No abstract

No abstract

The extant literature on exchange rate forecasting on the basis of the Dornbusch-Frankel, Frenkel-Bilson and Hooper-Morton models prominently reveals the dominance of the autoregressive models over the theory-based models. Some studies have however attempted to upturn the results by including the lagged dependent variable in the theory-based models which somewhat implies comparing a modified random walk with a traditional random walk. We follow a different approach both in terms of theory and methodology. We offer an innovative exposition of the Portfolio Balance theory to stock price-exchange rate nexus. Consequently, a predictive model for exchange rate where stock price is a predictor is formulated. The formulated model is expressed in both linear and nonlinear form in order to account for the role of asymmetric changes in stock prices in exchange rate forecasting. Thereafter, we employ the Lewellen (2004) and Westerlund and Narayan (2014) methods which account for any inherent statistical properties of the predictors. Our results validate the Portfolio Balance theory where we show that the sector-level stock prices consistently turn up as good predictors of the exchange rates. The predictive model proposed in this work does not require the inclusion of a lagged dependent variable to beat the autoregressive models which is the practice in the existing literature. We further demonstrate that asymmetry matters to a large extent in the nexus both for the in-sample and out-of-sample predictability.

No abstract

PurposeThis paper examines the return and volatility spillovers among major cryptocurrency using daily data from 10/08/2015 to 15/04/2018.Design/methodology/approachThe authors employ the Dielbold and Yilmaz (2012) spillover approach and rolling sample analysis to capture the inherent secular and cyclical movements in the cryptocurrency market.FindingsThe authors show that there is substantial difference between the behaviour of the cryptocurrency portfolios return and volatility spillover indices over time. The authors find evidence of interdependence among cryptocurrency portfolios given the spillover indices. While the return spillover index reveals increased integration among the currency portfolios, the volatility spillover index experiences significant bursts during major market crises. Interestingly, return and volatility spillovers exhibit both trends and bursts respectively.Originality/valueThis study makes a methodological contribution by adopting Dielbold and Yilmaz (2012) approach to quantify the returns and volatility transmissions among cryptocurrencies. To the best of our knowledge, little or no study has adopted the Dielbold and Yilmaz (2012) methodology to investigate this dynamic relationship in the cryptocurrencies market. The Dielbold and Yilmaz (2012) approach provides a simple and intuitive measure of interdependence of asset returns and volatilities by exploiting the generalized vector autoregressive framework, which produces variance decompositions that are unaffected by ordering.