Threshold cointegration, nonlinearity, and frequency domain causality relationship between stock price and Turkish Lira


Creative Commons License

Kassouri Y., Altıntaş H.

RESEARCH IN INTERNATIONAL BUSINESS AND FINANCE, cilt.52, ss.11-22, 2020 (SSCI İndekslerine Giren Dergi) identifier

  • Cilt numarası: 52
  • Basım Tarihi: 2020
  • Doi Numarası: 10.1016/j.ribaf.2019.101097
  • Dergi Adı: RESEARCH IN INTERNATIONAL BUSINESS AND FINANCE
  • Sayfa Sayıları: ss.11-22

Özet

Modelling complex asymmetric effects and non-linear relationships between exchange rate and

stock prices has challenged classical econometric methods. This study contributes to the relative

literature in the following distinct ways. First, we follow a variety of econometric approaches in

order to characterize the complex dynamic co-movements between Turkish stock market and

exchange rate from January 2003 to December 2018. Secondly, we show that the evidence for

asymmetric threshold cointegration in Turkey’s financial market can be hidden by following

linear time series methodologies. Thirdly, it is also worth noting that the real effective exchange

rate, USD-Turkish lira exchange rates, money supply and interest rates have large predictive

power for stock price fluctuations at various frequencies. Building on these insights, we claim

that asymmetry (nonlinearity) is particularly important in Turkey’s financial market because it

shows the need for a new pattern of policy measures to prevent financial market crisis risk in

Turkey.