John M.

Econometrician and economist. 44 contracts, 330+ hours and a 100% Job Success Score on Upwork. I build, estimate and defend econometric models in Stata, R, Python and EViews. Anyone can run a regression. The hard part is the identification - choosing a design that actually answers your question, testing the assumptions it rests on, and defending the specification when somebody challenges it. That is the part I get paid for, and it is the part I will not fake. If your data cannot support the claim you want to make, I will tell you before you build on it. WHAT I BUILD - Time series: ARIMA/SARIMA, VAR, VECM, cointegration, Granger causality - ARDL and NARDL bounds testing for mixed I(0)/I(1) series, including asymmetric effects - Volatility and risk: ARCH, GARCH, EGARCH, TGARCH, VaR and Expected Shortfall - Panel data: fixed and random effects, dynamic panels, Hausman and specification tests - Causal designs: difference-in-differences, instrumental variables, event studies - Asset pricing and market efficiency: CAPM, APT, Fama-French, Carhart, abnormal returns - State-space models and the Kalman filter - Limited dependent variables: Logit, Probit, Tobit, Poisson - Forecasting with rolling-origin and out-of-sample validation WHAT YOU GET - Reproducible, commented code - Stata .do file, R script or Python notebook. You can rerun every result yourself. - Publication-quality tables and charts - A plain-language write-up of what the results mean, including the diagnostics and why they matter - A clear statement of limitations, not just the numbers that look good WHO I WORK WITH Consultancies, NGOs and development organisations running impact and policy evaluation. Finance and investment teams needing volatility, risk or asset-pricing analysis. Companies that need forecasts they can plan around. Research teams and institutions that need an independent econometrician to build, replicate or verify a model. HOW WE START Send me the data, or just describe it, along with the question you need answered. I will tell you honestly whether the data can answer it, propose a specification, and agree the scope before any work begins. Tools: Stata, R, Python, EViews, SPSS, Excel. BSc Economics, The Catholic University of Malawi. Tell me what you are trying to measure.