What if the core assumption behind decades of macroeconomic modeling, treating GDP as a difference-stationary series was mathematically and empirically wrong? This article delivers a radical, data-backed challenge to one of the most entrenched beliefs in applied econometrics. We prove that long-run behavior of GDP, along with its core monetary counterparts and related macroeconomic aggregates, is governed by nonlinear dynamics bounded by structural ceilings, rather than by an unbounded linear trend. These nonlinear forces persist even under sustained monetary expansion and exogenous shocks, revealing a self-constraining architecture that challenges conventional difference-stationary assumptions. We demonstrate, both econometrically and mathematically, that GDP converges toward a definable limit inherent to the economic environment. We reconstruct the classical equation of exchange under nonlinear interdependence. The theory is applied to nominal and real time series of European countries as well as for the US. In a field where most models assume “more money equals more growth,” our findings, backed by rigorous time series analysis of 49 macro-economic time series from 37 different countries, including unit root tests, and nonlinear estimation reveal a ceiling exists. The findings are striking: GDP does not drift endlessly, but converges. Moreover, our results show that velocity, money supply, price level, and output exhibit interdependence as elements of the same nonlinear mathematical family. This systemic behavior challenges the dominant DS modeling paradigm and reshapes how macro-financial trajectories should be forecasted and understood. By identifying measurable upper bounds in GDP evolution and proposing a new analytical framework, this discovery introduces a decisive shift in the modeling of macroeconomic aggregates. The implications extend to monetary theory, inflation expectations, and the understanding of structural shocks in long-term forecasting. If you've ever questioned the long-run validity of DS models or struggled with unexplained forecasting errors in macro models, this article offers the missing link.
Tannoury et al. (Mon,) studied this question.