Home / Projects / Waterbridge Challenge 2025

Applied Macroeconomic Forecasting

Waterbridge Challenge 2025:
Modern Mercantilism Forecasting

"Forecasting the Future: A Modern Economics Challenge"

A comprehensive quantitative framework forecasting 25 binary global economic scenarios across 1–10 year horizons. Using advanced econometric modeling (ARIMA, VAR, logistic regression), Bayesian networks, and Monte Carlo simulation, this project delivers unprecedented analytical depth and reproducibility.


The Challenge

The Waterbridge Challenge 2025 asked teams to forecast the economic, technological, and geopolitical consequences of emerging Modern Mercantilism—a shift toward protectionist trade policies, supply chain restructuring, and technological bifurcation.

The Ask: Provide binary probability forecasts for 25 geopolitical and economic scenarios spanning trade restructuring, institutional changes, and systemic risks through 2030.

This project submits a complete quantitative implementation with 100% model coverage—every forecast backed by rigorous econometric analysis.

100% Quantitative Coverage

Unlike purely qualitative forecasts, every one of the 25 scenarios has a dedicated quantitative model:

Trade Restructuring (F1-F7)

ARIMA, linear regression, time series analysis of global tariffs, supply chain data.

Institutional Change (F8-F20)

Logistic regression, VAR models, network analysis, panel econometrics.

Systemic Risks (F21-F25)

GARCH volatility, cointegration testing, game theory, macroeconomic leading indicators.

Key Findings

Mean Forecast Probability

63.3%

Weighted by model confidence across all 25 scenarios.

System Fragility Index

0.39 (LOW)

Low risk classification; predictions robust to model variation.

Model Performance

R² > 0.65

Across primary econometric models; cross-validated.

Top Forecasts (Selected)

Forecast Probability Horizon Model Type
F1: Tariff Escalation – MFN tariffs rise ≥2pp above 2022 baseline 70% 2026 ARIMA
F2: Supply Chain Pivot – China import share <12%, Vietnam doubles 75% 2027 Linear Regression
F5: Tech Standards Bifurcation – Distinct US/China standards in 5+ tech verticals 85% 2027 Binary Classification
F7: USD Reserve Position – USD maintains >55.5% of global FX reserves 66% 2030 VAR Model
F6: Carbon Tariff Adoption – ≥7 G-20 economies implement carbon tariffs 64% 2029 Logistic Regression

Technology Stack

Core Analysis

  • Python – Primary implementation language
  • Pandas & NumPy – Data manipulation & computation
  • Statsmodels – ARIMA, VAR, cointegration testing
  • scikit-learn – Classification, cross-validation
  • pgmpy – Bayesian network analysis

Validation & Reporting

  • arch – GARCH volatility modeling
  • Seaborn & Matplotlib – Visualization
  • LaTeX – Professional report generation
  • Conda – Environment management
  • Pytest – Model validation tests

Explore the Analysis

Read the results summary, methodology walkthrough, and full technical documentation.

Read Full Documentation →