rfs-workflow

RFS Manuscript Workflow (rfs-workflow)

Overview

This is the router. It does not replace any specialized skill — it tells you which rfs- skill to use right now*.

Default assumption: unless the user says otherwise, treat the target as The Review of Financial Studies (RFS) — founded 1988 (first editor Michael Brennan), published by Oxford University Press for the Society for Financial Studies (SFS), one of the "top-3" finance journals alongside the Journal of Finance (AFA) and the Journal of Financial Economics (Elsevier). The current Executive Editor is Tarun Ramadorai (LSE / Imperial, term 2024–2027); the editor team includes Viral Acharya (NYU), Xavier Giroud (Columbia), Andrey Malenko (Boston College), Anna Pavlova (LBS), Clemens Sialm (UT Austin), David Sraer (Berkeley), and Jessica Wachter (Penn). RFS shares JF/JFE's high causal-inference and asset-pricing bar but has two distinctive levers JF/JFE do not share: (1) it is the first finance/economics journal to run Registered Reports (pre-results review; see Karolyi's 2014 editorial "Kick-Starting the Review Process," RFS 27(2)), and (2) it offers dual submission with the SFS Cavalcade conference. The governing tension at RFS is always novelty AND rigor — never one without the other.

When to trigger

• The user asks "what should I do next?"
• A draft arrives and you must diagnose the current bottleneck
• The user is thrashing between empirics, writing, and revision
• A decision letter from RFS arrives and you must switch into rebuttal mode
• The user is unsure whether the paper is novel enough OR rigorous enough for RFS

Routing table

Current symptom	Next skill
Idea feels incremental / "me-too"; unsure it clears RFS's novelty bar	rfs-topic-selection
Related-work positioning is weak; unclear delta vs. JF/JFE/RFS papers	rfs-literature-positioning
Core claim is causal but design is OLS + controls / endogeneity open	rfs-identification
Design choices unsettled: panel, factor model, structural, sample	rfs-empirical-design
Main result exists but is fragile / multiple-testing not addressed	rfs-robustness
Tables overloaded; figures not publication-grade; SEs unclear	rfs-tables-figures
Main paper is bloated; checks/derivations belong elsewhere	rfs-internet-appendix
Prose is dense, hedged, or buries the contribution	rfs-writing-style
Preparing to submit; need Editorial Express preflight + cover letter	rfs-submission
Choosing/excluding referees; anticipating reviewer objections	rfs-referee-strategy
Received an R&R or reject-and-resubmit; need a response letter	rfs-rebuttal

Default order

1. rfs-topic-selection — lock the novel question + the contribution to finance theory/practice
2. rfs-literature-positioning — situate against JF/JFE/RFS and define the precise delta
3. rfs-identification — pin the causal-inference / asset-pricing identification strategy
4. rfs-empirical-design — sample construction, estimators, factor/structural choices
5. rfs-robustness — alternative specs, placebo, multiple-testing, out-of-sample
6. rfs-tables-figures — finalize main exhibits and standard-error reporting
7. rfs-internet-appendix — move proofs, extra tables, and details to the IA
8. rfs-writing-style — sharpen contribution framing and prose (polish stage)
9. rfs-submission — Editorial Express preflight + cover letter + fee/format checks
10. rfs-referee-strategy — referee suggestions and objection pre-mortem
11. rfs-rebuttal — after the decision letter

rfs-writing-style is a late-stage polish trigger. Do not run it while the identification or novelty story is still unsettled.

Decision heuristics

• "I have clean data but no new question" → rfs-topic-selection
• "I cite the literature but cannot state my one-sentence delta" → rfs-literature-positioning
• "My DID is TWFE and I have not addressed staggered-adoption bias" → rfs-identification
• "My cross-sectional asset-pricing result rests on a self-selected factor" → rfs-empirical-design
• "I tested 40 predictors and reported the 3 that worked" → rfs-robustness
• "My main table has 11 columns and no clustering note" → rfs-tables-figures
• "Reviewers will want 20 more tables I cannot fit" → rfs-internet-appendix
• "The abstract hides what is new" → rfs-writing-style
• "I submit Friday" → rfs-submission
• "I want to suggest/exclude referees" → rfs-referee-strategy
• "I have three reports and an R&R" → rfs-rebuttal

Differences vs. JF / JFE stacks

RFS overlaps heavily with the JF and JFE skill stacks on the causal-inference bar, but three RFS-specific facts change the routing:

• Registered Reports route. RFS accepts Registered Reports (Stage 1 design review → in-principle acceptance → Stage 2 results). If the user has a strong design but no results yet, or fears a "results-driven" desk reject, this is a real path JF/JFE do not offer. Route design-stage work through rfs-identification + rfs-empirical-design before data collection.
• SFS Cavalcade dual submission. A paper routed to the SFS Cavalcade conference can be considered jointly with RFS — a JF/JFE-absent on-ramp. Flag it in rfs-submission.
• Code-release mandate. RFS requires authors to publicly release all code underlying a published paper as a condition of publication; this hits rfs-internet-appendix and rfs-submission earlier than at journals with softer policies.

A technically clean paper that merely re-runs a known design on new data without a new question is a weak RFS fit even if it would survive at a field journal. Use the Executive Editor / editor roster above when running rfs-referee-strategy (do not suggest a handling editor as a referee).

Anti-patterns

• Do not skip rfs-literature-positioning and jump to identification — reviewers judge the contribution first.
• Do not let rfs-tables-figures polish exhibits before identification and robustness hold up.
• Do not let rfs-rebuttal draft a response letter before the manuscript itself is revised.
• Do not pursue rigor while ignoring novelty (or novelty while ignoring rigor) — RFS rejects both halves alone.