Explaining the German hog price cycle: A nonlinear dynamics approach

Abstract

We investigated German hog-price dynamics with an innovative ‘diagnostic’ modeling approach. Hog-price cycles are conventionally modeled stochastically—most recently as randomly-shifting sinusoidal oscillations. Alternatively, we applied nonlinear time series analysis to empirically reconstruct a deterministic, low-dimensional, and nonlinear attractor from observed hog prices. We next formulated a structural (explanatory) model of the pork industry to synthesize the empirical hog-price attractor. Model simulations demonstrate that low price-elasticity of demand contributes to aperiodic price cycling – a well know result – and further reveal two other important driving factors: investment irreversibility (caused by high specificity of technology), and liquidity-driven investment behavior of German farmers.