A long-run relationship between stock price index and exchange rate: A structural nonparametric cointegrating regression approach

In this paper we estimate the relationship between stock prices and exchange rates in EU and USA during the period of recent financial crisis (2008–2012) and compare the results with those in a previous period where stock markets were operating under normal conditions. According to the portfolio balance effect theory in periods of financial crisis there is a causal relationship from stock returns to exchange rate returns. Previous studies detect this relationship in short-run but not in long-run level. However, they use, for testing the long-run relationship, standard linear cointegrating regression models that suffer from biased estimations and cannot detect non-linear relationships. To overcome these problems, we examine the long-run relationship adopting a more advanced econometric model, the structural nonparametric cointegrating regression. The results exhibit a causal relationship from stock prices to exchange rates that is long-run in EU and short-run in USA. The finding of long-run relationship with this direction is particularly important because it is presented for first time in relative literature and shows the need for a new pattern of economic policy in EU.

► We investigate the existence of long-run relationship between stock prices and exchange rates in EU and USA during the recent financial crisis (2008–2012).
► We employ euro-dollar and FTSE Eurotop 300 and euro-dollar and Dow Jones index.
► We use a more sophisticated econometric method the structural nonparametric cointegrating regression.
► The results exhibit a causal relationship from stock prices to exchange rates that is long-run in EU.
► This finding shows the need for the adoption of a new pattern of economic policy in EU.