Asian Journal of Economics, Finance and Management

Frontier markets such as Nigeria typically experience a range of problems, among them high volatility, currency pressure, inflation shocks, thin trading, and weak market efficiency. In such situations, it is very challenging for traditional asset-pricing models like CAPM and the Fama-French model to accurately quantify systematic risk and forecast stock returns. This paper employs AI techniques to improve asset pricing and risk estimation in Nigeria, spanning from 2010 to 2024. The study juxtaposes contemporary AI models such as XGBoost, Random Forest, and Long Short-Term Memory (LSTM) networks with their classical counterparts. To figure out which risk factors are the most significant, the paper deploys explainable AI (SHAP). The findings indicate that AI models offer far superior predictive accuracy and are able to capture non-linear market behavior to a much greater extent than traditional models. The SHAP analysis indicates that the factors causing the greatest systematic risk in Nigeria are exchange-rate volatility, inflation, oil prices, liquidity, and trading volume. The research determines that AI-enabled models provide a more dependable and transparent asset-pricing framework for frontier markets and, thus, can be of great help to investors, regulators, and policymakers in their financial ‍‌decisions.