Title: Does non-linearity in exchange rate hold in Nigeria? evidence from smooth transition autoregressive model

Authors: James Temitope Dada; Philip Akanni Olomola; Adebayo Adedokun

Addresses: Department of Economics, Obafemi Awolowo University, Ile-Ife, 230001, Nigeria ' Department of Economics, Obafemi Awolowo University, Ile-Ife, 230001, Nigeria ' Department of Economics, Obafemi Awolowo University, Ile-Ife, 230001, Nigeria

Abstract: This study employs smooth transition autoregressive (STAR) model to investigate the non-linearity in exchange rate process in Nigeria within the context of exchange rate parity theory (ERPT). Quarterly data from 1981Q1 to 2017Q4 is used. The outcome of the study confirms the presence of exchange rate parity in Nigeria using Augmented Dickey Fuller (ADF) and Kwiatkowski-Phillips-Schmidt-Shin (KPSS) unit root tests. The study rejects the null hypothesis of linearity in nominal exchange rate in Nigeria. Furthermore, the estimation result of non-linear least squares (NLS) regression indicates that most of the parameter estimates are statistically significant. The study reveals that non-linear models best explain exchange rate dynamics in Nigeria. Likewise, the study discovers that exchange rate process in Nigeria is best fit with smoothly asymmetric logistic smooth transition autoregressive (LSTAR). The study concludes that heterogeneous nature of participants and asymmetric information in the market cause exchange rate to adjust in non-linear version.

Keywords: exchange rate; exchange rate parity; STAR; smooth transition autoregressive.

DOI: 10.1504/IJMEF.2021.114024

International Journal of Monetary Economics and Finance, 2021 Vol.14 No.2, pp.152 - 165

Received: 10 May 2019
Accepted: 04 Aug 2020
Published online: 06 Apr 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article