Does Exchange Rate Volatility Dampen Imports? Commodity-Level Evidence From India

ABSTRACT

This paper studies the effect of exchange rates’ volatility on India’s imports on a balanced panel of 73 commodities spanned from April 2013 to October 2016. Rather than using cross-country bilateral import flows, we test the relationship at the commodity level using disaggregated trade data with monthly frequency. Generalized autoregressive conditional heteroscedasticity model is used for estimating exchange rate. We employ pooled mean group estimator for simultaneously assessing long- and short-run association between nominal exchange rate volatility and import volume. In the long-run, for all commodities, a 100% increase in volatility results in a 12% drop in India’s imports. A significant dampening impact of volatility of exchange rate on imports is evidenced also in short-run. However, at the disaggregate level, imports in the agricultural and allied sector are found to be relatively more sensitive to exchange rate volatility as compared to the manufacturing sector. We also conducted a time series analysis for the aggregate data covering both pre- and post-crisis period. The results validate the findings of commodities-level panel data analysis. This paper concludes with policy implications of our findings.

KEYWORDS:

• Exchange rate volatility
• imports
• pooled mean group
• India

JEL CLASSIFICATIONS:

• C33
• F10

Acknowledgements

The authors thank two anonymous referees for their useful comments and helpful suggestions on the previous version of this paper. Any errors or omissions are solely the authors.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Chandan Sharma http://orcid.org/0000-0002-0625-5237

Notes

1 The share of imported products subject to quantitative restrictions in India came down substantially from 87% during in 1987 to 45% during 1994, besides discontinuation of actual user condition on imports. Importantly, policymakers eliminated entire 26 import licensing lists along with introducing a negative list of products. All goods other than those included in the negative list, could be imported without any restriction. On average, tariffs fell from as high as 87% in 1990 to 43% in 1996 which further decreased to 32.4% in 1999 and even to 29% in 2002 (Virmani, Goldar, Veeramani, & Bhatt, 2004; Topalova & Khandelwal, 2011).

2 The exchange rate of Indian Rupee is a controlled float, wherein the central bank (Reserve Bank of India) allows price discovery in exchange rate markets; even though, prevents excessive volatility through required interventions However, the currency fluctuation indicates considerable volatility. For instance, on 22nd Feb 2016 Indian Rupees was as low as 68.8 from 58.5 against US $ in May 2014. Even, the depreciation between January 2018 and July 2018 experienced more than 8%.

3 Some previous studies, e.g. Bahmani-Oskooee and Satawatananon (2012) and Bahmani-Oskooee and Aftab (2017) have also used product-level data to analyze the impact of exchange rate volatility. However, they used annual trade data for the analysis. In our view, to capture the effect of volatility using annual data is not very appropriate as the effects are settled down in the lower frequency data. With this consideration, we utilize monthly frequency of commodities as well as exchange rate data that is conducive for capturing the volatility effects.

4 IIP is a reliable leading indicator of income and business cycles in India and the correlation between GDP and IIP is very strong, i.e. 0.97 at 1% significant level (Dash & Sharma, 2011).

5 Nevertheless, given the rigidity of domestic prices, as nominal exchange rate tends to move closely with real exchange rate over short- to medium-run horizon, the choice between real and nominal exchange rate is least likely for carrying significant effect on the estimated volatility.

6 Dumitrescu and Hurlin (2012) approach allows all coefficients to vary across cross-sections and performs the standard Granger causality equations for each individual cross-section. $\overline{W}$ statistics is based on the average of these test statistics, and $\overline{Z}$-statistics is the normal standardized version of this.

7 Unit root tests results are not reported because of space constraint however will be made available on a request.

8 Unit root tests results are not reported because of space constraint however will be made available on a request.

Additional information

Notes on contributors

Chandan Sharma

Chandan Sharma is Associate Professor of economics at Indian Institute of Management Lucknow, India. His research focuses on Trade, Industrial Economics, Infrastructure, Development Economics, Political Economy, and Shadow Economy. His work has appeared in several leading economics journals, such as, Economics Letter, Economics Modelling, Policy Modeling, International Review of Financial Analysis and Social Indicators Research. He received a PhD from University of Delhi. Debdatta Pal serves as Assistant Professor in Business Environment area with Indian Institute of Management Lucknow, India. He conducts empirical research in energy economics, transportation economics, and choice modelling. His works have been published in Energy Economics, Economic Modelling, Empirical Economics, Transportation Research Part A, Journal of Choice Modelling, and Journal of Policy Modeling.