Effective date. This code and data sharing policy (the Policy) applies to all submissions conditionally accepted on or after October 1, 2025.
Purpose. The Review of Financial Studies (RFS) is committed to transparency and the independent replication of published findings, thereby enhancing the visibility and impact of authors’ work. Accordingly, authors of accepted papers that include empirical, experimental, or computational work must provide the data, code, and documentation (the Package) necessary for a knowledgeable researcher to independently reproduce the published results, subject to the exemptions described in this Policy.
Definition. Reproduction means obtaining the published results using the authors’ provided Package. It differs from replication, which involves re-implementing analyses with new code or new data. By documenting the steps required for reproducibility, the Package facilitates replication.
Process. Upon conditional acceptance, authors work with the RFS Data Editors (DE) to ensure the Package meets Policy requirements. Subject to computational constraints, the DE will verify that:
- the submission’s description of computational steps and empirical methodologies matches the submitted materials;
- the code runs successfully in a dedicated virtual computing environment provided by the RFS; and
- the materials reproduce the submission’s key tables and figures using the authors’ provided data or pseudo-data where direct sharing is not legally or contractually possible (see below).
The DE confirm successful reproduction in a report that will be published alongside the Package following final acceptance. For further information on the DE process, see the RFS DE website.
Standards. RFS endorses the Data and Code Availability Standard (DCAS) v1.0, and this Policy is compatible with DCAS.
Data
Data Availability. A statement covering both the source data and any derivative data must be provided in a README file following the template README in Social Sciences v1.1. This statement should contain detailed information about how, where, and under what conditions an independent researcher can reproduce the steps needed to access the original raw data (potentially including required registrations, memberships, application procedures, monetary cost, or other qualifications). This information must be provided even when all raw data are included as part of the Package.
Raw Data. Primary data collected by the author and secondary data not otherwise available must be included in the replication package unless the exceptions for non-public data apply or unless the exact extract of the raw data used in the analysis is published in a trusted public repository with a permanent identifier (e.g., DOI).
Analysis Data. Analysis data should be provided as part of the replication package unless the exceptions for non-public data apply.
Non-Public Data. If raw or analysis data cannot be shared publicly due to legal, contractual, or ethical restrictions, authors must document the restriction in the README and provide as much publicly shareable information as possible (e.g., data schema, codebooks, summary statistics).
Authors who wish to retain exclusive usage rights to certain data (e.g., hand-collected data) are encouraged to create and apply an appropriate license. If a license alone does not adequately address the desired restriction, authors must notify the editor at the time of submission and propose a withholding period.
In all cases, the authors will collaborate with the DE to verify reproduction of the published results. As an additional measure, authors are encouraged to include synthetic, anonymized, or pseudo data that enables users to execute the code.
Format. The data files may be provided in a format compatible with commonly used statistical packages or software. Authors are strongly encouraged to provide data files in open, non-proprietary formats.
Metadata. Each dataset must be accompanied by metadata describing each variable, allowed values, and their meaning in the README. It is acceptable to reference publicly available documentation for these items.
Citations. All data used in the paper are cited where available.
Code
Transformation Code. The Package must include all programs required to clean, merge, and transform raw data into analysis data.
Analysis Code. The Package must include all programs that produce the computational results reported in the submission. These programs include scripts for parameter estimation, statistical testing, simulations, numerical solutions, and the generation of tables, figures, and other published outputs.
Non-Public Code. If code or programs cannot be shared due to legal or contractual restrictions, authors must document the restriction in the README and provide a suitable alternative, such as pseudo-code, that allows other researchers to execute the programs.
Authors who wish to retain exclusive rights to code that represents a significant intellectual contribution are encouraged to create and apply an appropriate license. If a license alone does not adequately address the restriction, authors must notify the editor at the time of submission and propose a withholding period.
Dependencies. All dependencies must be clearly listed in the README, including the software, packages, libraries, and version numbers, along with instructions required to execute the scripts. Where random number generation is involved, authors must set seeds in the code to ensure reproducibility.
Format. Code must be provided in an executable source format that can be interpreted or compiled by the appropriate software.
Citations. Citation of software packages (e.g., Stata / Python / R packages) is encouraged.
Supporting Material
Instruments. When applicable, authors include survey instruments or experiment instructions, as well as details on subject selection in the Package.
Ethics. When applicable, authors include details about ethics approval in the Package.
Pre-Registration. When applicable, authors document pre-registration of the research project in the Package.
Sharing
Location. After acceptance, the authors publish the package in the RFS Dataverse, while the DE host the code and their report in a public GitHub repository within the RFS organization.
License. The Package must include a license file to permit replication by independent researchers. A default license is provided on the RFS DE website, though authors may use an alternative license subject to DE approval.
Omission. The README clearly indicates any omission of the required parts of the package due to legal requirements or other approved agreements.
Post-Publication Responsibilities. Authors are not expected to provide assistance or ongoing support to users of the Package.