‘Currency war’ debates and growth
‘Currency war’ has been a recurring theme since the Global Crisis of 2008-09, as the monetary easing in advanced economies has pushed capital into emerging countries, fuelling appreciations of their currencies. The concern was raised by Brazil’s Finance Minister Guido Mantega, and underlies the ensuing debate between Fed Chairman Ben Bernanke and RBI Governor Raghuram Rajan on the spillover of US monetary policy (see Rajan 2014).
More broadly, policymakers often worry about the impact of appreciations due to their potentially damaging effect on competitiveness and domestic production. These discussions may be exacerbated by the sharp exchange rate movements recorded by the main global currencies in recent years (see Figure 1 for selected nominal effective exchange rates and Figure 2 for selected bilateral country pairs, taking a longer view). Interestingly, however, volatility is still not particularly high by historical standards (Figure 3).
Figure 1. Nominal effective exchange rates, selected economies
Figure 2. Bilateral nominal exchange rates, selected economies
Figure 3. Exchange rate volatility, selected bilateral nominal currency pairs
Note: standard deviation, 21-day rolling window.
Source: Reuters, Datastream.
The effect of appreciations on growth remains a much debated issue in the economic literature (e.g. Kappler et al. 2013). One of the reasons why this debate still hasn’t been settled is that this effect depends on the underlying factors that triggered the appreciation in the first place. For example, productivity gains may lead to an appreciation of the real exchange rate which, ceteris paribus, is detrimental to growth through the trade channel, but may have an overall positive effect on domestic production.
The exchange rate-growth link depends on the underlying shock
In a recent paper (Bussière et al. 2015), we undertake a broad empirical and theoretical assessment of the relation between appreciation and growth, paying particular attention to two underlying factors that may appreciate the exchange rate: a gain in domestic productivity and a surge in international capital flows, e.g. due to more accommodative monetary policy abroad.
- The main results suggest that the relation depends substantially of the underlying shock, with appreciations due to the surge of capital inflows being relatively less favourable for growth.
Our empirical analysis relies on a dataset of 68 advanced and emerging countries over 40 years. As growth and the real exchange rate are both endogenous variables, their co-movement is likely to depend on the underlying driver. Specifically, an appreciation due to higher productivity in the traded good sector is likely to be associated with higher growth, whereas an appreciation driven by a wave of capital inflows should be associated with less growth because of the adverse impact on international competitiveness. We therefore consider three types of appreciations: appreciations associated with a productivity boom, appreciations associated with a capital flow surge, and ‘simple’ appreciations.
The main well-known empirical challenge is to establish causality between exchange rate movements and growth, along with having a limited amount of information (observations). We address this by using a propensity score matching method to measure the average impact of the different appreciations on growth. This method allows to identify factors that affect the likelihood of a ‘treatment’ (an appreciation), and then compare countries that experience the treatments with countries that do not despite having a similar ex-ante likelihood. In other words, it constructs a counterfactual for the treatment, based on a set of observable macroeconomic characteristics such as GDP, VXO (Volatility Index of the S&P 100), commodity index, and inflation. To ensure the robustness of our results, we consider four alternative approaches in constructing the treated and control groups of countries.
Figure 4 provides the main point of our findings. It shows the ‘average treatment effect’ of an appreciation (i.e. the additional growth recorded by a country that experiences the ‘treatment’ compared to a similar country that doesn’t), depending on whether it happens alone (left panel), is associated with higher productivity (middle panel), or a capital inflows surge (right panel). Each panel shows the estimates for all countries, emerging economies, and advanced economies, reporting the average estimate across the four matching approaches, as well as the minimum and maximum estimates. As expected, the results differ depending on the type of countries (emerging versus advanced economies) and the circumstances of the appreciation.
Figure 4. Impact of an appreciation on growth
- Appreciations driven by higher productivity are linked to a clear increase in growth (the coefficients are statistically significant).
- While appreciations associated with higher capital flows are also associated with growth, the magnitude is much smaller and the coefficients are mostly not statistically significant.
- Appreciations taking place on their own are associated with somewhat higher growth in emerging countries (not statistically significant) and lower growth in advanced economies (significant).
In addition to these results, we also assess the impact of the appreciation by contrasting episodes with a productivity boom (or capital flows surge) and an appreciation against episodes with a boom (or surge) but no appreciation. This exercise shows that the appreciation per se reduces growth. Focusing on large appreciations (in search for potential non-linear effects) does not change the overall message.
These results highlight the necessity of understanding the underlying cause of the appreciation in order to assess its impact. As one could argue that the empirical pattern is not necessarily inefficient, and thus may not warrant a policy response, we develop a small open economy model generating a pattern of real exchange rate and growth consistent with our empirical findings. A key ingredient of the model is that fixed costs of production and a limited ability to create new firms in the short run can lead to inefficient firm destruction, following the work of Bacchetta and van Wincoop (2013).
The model shows that a policy response is not necessarily warranted depending on the shock considered. While an appreciation driven by a productivity boom leads to an efficient response, this is not the case for an appreciation driven by capital inflows. In the latter case, firms in the traded sectors face a loss of competitiveness that is not offset by higher productivity, leading to the shutdown of small firms. A social planner would recognise the inefficient feature of the shutdown and avert it. In addition, the relation between appreciation and growth is asymmetric; the weak growth observed during a capital flow driven appreciation does not imply that a mirroring depreciation is associated with high growth. Instead, the frictions in the model imply that:
- Exchange rate movements driven by capital flows shocks are always associated with low growth as they lead to inefficient firm destruction either in the traded sector (following an appreciation) or the non-traded sector (with a depreciation).
To the extent that firm destruction is quicker than firm creation, exchange rate movements may induce negative output movements in the short run. This supports policymakers’ concerns about both capital flows surges and sudden stops.
Our analysis shows that the concern about appreciations is well-founded when they are driven by shocks in global financial markets. Furthermore, being concerned about appreciations is not inconsistent with being concerned about depreciations. The proper policy response is, however, subtle. Specifically, the first best allocation following capital flows shocks does not dampen movements in the real exchange rate, but instead addresses the core friction of inefficient firm destruction. A policy aimed at the exchange rate could well be too blunt a tool to effectively address legitimate policy concerns.
Authors' note: The views expressed in this paper and those of the authors and do not necessarily reflect those of the affiliation with which they are affiliated.
Bacchetta, P, and E van Wincoop (2013), “The Great Recession: A Self-Fulfilling Global Panic”, mimeo, University of Virginia.
Bussiere, M, C Lopez and C Tille (2015), “Currency Crises in Reverse: Do Large Real Exchange Rate Appreciations Matter for Growth?”, Economic Policy 30(81).
Kappler, M, H Reisen, M Schularick and E Turkisch (2013), "The Macroeconomic Effects of Large Exchange Rate Appreciations," Open Economies Review 24(3), pp. 471-494.
Rajan, R G (2014), “Competitive Monetary Easing: Is It Yesterday Once More?” Speech at the Brookings Institution, Washington, D.C., 10 April.