The article argues that a factor can be useful even when its strongest information lies in identifying poor-performing stocks. It describes ranking stocks into daily groups by factor value and interpreting unusually strong performance in the worst-ranked group as evidence that the factor can help screen out likely laggards. This offers a way to apply short-side information in a market where precise stock shorting is difficult.
The examples compare a single-factor strategy, a combination of positively oriented factors, exclusion of stocks flagged by a weak-side factor, and score adjustments that penalize flagged stocks while retaining broad coverage. The reported backtests show improved performance in the examples, but the article cautions that results depend on factor choice and threshold tuning; adding factors can also reduce performance. These are selected historical tests, not evidence of out-of-sample robustness, and the article warns against using its sample strategies directly in live trading.
Key ideas
- A factor's ability to identify poor performers can be valuable even if its long-side ranking is weak.
- Daily cross-sectional ranking can reveal whether a factor distinguishes the weakest-return stocks.
- A strategy can use weak-side information by excluding flagged stocks or penalizing their composite scores.
- Factor thresholds should be tested incrementally because different factors can behave differently.
- Selected backtest improvements do not establish future performance or live-trading suitability.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.