Stochastic sensitivity and dynamical complexity of newsvendor models subject to trade credit

Abstract

To describe the dynamical ordering behaviour of a risk-averse newsvendor with trade credit in supply chain, two models are formulated for two scenarios based on the bounded rationality rule. Assuming the newsvendor chooses the ordering quantity by adaptively adjusting the choice of previous period, we first develop a deterministic model. Then, mathematical analysis of the model is provided, such as the stability of equilibrium points, Neimark–Sacker bifurcation and normal form. Besides, the diagrams such as bifurcation to chaos, Lyapunov exponent, attractors and time series are illustrated numerically. It is then followed by a stochastic model to reflect the unexpected random noises from external operational environment and the perturbation on the newsvendor behaviour. The sensitivity of equilibria, the confidence ellipse and the confidence band are investigated by utilising Stochastic Sensitivity Function technique. Analysis on stochastic sensitivity shows that the retailer’s dynamical ordering exhibits more complex behaviour than the deterministic counterpart.