Impact of Exchange Rate and Firm Heterogeneity on Exports: Empirical Evidence from Four ASEAN Economies.

• Author: Xie, Yizhe Daniel

1. Introduction

   Ever since China's labour costs have started escalating and the country's growth began slowing down, ASEAN has come under the global spotlight. The Association, composed of ten countries, (1) has a combined population of 650 million, making it the fourth most populous economy, while its nominal GDP stands at US$2.8 trillion, placing it sixth in the list of largest economies. In terms of international trade, it is the fourth largest exporting region--after the European Union, North America, and China/Hong Kong. The grouping plays a central role in Asian economic regional integration initiatives and its member economies are part of major regional free trade agreements such as the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP), Regional Comprehensive Economic Partnership (RCEP), ASEAN+1, inter alia.

   As ASEAN leaders seek to replicate China's economic growth miracle through export-driven expansion, the exchange rate, which measures the relative price competitiveness of a country's exports, has emerged as an extremely important factor. In fact, devaluation of the Chinese RMB is generally believed to be a key reason for the success of the country's exports (Ahmed 2009; Aziz and Li 2007; Cheung, Chinn and Fujii 2009). The Chinese evidence is consistent with the conventional view that depreciation or devaluation of a local currency will lead to more exports by making domestic goods more competitive in the international market. Not only in China, but the quick recovery seen in Europe and Japan following the Second World War, too, was largely attributed to the devaluation of their currencies (Eichengreen 2005, 2006). Moreover, currency deprecation has been found to have hastened the economic recovery process during the Great Depression of the 1930s (Eichengreen and Sachs 1985). In a nutshell, there is enough evidence to show that exchange rate devaluation is a useful tool to boost exports.

   Interestingly, in recent years, many ASEAN countries seem to defy this trend, as their total volume of exports has increased despite a steady appreciation of the real effective exchange rate (REER). As shown in Figure 1, between 2005 and 2015, five ASEAN countries (Indonesia, the Philippines, Vietnam, Laos and Cambodia) saw an upward trend in their respective REERs, which indicates that the exports from these economies were losing their price competitiveness (or were witnessing an increase in export prices) ceteris paribus at a certain level of exchange rate pass-through, but their aggregated exports were expanding almost throughout this period. With the exception of Indonesia, which saw some fluctuations, this is seen in all the other countries.

   This distortion in the trade and exchange rate nexus observed in these countries using aggregated data can have important implications for policymakers and researchers in the region and beyond. The observation suggests that the established link between trade and exchange rates may not hold true any longer, at least not in the rising ASEAN region. In other words, the devaluation or depreciation of a local currency cannot be seen either as a policy tool or a windfall for economies that bet on an export-based strategy for economic development.

   Therefore, it is meaningful and timely to ask whether, at a micro-level, the export performance of ASEAN firms is immune to the impact of exchange rate appreciation. And, if that is not the case, what type of firms are more sensitive to exchange rate movements? At the point of writing, no published studies have been found to address this particular phenomenon, especially using firm-level data.

   In order to fill the gap, this paper uses firm-level panel data from four ASEAN countries (Indonesia, Vietnam, the Philippines and Laos), as collected by the World Bank's IFC Enterprise Survey for selective years between 2009 and 2016. Due to the "zero trade" problem in the dataset, the Poisson-pseudomaximum likelihood (PPML) approach is employed as the main estimation method. Supplementary empirical analysis techniques, including Tobit, ad hoc solution and Heckman's two-stage model are also used in this study.

   For the baseline econometric analysis, change in a firm's real exports is assumed to be a function of the real effective exchange rate level and its volatility and firm characteristics (firm size and labour productivity), while controlling for unobserved time-invariant effects of year, country and sector. Next, the equation is extended by including other firm heterogeneities as explanatory variables, including foreign ownership, internationally recognized quality certification, and financial obstacles. Alternative exchange rate measurement tools--nominal effective exchange rate (NEER) and US dollar-local currency exchange rate--and data decomposition are employed for robustness checks. Finally, the interaction terms--exchange rate and firm heterogeneities (SMEs, prior export experience, foreign and domestic affiliation, foreign content in exports)--are constructed to analyse how firms with different characteristics respond to exchange rate appreciation.

