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-submission).| 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) |
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.
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.
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:
The deliverable is a one-page objection map — each likely reviewer, their attack, the table or IA section answering it — written before submission.
| 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 |
【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