High Performance Export Portfolio: Design Growth-Enhancing Export Structure with Machine Learning

IMF Working Papers, April 29, 2022

Source details

Canonical URL: High Performance Export Portfolio: Design Growth-Enhancing Export Structure with Machine Learning

Objective: Study the relationship between export structure and growth performance.
Approach: Design an export recommendation system using a collaborative filtering algorithm based on countries' revealed comparative advantages.
Coverage: Recommendations produced for over 190 economies and over 30 years.
Practical use: The recommendation system is presented as a practical tool for policymakers seeking actionable insights on export potential and diversification strategies.

Algorithm: Collaborative filtering based on countries' revealed comparative advantages.
Output: Export portfolio recommendations tailored to countries' comparative advantage signals.
Empirical scope: Applied across a global sample covering over 190 economies and over 30 years.

Alignment and growth:
- Economies with their export structure more aligned with the recommended export structure achieve better growth performance.
- Benefits measured in two dimensions:
- Higher GDP growth rate.
- Lower growth volatility.
Implication: Export structure matters for obtaining high and stable growth.

Use the export recommendation system as a policy tool to:
- Identify export potential grounded in revealed comparative advantage.
- Inform export diversification strategies that may be complex and hard to quantify.
- Support design of growth-enhancing export structures to raise GDP growth and reduce growth volatility.

High Performance Export Portfolio: Design Growth-Enhancing Export Structure with Machine Learning, IMF Working Paper No. 2022/075 (April 29, 2022)