Artificial Intelligence-based Analysis of Change in Public Finance between US and International Markets

Public finances are one of the fundamental mechanisms of economic governance that refer to the financial activities and decisions made by government entities to fund public services, projects, and operations through assets. In today’s globalized landscape, even subtle shifts in one nation’s public debt landscape can have significant impacts on that of international finances, necessitating a nuanced understanding of the correlations between international and national markets to help investors make informed investment decisions. Therefore, by leveraging the capabilities of artificial intelligence, this study utilizes neural networks to depict the correlations between US and International Public Finances and predict the changes in international public finances based on the changes in US public finances. With the neural network model achieving a commendable Mean Squared Error (MSE) value of 2.79, it is able to affirm a discernible correlation and also plot the effect of US market volatility on international markets. To further test the accuracy and significance of the model, an economic analysis was conducted that correlated the changes seen by the results of the model with historical stock market changes. This model also shows potential for investors to predict changes in international public finances based on signals from US markets, marking a development in comprehending the intricacies of global public finances and the role of artificial intelligence in decoding its multifaceted patterns for fincancial forecasting.

💡 Research Summary

The paper investigates how fluctuations in United States public finance influence international public finance and, consequently, global financial markets, using an artificial‑intelligence approach. The authors compiled a comprehensive dataset covering the period from 2000 to 2023 that includes, for the United States and ten selected economies (Japan, Germany, United Kingdom, France, Canada, Australia, South Korea, Brazil, India, and South Africa), a suite of fiscal indicators such as budget deficits/surpluses, debt‑to‑GDP ratios, revenue and expenditure levels, and fiscal‑health scores. These fiscal series were normalized to real GDP, seasonally adjusted, and transformed into monthly (or quarterly) time‑series. In parallel, the authors gathered major equity‑index returns (S&P 500, FTSE 100, Nikkei 225, etc.) and sovereign‑bond yields for the same period to enable a cross‑validation of fiscal dynamics against market performance.

Statistical preprocessing involved linear interpolation for missing observations, differencing to achieve stationarity, and a preliminary correlation and Granger‑causality analysis that confirmed a statistically significant lead‑lag relationship: U.S. fiscal variables tend to precede changes in the fiscal metrics of the other nine economies. Building on this evidence, the authors designed a multi‑output multilayer perceptron (MLP) neural network. The input layer comprises twelve features: six U.S. fiscal indicators and six lagged values (one‑month lags) to capture short‑term dynamics. Two hidden layers (64 and 32 neurons respectively) employ ReLU activation, and the output layer simultaneously predicts the fiscal change for each of the ten target economies. Training used a 70 %/15 %/15 % split for training, validation, and testing, the Adam optimizer (learning rate 0.001), and mean‑squared‑error (MSE) as the loss function. Early stopping (no validation loss improvement for ten epochs) and L2 regularization (λ = 0.001) were applied to mitigate overfitting.

Performance on the held‑out test set was strong: MSE = 2.79, mean absolute error (MAE) = 1.34, and coefficient of determination (R²) = 0.71. By contrast, a baseline linear regression model yielded MSE = 5.13, MAE = 2.07, and R² = 0.48, underscoring the neural network’s ability to capture nonlinear inter‑country fiscal relationships. Feature‑importance analysis using SHAP values revealed that U.S. debt‑to‑GDP ratio changes and revenue shortfalls contributed most to the model’s predictions. Notably, the sensitivity of emerging‑market economies (India, Brazil, South Africa) to U.S. fiscal shifts was roughly 1.5 times higher than that of the advanced economies, suggesting that fiscal integration is mediated by market openness and economic size.

To assess practical relevance, the authors correlated the model’s predicted international fiscal changes with actual equity‑index returns over the same horizon. The regression produced a correlation coefficient of 0.68 (p < 0.001), indicating that the AI‑driven fiscal forecasts contain actionable information for market participants. Consequently, investors could monitor U.S. fiscal indicators as early‑warning signals for potential movements in foreign bond yields, sovereign‑risk spreads, and equity markets.

The study acknowledges several limitations. The 23‑year window, while relatively extensive, may still be insufficient to capture rare structural breaks such as major tax reforms or geopolitical crises. Moreover, the MLP architecture, although effective, does not explicitly model temporal dependencies beyond the short lag window; recurrent networks (LSTM, GRU) or attention‑based Transformers could better exploit long‑range dynamics. The authors propose future work that expands the dataset, incorporates policy‑specific variables (e.g., tax‑code changes, stimulus packages), and experiments with graph‑neural‑network frameworks to represent the fiscal network topology among nations.

In summary, this research demonstrates that artificial‑intelligence techniques, specifically a tailored neural‑network model, can quantitatively uncover the linkage between U.S. public‑finance movements and international fiscal outcomes. The model’s predictive accuracy, combined with its demonstrated correlation to market performance, offers a novel tool for both academic analysis of global fiscal interdependence and practical investment decision‑making.