FX-constrained growth: Fundamentalists, chartists and the dynamic trade-multiplier

TL;DR

This paper develops a heterogeneous agents model for developing countries' FX markets, linking exchange rate dynamics with economic growth and demonstrating the importance of non-speculative demand in steady-state growth.

Contribution

It introduces a novel model distinguishing speculative and non-speculative FX demand, connecting FX supply growth to economic growth via a dynamic trade-multiplier.

Findings

Numerical simulations match stylized facts of exchange rate and growth distributions.

FX fluctuations in Latin American countries show similar statistical properties.

Time-varying estimates of the trade-multiplier align with observed growth rates.

Abstract

Behavioural finance offers a valuable framework for examining foreign exchange (FX) market dynamics, including puzzles such as excess volatility and fat-tailed distributions. Yet, when it comes to their interaction with the `real' side of the economy, existing scholarship has overlooked a critical feature of developing countries. They cannot trade in their national currencies and need US dollars to access modern production techniques as well as maintain consumption patterns similar to those of wealthier societies. To address this gap, we present a novel heterogeneous agents model from the perspective of a developing economy that distinguishes between speculative and non-speculative sectors in the FX market. We demonstrate that as long as non-speculative demand responds to domestic economic activity, a market-clearing output growth rate exists that, in steady-state, is equal to the ratio between FX supply growth and the income elasticity of demand for foreign assets, i.e., a generalised dynamic trade-multiplier. Numerical simulations reproduce key stylised facts of exchange rate dynamics and economic growth, including distributions that deviate from the typical bell-shaped curve. Data from a sample of Latin American countries reveal that FX fluctuations exhibit similar statistical properties. Furthermore, we employ time-varying parameter estimation techniques to show that the dynamic trade-multiplier closely tracks observed growth rates in these economies.