A Nonlinear Stochastic Mathematical Framework for Modeling NRI Investment Decision Dynamics and Risk Aversion Patterns in the Kerala Productive Sectors

Abstract

The research paper delves into the nonlinear stochastic modeling of NRI investment decisions in productive sectors of Kerala, including wealth, risk aversion, market volatility, and economic shocks. Regression analysis indicates that wealth is a significant factor in investment, with a coefficient of 0.18, suggesting that the more wealth one has, the more investment one makes. Market volatility is also relevant, with a positive coefficient (0.22), indicating that higher volatility is associated with greater investment in riskier areas, particularly among more affluent investors. These findings are further supported by Monte Carlo simulations with 10,000 iterations, which indicate an average investment in low-risk sectors of 62.35 and a standard deviation of 18.56. The average of high-risk sectors is 37.65 with a standard deviation of 14.89. The simulation also shows that richer investors spend more in risky areas when there is an increased volatility in the market. The nonlinear stochastic model works better than the linear model. The nonlinear model had a Mean Absolute Error (MAE) of 0.12 as compared to the 0.18 of the linear models. The nonlinear model had an R2 of 0.89, and the linear model had an R2 of 0.75, indicating that the nonlinear model is more predictive of the NRI investment behavior. These findings would be useful in policy development in order to increase NRI investments in the Kerala economy.