My Portfolio
Research archive
International Trade Econometrics & Geoeconomic Systems2001–2021 panel · Completed econometric study

Geopolitical Risk, Supplier Concentration, and Fuel Import Volumes

Poster summary

Stata · panel data · Driscoll–Kraay · WITS · HHI

A completed country-panel econometric study of the nonlinear relationship between geopolitical risk, fuel-supplier concentration, and import volumes.

01

Research question

How are geopolitical risk and supplier concentration associated with national fuel-import volumes after accounting for macroeconomic demand and common time shocks?

02

Methodology

Pooled OLS, random effects, country fixed effects with year effects, panel diagnostics, and Driscoll–Kraay fixed-effects inference

03

Dataset

714 country-year observations, 2001–2021, combining WITS fuel trade and concentration measures with geopolitical-risk and macroeconomic controls

04

Main finding

The preferred fixed-effects and Driscoll–Kraay models estimate a positive linear and negative quadratic geopolitical-risk term, consistent with an inverted-U conditional association; supplier concentration is positively associated with import volume.

Geopolitical Risk, Supplier Concentration, and Fuel Import Volumes

Overview

This project studies how geopolitical risk and import-source concentration relate to fuel-import volumes across countries. It uses a balanced set of panel estimators and reports diagnostics for heteroskedasticity, serial correlation, and cross-sectional dependence.

Research Question

How are geopolitical risk and supplier concentration associated with fuel-import volumes after controlling for GDP, population, energy use, trade openness, country effects, and year shocks?

Economic or Technical Motivation

Geopolitical risk can initially induce precautionary purchases or inventory accumulation, while sufficiently high risk may disrupt logistics, finance, and supply. A quadratic specification tests this nonlinearity rather than imposing a constant marginal relationship.

Data

The analytical table contains 714 country-year observations from 2001 to 2021. WITS trade files are used to construct fuel-import volumes and supplier HHI. Macroeconomic controls include GDP, population, energy use, and trade openness.

Methodology

The executed workflow compares pooled OLS, random effects, country fixed effects with year effects, and fixed effects with Driscoll–Kraay standard errors. It also runs the Hausman test and panel diagnostics for heteroskedasticity, autocorrelation, and cross-sectional dependence.

Main Findings

In the fixed-effects and Driscoll–Kraay specifications, the geopolitical-risk linear term is 0.909 and the squared term is -0.412; both are reported as statistically significant. This pattern is consistent with an inverted-U conditional association. Fuel-supplier HHI is 0.388 in the same models and is positively associated with import volume.

Robustness, Validation, or Model Assessment

The interaction pattern is stable between conventional fixed effects and Driscoll–Kraay inference. Pooled OLS differs in sign, reinforcing the importance of country effects. The project reports model-selection and panel-dependence diagnostics rather than relying on one estimator.

Tools and Technologies

Stata, WITS data, panel fixed effects, random effects, Driscoll–Kraay standard errors, and HHI construction are directly demonstrated.

Limitations

The estimates are associational. Geopolitical-risk measures and import volumes may be jointly affected by omitted global shocks, and the positive HHI association should not be interpreted as a welfare effect.

Deliverables

  • Country-year fuel-import and concentration panels
  • Stata estimation and diagnostic scripts
  • Four-model regression table
  • Data-construction notebooks and source workbooks