From jfi-skills
Guides analysis planning for JFI papers: bank/loan-level panel empirics, demand-absorbing specifications, robustness batteries, and numerical examples for theory.
How this skill is triggered — by the user, by Claude, or both
Slash command
/jfi-skills:jfi-data-analysisThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
- Building the empirical analysis on bank/firm/loan data, or its robustness battery
When the paper is a model, "data analysis" is lighter and means reproducible computation:
Prepare a Data Statement and link datasets via Editorial Manager; cite data with the [dataset] tag
(see jfi-replication-and-data-policy). Under Elsevier Option C, deposit/cite/link research data where
possible or explain why sharing is restricted.
Pick data for the intermediation mechanism, not the other way around — JFI referees notice when the dataset cannot carry the claimed channel:
| Mechanism under study | Workhorse data | What the merge must support |
|---|---|---|
| Relationship lending / information capture | Credit register or DealScan loan-level | Multi-bank firms, so firm×time absorption is feasible |
| Capital / regulation transmission | Call Reports, FR Y-9C, stress-test exposures | Bank-level shock measured before announcement |
| Deposit competition / franchise value | FDIC Summary of Deposits, branch-level rates | Market-level (county/MSA) shares and pricing |
| Runs, liquidity, interbank stress | Supervisory or payment-system records (typically restricted) | Daily/weekly frequency around the stress window |
| Fintech displacement of banks | Platform loan tapes plus bank comparators | Comparable borrower-risk controls across lender types |
A hypothetical JFI paper estimates that a 1pp rise in required capital cuts loan growth to the same firm by 2.1pp (s.e. 0.6, firm×time FE, clustered by bank). The battery a JFI referee expects, each row tied to a named threat:
【Track】empirical / theory
【Sample / parameters】<universe + filters, or calibration>
【Core spec / example】<FE structure, or the numerical illustration>
【Robustness】<the battery, or seed/determinism notes>
【Next skill】jfi-contribution-framing
npx claudepluginhub brycewang-stanford/awesome-journal-skills --plugin jfi-skillsBuilds or audits empirical data and estimation pipelines for Journal of Banking & Finance manuscripts: financial datasets, bank panels, winsorization, fixed effects, robustness checks, and reproducible scripts.
Audits the core analytical engine of a JFI banking/intermediation paper: causal design separating credit supply from demand in empirical work, or assumptions/equilibrium discipline/proofs in theory. Tests the design validity, does not run the analysis.
Runs and documents empirical finance analysis for JFQA papers: data construction (CRSP/Compustat/TAQ/IBES), winsorizing, fixed effects, clustered/Newey-West standard errors, robustness, and reproducibility.