brycewang-stanford Referee Strategy (jfm-referee-strategy)
When to trigger
- Pre-submission: you want to surface the objections a microstructure insider will raise before they do
- A report arrived and you need to triage which comments are fatal, fixable, or push-back
- You suspect the referee will challenge measurement or endogeneity rather than the idea
- You need to estimate whether the asks are within an R&R's reach or signal a reject
- The editor's letter prioritizes some comments and you must read the signal
The objections JFM referees actually raise
A JFM referee is a microstructure specialist. The predictable lines of attack cluster, and most live in measurement and identification, not novelty.
| Referee line | What they really doubt | Pre-empt with |
|---|---|---|
| "This is not really microstructure" | the mechanism is broad finance | jfm-topic-selection framing; lead with the trading-process object |
| "The result is driven by the liquidity measure" | measurement artifact | jfm-empirical-design + alternative constructs in jfm-robustness |
| "Liquidity is endogenous here" | reverse causality / confounding | jfm-identification: a market-structure shock or defensible IV |
| "Your event window / filters are arbitrary" | researcher-degrees-of-freedom | justify window from mechanics; document every filter |
| "Standard errors overstate precision" | un-clustered panel inference | two-way clustering / NW; state it in table notes |
| "Incremental to [known paper]" | contribution undersold | jfm-literature-positioning: explicit they/we gap |
| "PIN/structural estimates are fragile" | numerical instability | report solver settings, multistart, sensitivity |
Running the pre-mortem and the triage
Pre-submission pre-mortem. For each row above that could apply, write the strongest version of the objection and the exhibit/argument that answers it. If an answer requires analysis you have not run, run it now — it is cheaper than an R&R cycle. The output is a list of latent weaknesses, each with a planned defense.
Post-decision triage. Sort every comment into: (1) fatal-if-true — goes to the top of the response and gets the most work; (2) fixable — new analysis or exposition; (3) defensible — politely push back with evidence; (4) editor-flagged — the editor signaled priority, so it is effectively (1). Read the editor's letter for the real decision rule: which comments must be resolved for acceptance. Then hand the sorted list to jfm-rebuttal.
Timing the pre-mortem
Run the pre-mortem twice. The first pass is before submission, while there is still time to run the analysis that closes a foreseeable objection — this is its highest value, since adding an alternative-measure column now costs days, whereas discovering the gap in an R&R costs months. The second pass is on receiving the report, to triage the actual comments against your anticipated list: comments you predicted are already half-answered, and genuinely new ones tell you where your model of the field was incomplete. Skipping the pre-submission pass is the most expensive economy in the whole process.
Worked pre-mortem (illustrative)
A paper claims a new dark-pool venue worsened lit-market liquidity. The pre-mortem surfaces: (1) "It's just the measure" — the result holds only for quoted spread → run effective and realized before submitting; (2) "Endogenous venue choice" — flow self-selects into the dark pool → exploit the staggered venue launch and cluster by stock; (3) "Time-of-day" — dark trading concentrates midday → add intraday-bin fixed effects; (4) "Incremental to known fragmentation work" — name the closest 3 papers and the gap. Each is answered with a concrete exhibit before the editor ever sees the paper. The output is a short table of latent weakness → planned defense → exhibit number, which doubles as the skeleton of the eventual response letter.
Reading the decision letter
The editor's letter, not the referee reports, encodes the real decision rule. Look for: which comments the editor calls "essential" or "must address"; whether the tone is "revise" (fixable) or "we would consider a substantially revised paper" (closer to reject); and whether referees conflict (your cue to ask the editor to adjudicate rather than guess). An R&R that asks only for measurement and inference checks is far more tractable than one questioning the core identification — calibrate the effort and the timeline accordingly, and be honest with coauthors about which it is.
The two reviewer profiles to plan for
JFM submissions typically draw a microstructure specialist and, sometimes, a broader-finance reader. Plan for both. The specialist scrutinizes measurement, institutional accuracy, and the mechanism — answer them with construction detail, alternative measures, and a heterogeneity test. The generalist asks why the result matters beyond the plumbing — answer them with a clear economic-magnitude statement and a sentence on the broader implication for market quality. A paper written only for the specialist can read as narrow to the generalist; one written only for the generalist loses the specialist's trust. The intro and abstract should satisfy both in their first paragraphs.
Checklist
- Each predictable microstructure objection has a planned answer before submission
- Measurement-artifact and endogeneity objections are pre-empted with concrete exhibits
- Inference (clustering) objection is closed in advance
- Post-decision: every comment classified fatal / fixable / defensible / editor-flagged
- The editor's priorities are extracted from the letter, not just the referees'
- Push-backs are evidence-based, never tone-based
- The paper carries at least one advocacy hook (clean shock, new measure, exact institutions, or mechanism-predicted heterogeneity)
What a microstructure referee rewards, beyond catching flaws
Anticipating objections is defensive; the strongest submissions also give the referee reasons to advocate. A microstructure specialist becomes a champion when the paper (1) gets the institutional detail exactly right — they trust authors who clearly know the market's plumbing; (2) offers a clean identifying shock they had not thought to exploit; (3) introduces a measure or decomposition they will want to use themselves; (4) shows the effect where the theory predicts, demonstrating command of the mechanism. Build at least one of these into the paper deliberately, and surface it in the intro and cover letter. A pre-mortem that only patches weaknesses produces a defensible paper; adding an advocacy hook produces an accepted one.
Anti-patterns
- Treating a microstructure referee like a generalist and under-explaining the mechanism
- Ignoring the measurement-artifact objection because the idea feels strong
- Reading referee reports without separating editor-prioritized comments from the rest
- Planning to argue down a fatal endogeneity concern instead of redesigning
- Discovering a foreseeable objection only after the reject
Estimating whether an R&R is reachable
Not every revise request is equally winnable. Sort the editor's asks by the cost and risk of the work they demand. Tractable R&Rs ask for alternative liquidity measures, better clustering, more robustness, sharper exposition, or an added subsample — all additive, none threatening the core. Hard but doable R&Rs ask for a heterogeneity test of the mechanism or a new data source for generalizability. Near-reject R&Rs question the identification itself ("we are not convinced liquidity is exogenous here") — these require a redesign, not a patch, and you should decide honestly whether you can deliver one before sinking months into a revision. Tell coauthors which bucket the letter is in; mis-reading a near-reject as a tractable R&R wastes a cycle.
Output format
【Journal】Journal of Financial Markets (JFM)
【Skill】jfm-referee-strategy
【Mode】pre-mortem / post-decision triage
【Top objection】<microstructure line> → planned answer
【Measurement defense】alternative constructs ready? [Y/N]
【Endogeneity defense】shock/IV ready? [Y/N]
【Triage (if report)】fatal / fixable / defensible / editor-flagged counts
【Editor signal】what must be resolved for acceptance
【Next skill】jfm-rebuttal