Nonlinear Analysis in Economic & Financial

Abstract

Traditional economic and financial models often rely on linear frameworks to analyze relationships between prices and fundamentals. For instance, the cost of carry model suggests that stock futures prices should move in tandem with spot prices (Taylor et al., 2000), while log-linear present value models propose a direct linear relationship between log dividends and stock prices (Campbell and Shiller, 1988). Similarly, housing market models predict that the house price-to-income ratio should remain stable if the real value of residential property is a constant proportion of expected future real disposable income (Black et al., 2006). Despite these theoretical predictions, empirical research using linear unit root tests frequently reveals disappointing results, with deviations from fundamental values wandering without clear reversion to equilibrium. This discrepancy highlights a potential limitation of linear models and unit root tests in capturing the true dynamics of economic and asset markets. Recent studies suggest that the failure to observe economically meaningful results may be due to the assumption of linearity. Nonlinear dynamics might be a critical factor, as linear models and tests may yield misleading inferences in the presence of complex, nonlinear mechanisms influencing macroeconomics and asset markets.