Yamin Ahmad

Searching for Nonlinearities in Real Exchange Rates

A recent innovation in modelling exchange rates has been the use of nonlinear techniques such as threshold autoregressive models and its smooth transition variants. This article investigates the Smooth Transition Autoregressive (STAR) modelling strategy in an application to real exchange rates. The key findings are as follows. First, using the methodology advocated by Teräsvirta (1994), we find evidence of nonlinear dynamics for several of the spot dollar real exchange rates using monthly data on five of the G7 countries. However, once estimated, we find that the STAR specification is appropriate for only one of the three exchange rate series indicated to be an Exponential Smooth Transition Autoregressive (ESTAR) process. Moreover, using simulations, we show that the underlying methodology used to detect nonlinearities in the data exhibit substantial size biases, which we attribute to influential observations. We find, upon investigating alternative nonlinear specifications, that the open-loop Threshold Autoregressive (TAR) process is a more appropriate specification than the ESTAR process for the dollar–sterling and dollar–lira real exchange rates.

Outliers and persistence in threshold autoregressive processes

Abstract This paper uses Monte Carlo simulations to investigate the effects of outlier observations on the properties of linearity tests against threshold autoregressive (TAR) processes. By considering different specifications and levels of persistence for the data-generating processes, we find that additive outliers distort the size of the test and that the distortion increases with the level of persistence. In addition, we also find that larger additive outliers can help to improve the power of the test in the case of persistent TAR processes.

Government digital information discovery and exploration: the case of unraveling tourism-led-growth paradox in China

Purpose This study aims to examine the relationship between tourism development and economic growth while considering exports simultaneously. Governments in many countries have been developing and deploying strategies to attract tourism receipts as a means for economic growth. However, assessing the potential impact of tourism on economic growth among large economies is still in its infancy. Design/methodology/approach Using a vector error correction model framework, this study examines the relationship among exports, gross domestic product (GDP) and tourism receipts (including international tourism receipts and domestic tourism receipts in two separate models) with macro data that covers two recent decades (1994-2013) in China. Findings The empirical findings confirm the existence of a long-term equilibrium relationship in each of these two tri-variate models. The empirical findings reveal that (1) both tourism-led-growth and export-led-growth hypotheses are supported, (2) the growth rate of tourism receipts exhibit a higher relevance with GDP growth than export growth and (3) the growth rate of international tourism shows a higher relevance with GDP growth than domestic tourism growth. Originality/value Using macroeconomic data collected by the Chinese government, the current study employs an advanced econometric methodology to explore the potential benefits of tourism on economic growth in China.

Nonlinear Time Series Models and Model Selection

This chapter examines the consequences of model misspecification of nonlinear time series models in being able to appropriately characterize the data generating process of a particular economic time series. We conduct two sets of Monte Carlo experiments in order to examine how two classes of nonlinear models - Markov switching (MS) and smooth transition/nonlinear autoregressive models (NAR) - can be distinguished from one another within data. The first set assumes that nonlinearity in the data generating process results from an observable threshold arising from the class of NAR models, and we examine the extent to which MS models can capture these types of nonlinearities. The second set of Monte Carlo experiments assume that nonlinearity in the data generating process arises from a latent variable, and we examine the extent to which NAR models can capture this type of Markov switching process. Results from the first case indicate that the MS models do fairly well in capturing logistic smooth transition autoregressive (STAR) type of nonlinearity, although biases emerge for exponential STAR (ESTAR) and the multiple regime LSTAR (MRLSTAR) processes. We document the magnitude of these biases. For the second case, no consistent pattern emerges, which we attribute to a lack of power in the linearity test. However, for both cases that we examine, the power of identifying differences in persistence between regimes increases, the greater this difference in between the regimes.