A threshold model for the French franc/deutschmark exchange rate

Abstract

The behaviour of the French franc/deutschmark exchange rate is examined in this paper. During the time period studied, these currencies were constrained to lie within prescribed bands relative to one another and the usual random walk explanation of the exchange rate may not be appropriate. The data are examined for evidence of non‐linear structure and it is shown that a piecewise linear SETAR model provides a better explanation and superior forecasting performance than a random walk.

A Double-threshold GARCH Model for the French Franc/Deutschmark exchange rate

“…The models in this study are those used in Chappell et al (1996) estimated using 500 daily observations on the French Franc/German mark, spanning the period 1 May 1990 to 30 March 1992, obtained from Datastream International. Following Chappell et al, the models are estimated in the natural logarithms of the levels rather than the ®rst dierences, even though the null hypothesis that the former are unit root non-stationary cannot be rejected.…”

“…1 Another useful application of a SETAR model in economics or ®nance is in modelling movements in exchange rates which are subject to intervention by central banks. Chappell et al (1996), for example, apply a threshold model to the French Franc/Deutschmark exchange rate which is required to reside between two prescribed bands as part of the European Exchange Rate Mechanism (ERM). If the currency moves close to one of the bands, then the French and German central banks are obliged to intervene in the markets, buying the weaker currency and selling the stronger in an attempt to shift the exchange rate back closer to the central parity.…”

“…The purpose of this paper is to use the data series and SETAR speci®cations used in Chappell et al (1996), and to extend the model so that the conditional variance is also allowed to vary depending on the regime. Then work of Li and Li (1996) is also extended to use a GARCH rather than ARCH speci®cation for the conditional variance sine the former is more widely used and represents a more parsimonious way to represent persistent volatility series.…”

Threshold non-linear dynamics between Hang Seng stock index and futures returns

“…Consequently, this investigation adds to a developing literature that applies more complex modelling techniques (Eling and Toplek, 2009), such as Fractionally Integrated Error Correction models (Rajaguru and Pattnayak 2007), Exponential Generalised Autoregressive Heteroskedastic models (Butler and Okada 2008), Markov switching models (Okimoto 2008), or wavelet analysis (In and Kim 2006) Martens, Kofman and Vorst (1998) utilises a threshold error-correction model to investigate the complex dynamics between intraday Standard & Poors 500 futures and index returns, our main objective lies in the application of a Threshold AutoRegressive (TAR) model to the Hong Kong stock and futures markets. TAR models have been widely used in the finance and economics literature to explain various empirical non-linear phenomena that are observed in time series such as asymmetries and price jumps (Tong 1978(Tong , 1983Krager and Kugler 1993;Peel and Speight 1994;Tiao and Tsay 1994;Yadav, Pope, and Paudyal 1994;Potter 1995;Chappell et al 1996;Montgomery et al 1998;De Gooijer and Vidiella-i-Anguera 2003;Chan and Cheung 2005;Juvenal and Taylor 2008).…”

The Multi-Rule & Real-Time Training Neural Network Model for Time Series Forecasting Problem

“…④ Reference [3] suggested that this model improves the forecasting precision authentically. Alterable variance ⑤ model.…”