«OPTIMAL EXCHANGE RATE POLICY IN A GROWING SEMI-OPEN ECONOMY Philippe Bacchetta, Kenza Benhima and Yannick Kalantzis HKIMR Working Paper No.09/2014 ...»
HONG KONG INSTITUTE FOR MONETARY RESEARCH
OPTIMAL EXCHANGE RATE POLICY IN A GROWING
Philippe Bacchetta, Kenza Benhima and Yannick Kalantzis
HKIMR Working Paper No.09/2014
Hong Kong Institute for Monetary Research
(a company incorporated with limited liability)
All rights reserved.
Reproduction for educational and non-commercial purposes is permitted provided that the source is acknowledged.
Optimal Exchange Rate Policy in a Growing Semi-Open Economy* Philippe Bacchetta University of Lausanne Centre for Economic Policy Research Hong Kong Institute for Monetary Research and Kenza Benhima University of Lausanne Centre for Economic Policy Research and Yannick Kalantzis Banque de France May 2014 Abstract In this paper, we consider an alternative perspective to China's exchange rate policy. We study a semi-open economy where the private sector has no access to international capital markets but the central bank has full access. Moreover, we assume limited financial development generating a large demand for saving instruments by the private sector. We analyze the optimal exchange rate policy by modelling the central bank as a Ramsey planner. Our main result is that in a growth acceleration episode it is optimal to have an initial real depreciation of the currency combined with an accumulation of reserves, which is consistent with the Chinese experience. This depreciation is followed by an appreciation in the long run. We also show that the optimal exchange rate path is close to the one that would result in an economy with full capital mobility and no central bank intervention.
JEL Classification: E58, F31, F32, F38 * Philippe Bacchetta is a Swiss Finance Institute Professor at the University of Lausanne and a CEPR Research Fellow.
Kenza Benhima is a Professor at the University of Lausanne and a CEPR Research Affiliate. Yannick Kalantzis is a Senior Economist at the Banque de France. The authors would like to thank two referees, Menzie Chinn, Roberto Chang, Nicolas Coeurdacier, Giovanni Lombardo, Charles Engel, as well as participants at seminars at the City University of Hong Kong, the Hong Kong Monetary Authority, Università di Napoli Federico II, the London Conference on International Capital Flows and Spillovers, the Bank of Canada--ECB workshop on exchange rates, and the Summer Workshop in International Finance and Macro Finance, for useful comments. Bacchetta and Benhima gratefully acknowledge financial support from the National Centre of Competence in Research "Financial Valuation and Risk Management" (NCCR FINRISK). Bacchetta also acknowledges support from the ERC Advanced Grant #269573, and the Hong Kong Institute for Monetary Research.
The views expressed in this paper are those of the authors, and do not necessarily reflect those of the Banque de France, the Eurosystem, the Hong Kong Institute for Monetary Research, its Council of Advisers, or the Board of Directors.
Hong Kong Institute for Monetary Research Working Paper No.09/2014
In recent years we have seen a heated debate on Chinese exchange rate policy and the enormous accumulation of international reserves by its central bank. While the increase in reserves has been considered as a major contributor to global imbalances, the renminbi (RMB) has typically been viewed as undervalued. For example, Frankel (2010) clearly states "An appreciation would improve economic welfare". However, these views are not universally shared. For example, McKinnon (2010) gives two main arguments against more RMB flexibility. First, a flexible exchange rate is not desirable given the limited international use of the RMB. Second, an appreciation will not necessarily reduce the huge current account surplus, unless it reduces the difference between aggregate saving and aggregate investment.
This paper will focus on the second argument of McKinnon, namely the connection between the exchange rate level and net saving. We examine the optimal exchange rate policy in a dynamic intertemporal model that incorporates four basic features of the Chinese economy: i) limited capital mobility; ii) a net capital outflow taking the form of an accumulation of central bank international reserves; iii) underdeveloped financial markets; iv) a very high growth rate. Growth is assumed to arise from exogenous increases in endowments. In such a context the central bank is modeled as a Ramsey planner who can choose the optimal path of the exchange rate and of international reserves.
Our main result is that in a growth acceleration episode it is optimal to have an initial real depreciation of the currency combined with an accumulation of reserves. This depreciation is followed by an appreciation in the long run. We also show that the optimal exchange rate path is close to the one that would result in an economy with full capital mobility and no central bank intervention. The main reason for an optimal depreciation is financial underdevelopment implying a limited supply of financial assets. With a developed financial system, an initial appreciation would be optimal.
Studying the link between the real exchange rate and net saving naturally requires an intertemporal approach, in contrast to many analyses that examine the relationship between the exchange rate and the trade balance. The standard model analyzing this link is the representative-individual infinitehorizon model with traded and non-traded goods. We deviate from this benchmark model to incorporate the four features mentioned above. First, we assume low financial development, in the form of credit constraints. This may significantly affect saving behavior, especially with high growth rates. We follow Woodford (1990) and introduce credit-constrained heterogeneous households who alternate between high and low endowments. With strong credit frictions, low-endowment households cannot borrow and therefore have an incentive to save more in high-endowment periods. With a
growth acceleration episode, the desire to save is even higher as households will feel even more constrained in future low-endowment periods.
