jfe-workflow

JFE Workflow Router (jfe-workflow)

Overview

This is the router. It does not do the work of any specialized skill; it tells you which jfe- skill to use at the current stage* of a Journal of Financial Economics (JFE) manuscript.

Default assumption: unless you say the target is a different outlet, work to JFE's culture — nuts-and-bolts methodological rigor, exhaustive robustness, and credible identification. JFE papers are typically more technical, longer, and more robustness-heavy than JF, and the referee culture rewards thoroughness over a single clever result.

JFE at a glance (verify on jfinec.com): Elsevier/North-Holland, ISSN 0304-405X, founded 1974 by Michael C. Jensen; Editor-in-Chief Toni M. Whited (Michigan) with co-editors at Kellogg, Wharton, and NYU Stern; submit via editorialmanager.com/finec with a US$850 fee (discounted by referee-earned submission rights; no fee on R&R); 100-word abstract, double-spaced author-date format; mandatory code + non-proprietary-data deposit to Mendeley Data for post-2021 submissions; the Jensen Prize (corp fin) and Fama-DFA Prize (asset pricing) are its best-paper awards.

When to trigger

• You ask "what should I do next?" on a JFE-bound paper
• You hand over a draft and need the current bottleneck diagnosed
• You keep bouncing between empirics, writing, and the response letter and have lost the thread
• You received a JFE decision letter and need to switch into revision mode

Routing table

Default order

1. jfe-topic-selection — lock the contribution and economic question
2. jfe-literature-positioning — situate against the top-3 finance frontier
3. jfe-identification — make the causal/inference design defensible
4. jfe-empirical-design — factor construction, estimators, standard errors, GMM
5. jfe-robustness — stress every result until it does not break
6. jfe-tables-figures — finalize main exhibits
7. jfe-internet-appendix — move proofs and supplementary tests off the main text
8. jfe-writing-style — polish prose to JFE register
9. jfe-submission — pre-submission preflight
10. jfe-referee-strategy — pre-empt objections before and during review
11. jfe-rebuttal — after the R&R

jfe-writing-style and parts of jfe-tables-figures are late-stage polish. Do not burn effort there while identification or the core design is still unsettled — JFE referees will not be charmed by clean tables sitting on a weak design.

Bottleneck triage card

Answer the questions in order and stop at the first "no" — that is the skill to invoke now. Everything below the first "no" is premature polish.

JFE bottleneck triage — run top to bottom, stop at first "no"
-------------------------------------------------------------------------
Q1  Can you state the contribution in one sentence a co-editor
    would care about?                          no -> jfe-topic-selection
Q2  Does related work say why JF/RFS haven't already covered it?
                                               no -> jfe-literature-positioning
Q3  Would the identification survive a hostile referee report?
    (staggered-DID bias addressed? instrument defended?)
                                               no -> jfe-identification
Q4  Are factor construction, estimator, SEs, and GMM choices
    settled and justified?                     no -> jfe-empirical-design
Q5  Does the result survive the obvious alternative specs a
    thorough JFE referee will run?             no -> jfe-robustness
Q6  Are main exhibits readable and in JFE house style?
                                               no -> jfe-tables-figures
Q7  Are proofs and extra tests offloaded to the Internet Appendix?
                                               no -> jfe-internet-appendix
Q8  Is the prose at JFE register, contribution up front,
    abstract at 100 words?                     no -> jfe-writing-style
-------------------------------------------------------------------------
All yes -> jfe-submission (editorialmanager.com/finec preflight: fee,
           Mendeley Data code/data deposit), then jfe-referee-strategy.
Decision letter arrived -> revise first, then jfe-rebuttal.

Decision shortcuts

• "I have data but no sharp economic question" -> jfe-topic-selection
• "Related work doesn't say why JF or RFS wouldn't already cover this" -> jfe-literature-positioning
• "DID is two-way FE with staggered adoption and I haven't addressed the bias" -> jfe-identification
• "I sort on a characteristic but haven't justified the factor or the breakpoints" -> jfe-empirical-design
• "Result holds in the baseline but I haven't tried obvious alternative specs" -> jfe-robustness
• "Main table has 12 columns and no one can read it" -> jfe-tables-figures
• "Referee will want 20 extra tables I can't fit" -> jfe-internet-appendix
• "Submitting this week" -> jfe-submission
• "Reviewer 2 will ask about endogeneity / multiple testing" -> jfe-referee-strategy
• "Got three referee reports back" -> jfe-rebuttal

Differences vs. the JF and RFS stacks

JFE is published by Elsevier and is editor-driven (Whited + co-editors), unlike JF (American Finance Association, Wiley) and RFS (Society for Financial Studies, Oxford UP) — so the fee structure, the finec Editorial Manager portal, and the masthead are all JFE-specific; do not recycle a JF/RFS cover letter. If the paper is a shorter, idea-forward piece where a single clean result carries the day, the JF stack may fit better; if it leans heavily on theory or method, the RFS stack may fit better. The core JFE difference: referees reward methodological thoroughness and exhaustive robustness, the Internet Appendix is expected to be deep and is appended to the main file, and the US$850 fee plus referee-earned submission rights shape the whole submission economy. Verify each journal's current scope on its official page.

Anti-patterns

• Do not skip jfe-literature-positioning and jump to identification — referees first ask "why is this not already known?"
• Do not let jfe-tables-figures beautify exhibits while the design is still contested
• Do not let jfe-rebuttal draft a response letter before you have actually revised the manuscript and re-run the affected results