brycewang-stanford Referee Strategy (jf-referee-strategy)
When to trigger
- Before submitting: anticipating the objections a JF referee will raise
- The portal asks for suggested / opposed reviewers
- A conflict of interest among editors, AEs, or referees needs flagging
How JF review works (plan around it)
JF, the AFA flagship, runs a demanding multi-round process: a handling editor (currently Antoinette Schoar (MIT), with co-editors and 50+ Associate Editors — verify the masthead) screens first, then assigns typically 2–3 referees. The reality is severe — ~33–45% desk-rejected, ~5% accepted (afajof.org editor reports, accessed 2026-05-30). Most papers die at desk review, so the first job is surviving the editor, not the referees.
JF-specific channels
- A cover letter is normally unnecessary at JF — use one only to flag a conflict of interest (referees/AEs/editors who should be excluded) or to request a code-sharing exemption. Do not pad it (see
jf-submission). - Suggested/opposed reviewers, if the portal asks, should be specific and justified.
Anticipating objections (pre-empt in the paper)
| Likely referee objection | Where to neutralize it |
|---|---|
| "This is already in [paper]" | Contribution paragraph (jf-literature-positioning) |
| "Identification is not credible" | Design section + IA tests (jf-identification) |
| "Just data mining / multiple testing" | Disclosed search + adjusted thresholds (jf-robustness) |
| "Not general-interest enough" | Reframe for the broad AFA reader (jf-topic-selection) |
Reading the AFA screening funnel
Two gates sit in series. The handling editor reads the abstract, introduction, and identification before spending referees' time; you only reach the referee panel once the editor judges the paper plausibly flagship-publishable. Because the editor screens for general interest and first-order importance, the highest-leverage pre-submission work is making the first three pages survive an editor who is not a specialist in your niche. Treat the editor as the audience for the introduction, the referee for the design and Internet Appendix.
Desk-reject patterns at a top-3 finance journal
Recurring shapes that draw a fast editorial "no" at JF (and JFE/RFS); pre-empt them before the portal, not in the rebuttal:
| Desk-reject pattern | What the editor sees | Pre-submission fix |
|---|---|---|
| Narrow / single-subfield question | "Interesting to microstructure people only" | Reframe the first-order stake (jf-topic-selection) |
| Causal verb, correlational design | No named shock/instrument in the intro | Lead with the source of variation (jf-identification) |
| Anomaly with naive t > 1.96 only | Smells of the factor zoo | Multiple-testing-adjusted threshold (jf-robustness) |
| Statistically significant, economically tiny | A 2-bp effect dressed as a discovery | Report Sharpe gain / % of market cap up front |
| Wrong-journal fit (reads like a JFE paper) | Specialized, results-driven, no broad hook | Re-center on the general-interest reader |
Confirm the current desk-reject rate language against the journal's editor reports rather than quoting a fixed number.
Worked vignette — pre-empting the panel on an anomaly paper
Illustrative figures. A new "supplier-concentration" return predictor shows a 55 bps/month long-short spread, raw t = 3.1. War-game the panel before submitting:
- AE / asset-pricing referee: does 55 bps survive Fama–French five-factor + momentum alphas, and is t = 3.1 above the factor-zoo bar (~3.0, illustratively)? Pre-empt with the post-adjustment alpha (say 38 bps, t = 2.7) in the body and the full factor grid in the IA.
- "Is it real" referee: microcap artifact? Pre-empt with a NYSE-breakpoint, value-weighted version (say 31 bps).
- General-interest editor: why would a corporate reader care? Tie the signal to a mechanism (supply-chain risk pricing) that reaches beyond the anomaly catalog.
The deliverable is a one-page objection map — each likely reviewer, their attack, the table or IA section answering it — written before submission.
Referee-pushback patterns and the JF-specific fix
| Pushback you will hear | JF-specific fix |
|---|---|
| "Run it value-weighted with NYSE breakpoints" | Pre-build the VW result; IA holds EW and alternative breakpoints |
| "The effect is economically trivial" | Lead with Sharpe gain / bps, not the t-stat |
| "Your shock coincides with the 2008 crisis" | Excluded-period and placebo windows in the IA |
Checklist
- Top 3 likely referee objections listed and pre-empted in the manuscript
- Desk-reject risk assessed (general interest, identification, fit)
- Opposed reviewers (genuine COI) noted only if the portal asks
- Cover letter used only for COI or code-exemption — otherwise omitted
- Suggested reviewers are specific and defensible
Anti-patterns
- Writing a long cover letter JF did not ask for
- Listing opposed reviewers without a real conflict (reads as gaming)
- Ignoring desk-reject risk and over-optimizing for referees
- Leaving the obvious "already known" objection unaddressed
Output format
【Top objections + where pre-empted】...
【Desk-reject risk】low / med / high — why
【COI to flag in cover letter?】yes / no
【Suggested/opposed reviewers (if asked)】...
【Next step】jf-submission → (after decision) jf-rebuttal