Nonlinear Dynamics of US Inflation, Money Supply, and Growth Time Series

Recently attention has been paid to significant though incipient inflationary effects in core economies, e.g. the Euro Zone and US. ‘Supply chain bottlenecks’ have been cited as a causative factor, as well as inflation in energy prices.

The problem with the bottleneck metaphor is that it implies there is a voluminous fluid in one space which can not reach another space, and while this may explain price dynamics in some markets, the market at large is better conceived of as a network, and the global shock of the pandemic as a significant and multi-level pruning of the market (a sledge hammer to a mycelium network, if you will).

In this paper we consider nonlinear time series analysis with respect to inflation, money supply, and growth in the US, as individual series and in tandem. In particular we consider the maximum characteristic Lyapunov exponent, embedding dimension with principal component analysis, embedding time delay using average mutual information and the minimum sufficient embedding dimension using the false nearest neighbour method, effects of the Theiler window using space-time separation plots, thresholded and unthresholded recurrence plots, and recurrence quantification analysis.