In an open economy, household could save more by buying foreign assets. This would allow aggregate saving to increase and would lead to a current account surplus. Moreover, this would initially depreciate the real exchange rate: a higher saving rate reduces demand and pressure on domestic prices, which implies a real depreciation. However, we consider a "semi-open" economy, which is an economy where the private sector does not have any access to the international capital market, but the central bank does. Therefore, there is a lack of financial assets available for 3,4 consumers. In this context of an "excess" demand for saving, or asset scarcity, the government or the central bank can provide domestic assets to accommodate the saving need. A natural way of changing the amount of domestic assets is for the central bank to serve as intermediary between the international capital market and domestic savers. Thus, an accumulation of international reserves at the central bank can be translated into an increase in the supply of domestic assets and an increase in private saving.
This policy will also affect the real exchange rate. As in the open economy, a higher saving rate implies a real depreciation. Therefore the optimal exchange rate policy is directly tied to asset provision and reserve policy. A natural question will be to compare the optimal policy in the semiopen economy to the decentralized equilibrium in the open economy.
To analyze the optimal exchange rate policy in the context of a semi-open economy, we take a dynamic optimal taxation approach, by modelling the central bank as a Ramsey planner. The central bank takes the government behavior as given and therefore has much fewer instruments than in standard optimal taxation analyses. Although this approach has not been used for exchange rate policy (and even less in the Chinese context), there is growing interest in using these tools in international macroeconomics. In a growth acceleration episode, we find that it is optimal to increase the supply of domestic assets financed by international reserves. Therefore it is also optimal to let the currency depreciate. We also find that the optimal depreciation with capital controls is close to the depreciation that would occur in an open economy. The need for reserve accumulation and currency This implies that a capital account liberalization would lead to a net private capital outflow. Several papers in the literature predict such an outcome for China using totally different perspectives. E.g., see He et al. (2012a).
This simple framework with credit constraints enables to capture two key features found in the recent literature on global imbalances: insufficient supply of domestic assets (see Caballero et al., 2008) and precautionary saving (see Mendoza et al., 2009).
Since 2000, the Chinese central bank has increased its liabilities with the domestic banking sector at about the same rate as international reserves. These liabilities mainly take the form of central bank bonds and commercial banks reserves.
In a recent paper, Jeanne (2012) considers a semi-open economy with traded and non-traded goods and shows how exogenous changes in international reserves alter intertemporal consumption choices, as well as the real exchange rate.
Farhi et al. (2012) show that a simple combination of taxes can replicate nominal exchange rate policy, but they do not consider a Ramsey planner.
Hong Kong Institute for Monetary Research Working Paper No.09/2014 depreciation will be stronger when the lack of saving instruments is acute, i.e., with low financial development.
The structure of the semi-open economy is similar to Bacchetta et al. (2013), but we consider traded and non-traded goods to determine real exchange rate movements. In our previous paper with a single good, the optimal policy was determined by various trade-offs caused by changes in the interest rate. Indeed, in the presence of credit constraints and growth, the role of policy is to help agents consume more in early periods, when the constraint is more binding. This can be achieved by either a high or a low interest rate, depending on the level of growth, risk and credit constraint. The optimal semi-open economy then consists in accumulating more or less reserves than the open economy. The introduction of the real exchange rate adds an incentive to appreciate the currency in order to stimulate income and the value of collateral in early periods, which implies that the central bank would tend to accumulate less reserves. Our results suggest that this either mitigates or reinforces the policy induced by the interest rate trade-offs, but does not dramatically alter optimal policy. Exchange rate dynamics is more a consequence of reserve policy than a driver of that policy.
Our results are broadly consistent with the experience of China after it joined the World Trade Organization (WTO) in December 2001. Figure 1 documents the relevant stylized facts. First, China experienced a growth acceleration: the growth rate of GDP per capita increased from 7% in 2001 to 14% in 2009. Second, the central bank started accumulating large amounts of international reserves, from 16% of GDP in 2001 to almost 50% at the end of the decade. Third, over that same period, the real effective exchange rate initially depreciated, from 2001 to 2005, before appreciating after 2005.
Fourth, these dynamics coincided with an increase in aggregate net saving as represented by the current account (from 1.3% of GDP in 2001 to 10% in 2007), consistent with the mechanism we describe below.
A standard perspective for the last decade is that China experienced an exogenous increase in export demand. To prevent a nominal appreciation, the central bank intervened in the foreign exchange market and accumulated reserves. Moreover, it sterilized the increase in reserves to avoid inflation.
Our alternative perspective starts from an exogenous growth acceleration that increases saving by Chinese consumers, mainly in the form of bank deposits. The implied increase in liabilities of the Chinese banking sector was translated into an increase in central bank liabilities, through required reserves and central bank bills. In this context, the optimal policy of the central bank is to purchase foreign currency assets and to let the real exchange rate depreciate. Thus, the central bank served We focus on the years 2000 as China was not truly a market economy until the late nineteen-nineties. For instance, a significant share of producer and retail prices were not market-determined until the second half of the nineties. The People's Bank of China only became an autonomous central bank in the modern sense after a law was passed in March
1995. See OECD (2009) for details on the reform process.
We consider a real model and do not model inflation explicitly. Introducing a nominal sector with flexible prices would allow to distinguish between nominal and real exchange rate fluctuations, but would not change our main analysis. Notice, however, that the nominal trade-weighted RMB has moved closely to its real value since 2000.
Hong Kong Institute for Monetary Research Working Paper No.09/2014 as intermediary between the private sector and the international capital market, as argued in particular by Song et al. (2011). We notice that both the standard and our alternative approach are consistent with the increase in international reserves and current account illustrated in Figure 1.
However, the standard perspective is not consistent with the real depreciation between 2001 and 2005, as an increase in export demand should lead to a real appreciation.