   The findings of this study suggest that, first, contrary to conventional wisdom, exchange rate appreciation actually discourages exports of ASEAN-based companies. Second, firms' responses to currency appreciation vary; while SMEs and first time exporters are more fragile when it comes to exchange rate movements, exporters can reduce such risks through foreign or domestic affiliations (although foreign ownership proves to be more helpful). Third, firms whose exports consist of foreign inputs are less affected by the rise in the value of local currency. And fourth, firms in the services sector are more sensitive to currency appreciation than those in manufacturing.

   In addition to the studies mentioned above, the results of this paper are also related to literature on: firm heterogeneity in export decision (Melitz 2013; Bernard and Jensen 1999 and 2004); exchange rate, sunk cost, and hysteresis in trade (Baldwin and Krugman 1989; Dixit 1989; Campa 2004); and the connection between exchange rate and trade (Leigh et al. 2015; Amiti, Itskhoki, and Konings 2014; Bernini and Tomasi 2015).

   This paper is structured as follows. The next section covers a brief overview of the existing literature, while the third section focuses on the data used in the study. The fourth section includes the empirical framework and methodology. The empirical results are presented in the subsequent section. Limitations of the research are acknowledged in the sixth section, and the final section concludes.

2. Literature Review

   Real exchange rate movements affect exports through two channels: first, via the level/magnitude of the real exchange rate; and second, via real exchange rate volatility. The former is an important component of relative price for firm exports, while the latter is a vital element of the fixed costs incurred by firms to participate in international trade. Although there are many studies on the impact of exchange rate on exports, the vast majority of them do not--theoretically or empirically--conclude whether or not the level of real exchange rate and real exchange rate volatility increase exports. Given this background, this section reviews some of the existing studies on the effect of real exchange rate depreciation and exchange rate volatility on exports.

   In theory, real exchange rate depreciation is said to lower export prices and raise import prices (Krugman 1999; Kandil and Mirzaie 2002; Greenaway, Kneller, and Zhang 2010). Therefore, the effect of exchange rate depreciation on exports depends on this offset effect. A large number of recent empirical studies also show that real exchange rate depreciation increases exports (Bahmani-Oskooee and Kara 2003; Hausmann, Pritchett, and Rodrik 2005; Ng et al. 2008; Freund and Pierola 2012). In literature devoted to the Asian region, too, many studies find that real exchange rate depreciation increases exports. Fang, Lai, and Miller (2006), for example, find that real exchange rate depreciation has a significant, positive effect on export growth of seven Asian countries (Indonesia, Japan, Korea, Malaysia, the Philippines, Taiwan and Thailand), but not Singapore. Jongwanich (2009), who analysed eight countries (China, Hong Kong, India, Indonesia, Korea, Malaysia, Singapore and Thailand), also notes that real exchange rate depreciation generates a higher volume of exports. Mirroring the opposite trend of increase in exports on account of real exchange rate depreciation, many studies also document the negative effect of real exchange rate appreciation on exports (Bugamelli and Infante 2003; Campa 2004; Das, Roberts, and Tybout 2007; Greenaway, Kneller, and Zhang 2008; Thorbecke and Smith 2010). In contrast, Abeysinghe and Yeok (1998) and Wilson and Tat (2001) argue that real exchange rate appreciation does not reduce the volume of exports in some Asian countries. Likewise, Bernard and Jensen (2004) analysed US exports and Wilson (2001) studied the export markets of Singapore, Malaysia and Korea, and both studies report that real exchange rate does not have a significant impact on bilateral trade.

   A large number of studies also investigate the effect of real exchange rate volatility on exports. The traditionally accepted view is that exchange rate volatility is negatively associated with trade flows, given the idea of uncertainty (Ethier 1973; Arize 1997). In fact, in the "new" new trade theory (NNTT) developed by Melitz (2003), exchange rate volatility serves as a fixed cost for exporting firms. Specifically, higher fixed costs from high exchange rate volatility negatively affect firm exports.

   However, Broil and Eckwert (1999) highlight the theoretical possibility of a positive relationship between exchange rate volatility and exports, as higher exchange rate volatility may also increase the real option to export to the...