Replication Data for: A threshold dynamic-programming rule for India's strategic petroleum

2014–2026 backtest data and code replicates a threshold dynamic-programming model for India's strategic petroleum reserve policy. The dataset includes Brent price, an oil-specific geopolitical-risk index (Oil-GPR), and a disruption-probability calibration with an AUC of 0.836. It was authored by Adesh Mishra and published on Harvard Dataverse in June 2026.

Use Cases

• Calibrating disruption probability models based on the logistic regression on the Oil-GPR index.
• Backtesting strategic petroleum reserve release and refill policies against historical price and risk data.
• Conducting welfare analysis normalized by the value-of-lost-load (VOLL) construction.
• Reproducing the empirical price–risk transition kernel for forecasting.
• Testing the robustness of energy policy models under different geopolitical risk scenarios.

Strengths

• Includes a backtest covering the period from 2014 to 2026.
• Features a calibrated disruption-probability model with a reported AUC of 0.836.
• Integrates public data from sources like PPAC, UN Comtrade, MoSPI, and the Iacoviello & Tong Oil-GPR index.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Public data from PPAC; UN Comtrade HS 2709; MoSPI; Iacoviello & Tong Oil-GPR.

Collection Method

Data compilation and model calibration for replicating academic paper results.

Time Range

2014–2026

Freshness

Last updated 2026-06-25 01:53:02; freshness should be verified.

Geography

India