Stats › Intraday and market microstructure › Anomalies after costs
After publication and after 2005, the average anomaly nets four basis points a month on a standard error of three
| Implementation and period | Gross | Net | Standard error |
|---|---|---|---|
| In sample, cost optimised | 66 bp | 38 bp | not printed |
| In sample, academic implementation | 66 bp | about 5 to 6 bp | not printed |
| Post publication, academic | 30 bp | -3 bp | 5 |
| Post publication, cost optimised | 20 bp | 13 bp | not printed |
| Post publication and post 2005, equal weighted | not printed | 8 bp | 4 |
| Post publication and post 2005, value weighted | not printed | 4 bp | 3 |
Monthly returns to 120 long short anomaly portfolios, net of effective spreads. The academic implementation is equal weighted quintiles, which is how most published anomalies are constructed. Cost optimised adds buy and hold bands. Price impact and short sale fees are excluded from these cost estimates, which the authors state explicitly.
Across 120 anomaly portfolios, the average gross return of 66 basis points a month becomes about five after an academic implementation's trading costs. Post publication and value weighted, the net is four basis points against a standard error of three.
Members only
The rest of this page is for members
Below this point there are 6 sections, 1 table and 2 named sources, roughly 650 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.