رابطه پویا میان نرخ های ارز و جریان های پرتفوی بین المللی : شواهدی از بازار سرمایه آفریقا
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|14564||2012||17 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : International Review of Economics & Finance, Volume 24, October 2012, Pages 71–87
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.
How volatile and/or persistent are portfolio flows? Do fluctuations in real exchange rates contain information that informs the decisions 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 theory, 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 et al., 2005, de Blas, 2010, Meese and Rogoff, 1983 and Yuan, 2011).1 On the other hand, the many failures of macro-based models to explain exchange rates changes have given rise to the burgeoning 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 flow2 indeed predicts exchange rate movements (e.g., Daníelsson et al., 2002 and Evans and Lyons, 2002). Further, there is evidence that the order flow is the channel through which macroeconomic information that is publicly and simultaneously released to all market participants (Love & Payne, 2008), and expectations of future macro-variables (Evans & Lyons, 2007), are partially impounded into asset prices and exchange rates. Given the empirical success of the microstructure approach, there is clear need to better understand the link between balance of payments movements (probably a major cause of order flows in the foreign exchange markets) and exchange rate movements. Indeed, as Hau and Rey (2006) argue, the foreign exchange order flow is itself tied to endogenous portfolio flows which emerge under optimal dynamic investment in an incomplete market setting; thus, the exchange rate dynamics is based largely on the financial market structure rather than traditional macroeconomic determinants. This argument is strengthened by Müller-Plantenberg (2010) who explains that the exchange rate is driven by the demands and supplies of currencies in the foreign exchange market which, in turn, are influenced by the different types of transactions that economic agents undertake such as sales of domestic financial assets. Empirically, many studies have found a relationship between portfolio flows and exchange rates. Froot, O'Connell, and Seasholes (2001) find a positive contemporaneous covariance between net portfolio inflows and dollar equity and currency returns; they attribute the co-movement to the information content of flows. Siourounis (2004) finds evidence suggesting that a good portion of exchange rate movements for Germany, Switzerland, the UK and USA can be explained by net cross-border equity flows. Brooks, Edison, Kumar, and Sløk (2004) find that inflows into the US equity markets and direct investment flows finance the current account deficit and allow the dollar to remain strong; conversely, large and initially unanticipated outflows from the euro area appear to account for a substantial part of the dollar's fall and persistent weakness. More recently, Heimonen (2009) presents results indicating that an increase in the euro area equity returns with respect to USA equity returns causes an equity capital outflow from the euro area to the USA; and that this equity flow generates an order flow in the foreign exchange markets, which leads to an appreciation of the US dollar.3 Emerging market economies have also attracted some studies. The influential work of Edwards (1998), conducted with quarterly data of Latin American countries, shows that in seven out of the eight countries studied, it is not possible to reject the hypothesis that capital flows influence real exchange rates; and, in three of the eight countries, it is not possible to reject the two-way causality hypothesis. Strikingly, the data shows no evidence that the real exchange rate causes capital inflows. Kim and Singal (2000), with data from twenty emerging market economies over 1976–1996, find that foreign investment flows exert a calming influence on foreign exchange rate volatility; the resulting lower currency risk in turn, encourages foreign investors to invest more at a lower required rate of return. Jongwanich (2010) finds that both portfolio investment and other financial assets investment (including bank loans) elicit a faster speed of real exchange rate appreciation than foreign direct investment inflows. In the African region, where developments in international capital flows have not received much research attention, the literature is thin. Ndung'u and Ngugi (1999) find that a unit shock in the volatility of Kenya's capital flows leads to an initial decline in the real exchange rate followed by a continuous rise, with no signs of the effects dying out: however, volatility in capital flows only accounts for a small percentage of the innovations from the real exchange rate, implying that there is weak feedback from the real exchange rate movements to the volatility in capital flows. Buffie, Adam, O'Connel, and Pattillo (2004) observe that incipient capital inflow to Africa places acute short-run pressure on the foreign exchange market, dramatically undermining the case for a floating exchange rate. Early studies, conducted on the back of the liberalization campaign by the Bretton Woods institutions, focused largely on official development assistance (e.g., Kasekende, Kitabire, & Martin, 1996). Other