jfqa-topic-selection

JFQA Topic Selection (jfqa-topic-selection)

Use this skill to test whether a finance question belongs in the Journal of Financial and Quantitative Analysis (JFQA) before you build the paper and pay the $350 submission fee (only $275 refundable if it is not sent to a reviewer).

What JFQA publishes

JFQA covers theoretical and empirical research in financial economics, with a quantitative core:

• Corporate finance — capital structure, payout, governance, M&A, investment.
• Investments / asset pricing — cross-section of returns, factors, anomalies, portfolio choice.
• Capital and security markets — market microstructure, liquidity, price discovery.
• Financial institutions — banks, intermediaries, regulation.
• Quantitative methods relevant to finance.

The name is load-bearing: the journal rewards quantitative analysis — disciplined data, models, and inference — over purely descriptive or institutional essays.

Fit checklist

• Question sits squarely in financial economics, not adjacent (pure macro, accounting-only, generic econometrics).
• There is a quantitative empirical or theoretical answer — not just a narrative.
• The data/design can deliver a clean, defensible result (see jfqa-identification-strategy).
• The contribution is sharp enough to survive a journal that prints < 9% of 1,000+ annual submissions.
• If empirical, the data can be archived (raw or pseudo dataset) under the JFQA Code Sharing Policy.

Anti-patterns

• A descriptive industry study with no quantitative test or model.
• A method paper with no genuine finance application (belongs in an econometrics outlet).
• An incremental anomaly with no economic mechanism or out-of-sample discipline.
• Excessive length that invites desk rejection (JFQA discourages over-long papers).

Fit-scoring rubric (score before you build)

Dimension	0 points	1 point	2 points
Finance object	none identifiable	adjacent (accounting/macro proxy)	a return, spread, ratio, or institution JFQA readers own
Quantitative core	narrative only	descriptive statistics	estimation or a model with testable implications
Identification feasibility	pure correlation	plausible design, untested	a named shock, threshold, or restriction
Data archivability	data cannot be shared or simulated	pseudo data possible with effort	raw or pseudo data straightforward
Novelty at < 9% selectivity	replication-grade	extends a known result	changes a number or conclusion the field uses
Length discipline	sprawling multi-question paper	trimmable	one question, one design

Read the total: 10-12, build for JFQA; 7-9, repair the weakest dimension before writing; 6 or below, retarget the venue or redesign the project.

Two candidate questions scored (illustrative)

• Candidate A — "Does option-implied information subsume post-earnings-announcement drift?" Finance object 2, quantitative core 2 (options plus stock-return data), identification 1 (predictive design with multiple-testing exposure), archivability 2 (pseudo data is routine), novelty 1, length 2 → 10. Verdict: build it, but write the multiple-testing defense into the design before the first regression.
• Candidate B — "How do fintech lenders talk about their culture?" Scores roughly 3: no finance quantity is measured and nothing is estimated. Verdict: either redesign around measurable lending outcomes (rates, default, approval gaps) or send the descriptive version to a field outlet.

Borderline calls from adjacent fields

• Accounting-flavored questions qualify when the outcome is a finance quantity (cost of capital, returns, spreads) rather than reporting quality for its own sake.
• A pure econometrics advance qualifies only if it changes a finance conclusion in a real application.
• Macro-finance fits when the asset-market or intermediary channel is the object, not the backdrop.
• Household finance fits when portfolio, credit, or pricing behavior is quantified at scale.

Portfolio thinking under the fee structure

• Score every candidate project on the rubric before any is built; the journal's fee-and-refund design effectively prices a failed screen, so weak candidates should die at this stage, not at submission.
• A 7-9 project with a repairable dimension (usually identification or novelty) often beats starting a fresh 10 — the repair plan itself can become the paper's design section.
• Re-score after the first full results pass: projects drift, and a question that scored 11 as proposed can be an 8 as executed.

Output format

【Scope fit】corporate finance / investments / markets / institutions / methods?
【Quantitative core】Y/N — what is measured/estimated
【Selectivity check】is the contribution sharp enough for <9%?
【Next step】jfqa-literature-positioning