From joe-skills
Guides Monte Carlo simulation design and empirical illustration for Journal of Econometrics submissions, covering size/power, DGP stress tests, and computational hygiene.
How this skill is triggered — by the user, by Claude, or both
Slash command
/joe-skills:joe-data-analysisThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
- The theorems are settled but the finite-sample evidence is thin or one-off
At the Journal of Econometrics the empirical work serves the method, not the other way around. A theorem describes behavior as $n\to\infty$; the Monte Carlo shows the asymptotics bite at realistic sample sizes, and the empirical illustration shows the method is usable and yields a sensible answer on real economic data. The applied illustration is a demonstration, not the paper's primary contribution — purely applied work without a methodological advance is out of scope here. Build both as evidence that the formal claims hold.
joe-literature-positioning).Build the Monte Carlo grid around the theorem's weak points, not around flattering defaults:
| Dimension | Minimum stress case |
|---|---|
| Sample size | A small or moderate $n$ where the asymptotic approximation is plausibly strained. |
| Identification strength | Weak instruments, near-collinearity, boundary parameters, local-to-zero effects, or sparse support as relevant. |
| Error process | Heavy tails, heteroskedasticity, serial/cross-sectional dependence, or clustering that matches the target application. |
| Tuning | Bandwidth, penalty, lag, moments, sieve dimension, or bootstrap choice varied enough to show stability. |
| Competitor | The closest existing estimator/test run on exactly the same DGP and reporting scale. |
Pre-register the cells in the simulation plan, then mark any post-hoc additions as diagnostics. JoE referees punish Monte Carlos that prove only that the authors found a friendly DGP.
[dataset] tag and prepare materials for the archive (see joe-replication-and-data-policy).【MC estimators】bias / RMSE / coverage reported? [Y/N]
【MC tests】size at 5%/10% + size-adjusted power? [Y/N]
【DGP stress】distributions / dependence / tuning / boundary? [list]
【Benchmark】compared to nearest method on same DGP? [Y/N]
【Reproducibility】seeds + reps + MCSE reported? [Y/N]
【Illustration】method changes/sharpens a real conclusion? [Y/N]
【Next step】joe-tables-figures
npx claudepluginhub brycewang-stanford/awesome-journal-skills --plugin joe-skillsDesigns Monte Carlo simulations and numerical illustrations for Econometric Theory papers, showing finite-sample behavior tracks asymptotics.
Guides design and audit of Monte Carlo simulations, empirical applications, and estimator comparisons for The Econometrics Journal, focusing on reproducibility and theoretical alignment.
Designs and audits Monte Carlo simulation evidence for Econometrica manuscripts, covering finite-sample performance, regularity-condition stress tests, and degenerate cases.