Making data and/or programs available is prima-facie evidence of good faith.
The data policy of the journal requires replicability of studies that are published in the journal. Authors are required to make available, upon request or by posting, data snippets [examples] and programs in order to facilitate third-party replication. They are not obliged to provide end-user support for inquiries from the general public. They are required to respond to appropriate inquiries from the journal if it becomes clear that the public data and programs cannot be easily used to replicate the data.
- Open Distributable Data:
- Please include all distributable data. Note that CRSP and Compustat will allow distribution of non-systematic data sets. (That is, you cannot include all of Compustat's "total assets" variables for all firms, but they will not object to including, say, total assets of 10 firms in 20 years. CRSP will allow you to include a few stocks or the market return, but not all returns. Make sure to attribute the data to the vendor.)
- Distributable Non-Open Data:
- You can create data sets with all total assets used. You just cannot share it with researchers that do not have a full Compustat license.
- Placebo Support:
- In lieue of distributable non-open data, authors should provide just a few sample points in open format (to make it possible for a readers to check their own variables) and a placebo data set used to illustrate the functioning of the code.
- Code must be provided to take the data from the data sets to the key findings in the paper. If the data cannot be shared, see the placebo support
- Confidential Data:
- The CFR does not publish research that depends on truly confidential data that cannot be shared even in an anonymized manner.
More details can be found here.
The data policy applies even though we do not collect the data at the time of final acceptance. You must agree to provide it later upon request. Of course, it will take you three times as long then if you are not putting your data together in time.
Authors who do not want to agree to these terms are hereby asked not to submit to the CFR. Authors who cannot fulfill the requests later may see the journal retract the acceptance.
We apologize, but administrative limits prevent us from providing a systematic archive of data from prior publications. Eventually, we plan to provide this.