studies have investigated aggregate private capital flows of Africa (Buffie et al., 2004 and Ndikumana, 2003) but have not decomposed them into foreign direct investments (FDI) and portfolio flows. These omissions, commonly ascribed to data difficulty, have meant that policy measures cannot be prescribed that touch specifically on the various components of the financial account. Therefore, in order to aid the formulation of directed policy, it is important to understand the linkages between real exchange rates and portfolio flows in Africa.4 Portfolio flows are chosen for this analysis for two reasons. First, portfolio flows, by their nature, are more temporary and more volatile than foreign direct investment and “other investment flows” (see e.g., Kodongo & Ojah, 2011) and are therefore likely to pose a greater challenge to monetary policy if they inform changes in real exchange rates. Second, the literature has concentrated on the relationship between real exchange rates and capital flows in general (in the aggregate), leaving a knowledge gap on the role of portfolio flows, in isolation, on real exchange rates changes, and vice versa. Because they are relatively less explored, Africa's capital and foreign exchange markets provide a fitting laboratory for examining these possible important linkages. The rest of this paper is structured as follows. Section 2 presents the data and examines their basic characteristics. Section 3 presents and discusses results of our baseline tests and robustness checks. Section 4 concludes.
نتیجه گیری انگلیسی
We investigated the dynamic relationship between real exchange rates and net portfolio inflows of four major countries in Africa. The results suggest that international portfolio flows to African countries are characterized by high volatility and low persistence. Overall, our results do not indicate a direction of causality between real exchange rates and net portfolio inflows that can be generalized across all the countries investigated. Rather, causality between the two macroeconomic variables appears to be both country-specific and time-varying. These results are consistent with Morrissey, Lloyd, and Opoku-Afari (2004), who find that neither “permanent” (foreign direct investments, remittances, grants) nor “non-permanent” (loans, portfolio investments) capital inflows have a short-run effect on the real exchange rate in Ghana. Our results also mirror Ndung'u and Ngugi (1999), who find only a weak feedback from real exchange rate movements to Kenya's capital flows volatility. The country-specific nature of the documented relationship may be attributed to idiosyncrasies around attractiveness of the specific capital markets to foreign investors, as explained by their perceived openness to international capital flows and financial markets sophistication. Thus, the South African capital market, which exhibits the highest level of sophistication of the markets investigated,10 is more likely to command greater awareness of foreign investors than the other markets. As such, the decisions of international portfolio investors in South Africa's financial markets are likely to be informed by changes in macroeconomic fundamentals including fluctuations in foreign exchange rates. Similarly, because of her relatively more flexible foreign exchange regime, South Africa's exchange rate fluctuations may draw more closely from the flow of international money. This may explain the bidirectional causality finding for South Africa for the full sample period. Next, it is important to point out that the order flow in the foreign exchange market depends both on the financial account and on the current account of the balance of payments. As Müller-Plantenberg (2010) explains, the balance of payments consists primarily of the current account and the financial account, with the latter composed of monetary flows and non-monetary flows such as portfolio flows, foreign direct investment, bank lending and other investments. Given that capital flows, and in particular portfolio flows, are far more volatile, in general, than the current account, it is natural to expect significant correlations between portfolio flows and exchange rate movements. However, in the medium-term or longer-term horizons, whether the exchange rate moves in one way or the other depends crucially on whether net trade flows outweigh net capital flows or vice versa. That is, there is the likelihood that the relative persistent dominance of the current account (i.e., current account surplus/deficit) might counter or even trump the portfolio flow effects. Thus the asymmetries between balance of payments items may explain the relatively weak evidence of causality in our findings. Finally, expectations may also play a role. For example, if, for some reasons, portfolio inflows are expected to rise at some future date, the exchange rate would have to rise now rather than in the future; otherwise, investors could make relatively riskless profits by buying the domestic currency now and waiting for it to appreciate. Given their non-persistence and high volatility, it is relatively difficult, in general, to predict future portfolio flows; with the possibility of anticipated exchange rate changes coming into the mix of things, it is even more difficult to reach conclusions about the empirical causal nexus between the two variables.