The dynamic relation between foreign exchange rates and international portfolio flows: Evidence from Africa's capital markets Odongo Kodongo a , Kalu Ojah b, ⁎ a Department of Economics and Business Studies, Jomo Kenyatta University of Agriculture and Technology, P.O. Box 62000 City Square, Nairobi 00200, Kenya b Graduate School of Business Administration, University of the Witwatersrand, Private Bag 98 Wits, Johannesburg 2050, South Africa article info abstract Article history: Received 29 January 2011 Received in revised form 9 January 2012 Accepted 10 January 2012 Available online 18 January 2012 We examine the nexus between real foreign exchange rates and international portfolio flows using monthly data for the period 1997:1 to 2009:12 for Egypt, Morocco, Nigeria, and South Africa. We analyze the full sample period and two sub-periods, distinguished by the relative volume and volatility of portfolio flows. We find international portfolio flows, in Africa, to be non-persistent and relatively volatile. Granger causality tests and innovation accounting from vector autoregressions show that the dynamic relationship between portfolio flows and foreign exchange rates is both country-dependent and time-varying; and these findings are robust to alternative VAR specifications. © 2012 Elsevier Inc. All rights reserved. JEL classification: G15 F21 F31 F32 Keywords: Real exchange rates Net portfolio inflows Africa's capital markets 1. Introduction How volatile and/or persistent are portfolio flows? Do fluctuations in real exchange rates contain information that informs the de- cisions of international portfolio investors? Do international portfolio investment flows have a discernible influence on observed changes in the real exchange rates? This study is the first serious attempt at answering these questions in the African context. In the- ory, the relationship between balance of payments items and foreign exchange rates can be explained by two major strands of the literature. On the one hand are the traditional macroeconomic models of exchange rates determination categorized as the monetarist approach (Frenkel, 1976), the portfolio balance approach (Dooley & Isard, 1979), or a variety of general equilibrium models (Stockman, 1980). The traditional models focus largely on medium- to long-run foreign exchange rate determination but have not performed very well in empirical tests (Cheung, Chinn, & García Pascual, 2005; de Blas, 2010; Meese & Rogoff, 1983; Yuan, 2011). 1 On the other hand, the many failures of macro-based models to explain exchange rates changes have given rise to the bur- geoning strand of literature that springs from market microstructure. The microstructure approach, concerned with the empirical link between order flow in foreign exchange markets and foreign exchange rate changes, focuses on the short-run, and has shown that foreign exchange order flow 2 indeed predicts exchange rate movements (e.g., Daníelsson, Payne, & Luo, 2002; Evans & Lyons, International Review of Economics and Finance 24 (2012) 71–87 ⁎ Corresponding author. Tel.: +27 11 717 3764; fax: +27 11 717 3849. E-mail addresses: kodongo03@gmail.com (O. Kodongo), kalu.ojah@wits.ac.za (K. Ojah). 1 However, evidence that some macroeconomic factors may have some predictive power on exchange rates exists in recent literature. For instance, Qiu, Pinfold, and Rose (2011) demonstrate that purchasing power parity has a predictable and sizable influence on future exchange rates in market-based economies with freely floating exchange rates. 2 Order flow, a measure of buying/selling pressure, is the net of buyer-initiated orders and seller-initiated orders. 1059-0560/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.iref.2012.01.004 Contents lists available at SciVerse ScienceDirect International Review of Economics and Finance journal homepage: www.elsevier.com/locate/iref