Let us just visualize the cumulative return associated with these 208 anomaly-based long-short strategies. This may give us an inspiring insight on how the publication of anomalies generate impacts on the cumulative return of these strategies.

There are totally 208 anomalies documented in Chen and Zimmermann (2020). Here I just calculate the cumulative return for each anomaly and categorize every 5 anomalies into one group. That is, in the following figure there are 5 cumulative return plots in each panel respectively. The vertical line refers to when the intervention (publication of specific anomaly for this setting) was in place. For each time-series plot of the cumulative return associated with a specific anomaly, the solid part corresponds to the cumulative return during period before the intervention was in place while the dashed part corresponds to the realized cumulative return of long-short portfolio constructed from that anomaly (characteristic) during period after intervention was in place.

It would also be useful to compare the average returns of different anomaly-based portfolios before and after the publication of specific anomaly. It is visually clear from the following plot that generally publication effect is there and the ex-post average returns of anomaly-based portfolios are much less than the corresponding ex-ante average returns. This data-driven observation also implies that publication of anomaly would potentially serve as good alternative for the treatment.

References

Chen, Andrew Y., and Tom Zimmermann. 2020. “Open Source Cross-Sectional Asset Pricing.”