From jmcb-skills
Crafts JMCB-ready exhibits: IRF plots, coefficient figures, regression and balance-sheet tables that are self-contained and economically legible. Does not re-run analysis or write prose.
How this skill is triggered — by the user, by Claude, or both
Slash command
/jmcb-skills:jmcb-tables-figuresThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
- An IRF figure has bands but no statement of what shock, what units, or what horizon
JMCB is a quantitative monetary/banking journal, so its signature exhibits are impulse-response functions, local-projection coefficient plots, and event-study figures, alongside the standard regression and balance-sheet tables. The bar is that each exhibit is self-contained and economically legible: a reader skimming should grasp the shock, the units, the horizon, the magnitude, and the uncertainty without hunting through the text. Because the journal recommends ≤40 pages including tables and figures, exhibit economy is also a length discipline — every panel must earn its place.
Referees skim exhibits before prose. Order them so the argument is legible from the figures and tables alone:
If a referee can reconstruct the paper's claim from this sequence without the text, the exhibits are doing their job.
Table and figure notes are not decoration at JMCB — given the journal's replication heritage, they are where a careful referee checks that the exhibit means what the text claims. Each note should state the sample period and frequency, the estimator, the inference method (clustering dimensions, bootstrap, HAR), and the units of the reported effect. For IRFs, state the shock normalization explicitly (a 25bp or one-SD surprise). Ambiguous units or missing inference details invite the referee to assume the worst.
A dynamic response (an IRF or event study) almost always belongs in a figure — the shape and the timing are the point, and a table of horizon coefficients hides both. A cross-sectional comparison or a precise magnitude with its standard error belongs in a table. JMCB papers often over-table dynamics and under-figure them; converting a 12-row horizon table into a banded IRF plot both clarifies the result and saves space against the 40-page recommendation. Reserve tables for where the exact number, not the pattern, is what a reader needs.
A draft shows a local-projection figure of lending after a monetary shock with bare y-axis "coefficient" and no shock size. The JMCB fix: relabel to "% response of bank lending to a 100bp contractionary surprise," extend the horizon to 12 quarters to show the −1.8% peak at quarter 4 and the return to zero by quarter 10, add 90% bootstrap bands, and state the magnitude in the caption. A second figure splits by capital quartile so the heterogeneity the paper argues is visible at a glance. Two clean figures replace four cluttered tables, freeing pages under the 40-page cap.
【Journal】Journal of Money, Credit and Banking
【Skill】jmcb-tables-figures
【Exhibit type】IRF/LP figure / regression table / balance-sheet table / event study
【Self-contained?】shock + units + horizon + magnitude readable without the text? [Y/N]
【Headline clarity】which column/panel is the result; is it flagged?
【Magnitude shown】economic size alongside coefficients/bands? [Y/N]
【Specs visible】fixed effects + clustering + N present? [Y/N]
【Length fit】exhibits within the 40-page core; overflow mapped to appendix
【Next skill】jmcb-internet-appendix
npx claudepluginhub brycewang-stanford/awesome-journal-skills --plugin jmcb-skillsGenerates JBF-ready tables and figures: descriptive stats, regression tables, event-study plots, mechanism/heterogeneity exhibits, and appendix robustness inventories with proper notes.
Formats tables and figures for Journal of International Money and Finance manuscripts. Builds panel regression tables, event-study plots, and country-heterogeneity displays following Elsevier/finance presentation norms.
Designs exhibits for JFI papers: summary tables, main results with FE progression, event-study figures, and robustness grids. Useful when building or revising a submission's table/figure set.