Do exchange rates have any impact upon UK inward foreign direct investment?

Abstract

This article examines the impact of the level and volatility of the real exchange rate on UK foreign direct investment (FDI) inflows from the seven major countries of origin of the investment over the period 1975–2001. We use both fixed effects and dynamic generalized methods of moments (GMM) panel estimation techniques, and manufacturing data disaggregated by high and low R&D content of the sector of destination. Our results provide strong evidence that exchange rate volatility has a negative impact on FDI flows into the UK, irrespective of the sector of destination of the investment. On the other hand, the level of the real exchange rate is found to have a statistically insignificant effect on FDI after controlling for endogeneity of the regressors.

Notes

¹ To the authors’ knowledge, the only published study examining the determinants of inward FDI into the UK is that by Pain (1993). However, in his analysis, the only exchange rate variable considered is the rate of change of the exchange rate.

² Where the exchange rate is defined as the price of pounds sterling in terms of units of foreign currency.

³ This allowed us to establish the presence of any country-specific effects. Variations between countries are modelled using dummies. We also estimated (1) using a random-effects technique, which treats country specific effects as random variables. We found the results of the fixed-effects procedure to be superior in that it produced an estimated model with higher R ² and no serial correlation.

⁴ In searching for the optimal volatility measure, during the pre-testing phase we experimented with several ARCH, GARCH and SD specifications. Using a range of model selection criteria, we found that the GARCH (1,1) process provided the best overall fit to the data. For a detail illustration of an analogous selection tournament among competing proxies of exchange rate volatility, see the authors’ previous work on volatility and trade (De Vita and Abbott, 2004).

⁵ From 1975 to 2001 sterling appreciated on average by 46.3% in real terms, with a 16.8% appreciation between 1990 and 2001 and a 3.8% appreciation since 1995. During the same periods inward FDI increased significantly: 829% from 1975 to 2001, 288% from 1990 to 2001, and 119% from 1995 to 2001. The largest rises during the boom period of the 1990s took place in low R&D sectors (from 1995 to 2001 high R&D FDI increased by 43%, whereas low R&D FDI increased by 177%) for which a larger real exchange rate elasticity is found.

⁶ The decision to split the sample at 1987 was motivated by the visual inspection of Fig.1, which suggested a possible structural break in the FDI series at that time.

⁷ Using dummies for each time period multiplied by the real exchange rate, we also tested for the joint significance of the dummy coefficients. The insignificance of the Wald test for exclusion restrictions confirmed that the exchange rate effect was relatively constant over the sample period.

⁸ To implement the Arellano-Bond estimator we took the first difference of all of the variables in (1), dropping distance and common language since they are time-invariant. Since Δε _it and ΔY _it−1 are correlated, we used lagged levels as instruments for the dependent variable and the regressors and then estimated by GMM. The validity of the instruments was assessed using the Sargan's test for over-identifying restrictions (Sargan, 1958). The unbiasedness of the estimates was addressed by testing for serial correlation up to 2nd order.