International Journal of Economics and Financial Issues ISSN: 2146-4138 available at http: www.econjournals.com International Journal of Economics and Financial Issues, 2018, 8(5), 210-219. International Journal of Economics and Financial Issues | Vol 8 • Issue 5 • 2018 210 Gold - Silver Nexus: A Threshold Cointegration Approach Zouheir Ahmed Mighri 1 *, Majid Ibrahim Al Saggaf 2 1 Department of Finance and Insurance, College of Business, University of Jeddah, Saudi Arabia, 2 Department of Finance and Insurance, College of Business, University of Jeddah, Saudi Arabia.*Email: zmighri@gmail.com ABSTRACT We investigate the dynamic relationship between the gold and silver prices using the Enders-Siklos threshold cointegration approach. Our data are the weekly prices of the gold and silver from January 1968 to May 2016. We fnd a, asymmetric threshold cointegration between these two series, instead of the linear cointegration well established in the literature. The short-term adjustment to the equilibrium shows an asymmetric effect according to the price deviation from the long-run equilibrium. Moreover, using an equilibrium adjustment path asymmetry test, we fnd that, in the short term, gold has a much faster reaction to negative deviations from long-term equilibrium than positive deviations. Keywords: Threshold Cointegration, Price Transmission, Gold, Silver JEL Classifcations: C13, C22, C32, C52, C53, G15. 1. INTRODUCTION Precious metals such as gold and silver are strategic commodities whose prices have received much attention. They have been used as a store of value and currency for thousands of years, suggesting that there is a long-run relationship between the two precious metals (Baur and Tran, 2014). However, several factors may drive their prices away from each other, namely the industrial demand for silver and jewelery, and the dental demand and central bank demand for gold 1 . The recent popularity of commodities as an investment and a hedge (Capie et al., 2005; Levin and Wright, 2006; Baur and Lucey, 2010; Baur and McDermott, 2010) against adverse fnancial or economic events may constitute an additional force that either creates an otherwise non-existent long-run relationship or strengthens a preexisting long-run relationship (Batten et al., 2013). The commodity markets are broadly characterized by movements in their prices that naturally depend on a number of exogenous and endogenous factors. These movements may be upwards or downwards in response to changes in the predictors. Escribano and Granger (1998) analyzed the relationship between gold and 1 For more details, see World Gold Council (www.gold.org). silver prices and found that gold and silver are co-integrated mainly due to a specifc bubble and post-bubble period. However, the magnitude of positive and negative responses may differ for similar positive and negative variations in the predictors, in which case the variables display asymmetric adjustment over the business cycle. Cointegration and causality analyses are broadly used by some studies to investigate the interdependence and the long-run cointegration relationships between gold and silver prices. Most of these studies either use data that make a comparison with the Escribano and Granger (1998) study either impossible or analyze samples that are too short to answer the questions raised by the authors (Adrangi et al., 2000; Ciner, 2001; Lucey and Tully, 2006; Liu and Chou, 2003; Hammoudeh et al., 2010; Baur and Tran, 2014; Kucher and McCloskey, 2017; Eryiğit, 2017). The standard cointegration analysis (Engle and Granger, 1987) assumes that the adjustment mechanism of the error correction term is symmetric. This indicates that the adjustment coeffcients are the same no matter if the equilibrium error is positive or negative. That is, the adjustment speed of prices is the same regardless of the kind of shocks. Nevertheless, using positive and negative error terms to denote the positive returns (good news) and negative returns (bad news), the adjustment speed of prices may be slower for positive shocks than for negative ones.