A Taxonomy of Anomalies and Their Trading Costs
Read the paperopens doi.org in a new tab
What they found
The authors estimated realistic transaction costs for 23 well-known stock anomalies using an effective-spread measure and asked which survive. Low-turnover anomalies (value, size, profitability) kept most of their returns after costs; mid-turnover ones (momentum) were still profitable but with much smaller margins; high-turnover anomalies (short-term reversal, industry momentum) had gross profits that were entirely consumed by costs. They also showed that simple cost-mitigation techniques, such as trading only when a stock moves far enough into the portfolio and holding positions longer, could roughly halve costs.
What you can use
- The higher a strategy's turnover, the more of its paper return is fiction; short-term reversal strategies are essentially untradable after costs.
- Momentum survives costs but with a thin margin; implementation details decide whether it is profitable for you.
- Buffer rules (wait for a stock to move well into the buy zone before trading, and hold it until it moves well out) cut costs substantially without much loss of signal.
Caveats
Costs are estimated for institutional-scale trading in U.S. stocks; retail costs differ (smaller size, but wider effective spreads via payment for order flow). Sample 1963 to 2013. Free NBER version exists.
Tags: backtesting, transaction-costs, anomalies, turnover
Summaries are our own reading of the paper, not the authors' words. Educational only, not advice. Discuss it in Book Club.