Exchange Rate Volatility on the Timing of Foreign Direct Investment: Market-seeking versus Export-substituting
Chia-Ching Lin* Kun-Ming Chen Hsiu-Hua Rau March 18, 2006 *Department of International Trade, National Chengchi University, Taiwan
Stylized Facts
The flows of FDI have been increasing dramatically and fluctuating sharply since the 1970s. Since the breakdown of the Bretton Woods system in 1973, the exchange rates of many countries have been fluctuating considerably over time.
Motivation
How to explain the short-run movements of FDI. It has been suggested that the exchange rate changes may be a good explanation for it. However, there is no consensus either in theory or empirical studies. Few studies focus on the newly industrializing countries or the developing countries
Literature Review
Relationship between Exchange Rate Uncertainty and FDI
Cushman (1985) and Goldberg and Kolstad (1995)
– They emphasize the importance of considering the postFDI changes in the exposure of a firm’s profits to exchange rate risk.
– If the investing firm can choose to serve foreign markets via exports or FDI, then an increase in exchange rate volatility might lead the firm to substitute FDI for exports. – It is because FDI activity reduces the exposure of its profits to exchange rate risk.
Literature Review
Relationship between Exchange Rate Uncertainty and FDI
Dixit (1989a,b)
– A real options model. – He indicates that the waiting value increases as the uncertainty rises even for a risk-neutral firm. – Hence, an increase in exchange rate uncertainty will defer the FDI activity of the firm.
Objective
Using Real Options Approach with risk aversion, this paper investigate theoretically the effects of exchange rate uncertainty on the FDI activity under different motives of the firms.
The relationship between exchange rate uncertainty and FDI varies with the extent of the exposure to exchange rate risk which is determined by investing motives.
Objective (cont’)
Firm-level data on Taiwan’s outbound FDI in China over the period 1987-2002 are employed to test the validity of the theoretical results. Survival analysis is used to examine the relationship between the timing of FDI and covariates.
A Real Options Model
Assumptions – The exchange rate R, expressed in units of home currency per foreign currency, follows an exogenously geometric Brownian motion.
dR dt dz R
– Risk aversion firms. – A potential entrant stays in the market forever after entering the market. – Perfect competition.
A Real Options Model (cont’)
Mean-variance expected utility function
– ap is Arrow-Pratt’s absolute risk aversion coefficient
1 EU E a p Var 2
Two Types of Firms
Export-substituting Firm
– An exporting firm, originally producing at its home country and serving a foreign market via exports, relocates its whole production abroad to serve the foreign market. – Profit flows pre-FDI:
0 Pf R Wd
– Profit flows post-FDI
1 Pf R Wf R
Two Types of Firms (cont’)
Market-seeking Firm
– A domestic firm, originally not serving a foreign market via exports, chooses to set up a foreign subsidiary to produce and sell in a given foreign market.
– Profit flows pre-FDI:
0
0
– Profit flows post-FDI
1 Pf R Wf R
Proposition 1
In the case of export-substituting FDI, an increase in exchange rate volatility will stimulate FDI activity of a firm if the firm cannot delay its investment.
– The risk attitude is the only channel through which exchange rate uncertainty affects FDI. – Substituting FDI for exports reduces the firm’s exposure to exchange rate risk, and this gain from risk reduction is larger if the exchange rate is more volatile.
Proposition 2
A risk-neutral export-substituting firm will delay its FDI activity when the exchange rate volatility rises.
– The option value of investment flexibility is the only channel through which exchange rate uncertainty affects FDI.
– An investment is like a call option whose value rises if the underlying uncertainty increases.
Proposition 3
In the case of export-substituting FDI, the effect of exchange rate volatility on the timing of FDI is ambiguous. However, there exists a threshold in the degree of risk aversion a such that this effect is positive (negative) if the firm’s risk-aversion coefficient is greater (smaller) than a.
– Given the negative effect on FDI activity from the option value of investment flexibility, if the positive effect resulting from the risk aversion as well as the change in the exposure to exchange risk resulting from the firm’s FDI becomes large enough, the net effect will be positive .
Proposition 6
In the case of market-seeking FDI, an increase in exchange rate volatility will delay the FDI activity of the firm.
– FDI activity will make the firm’s exposure to exchange rate risk increase.
– An increase in exchange rate volatility will increase the option value of delaying the investment so as to deter the FDI activity further.
Expected signs of the determinants of FDI with different motives
Types Exchange Exchange Exchange Variables Rate Level Rate Trend Rate Volatility (R) (μ) (σ) risk neutrality risk aversion risk neutrality risk aversion － － ＋ ＋ － ？ ＋ ？ － ？ － － Sunk Costs (k) － － － － Foreign Domestic Wage Rate Wage Rate (Wf) (Wd) － － － ? ＋ ＋
Exportsubstituting FDI
Marketseeking FDI
Empirical Model: Survival Analysis
This paper focuses on the analysis of how exchange rate volatility affects the timing of foreign entry. One widely applied method to examine the issue about timing is to conduct survival analysis (event history analysis).
– The event is a firm’s entry into a foreign market. – The waiting time for a firm to enter a foreign market can be treated as the survival time of the firm.
– The timing of entry can be treated as the timing of event occurrence.
Survival Analysis (cont’)
Cox’s proportional hazard model
Hazard function
h t | xi1,x2i (t ) h0 t exp x1i x2i (t )
Partial likelihood function
e Lp x1j x 2j ( t ) e i 1 j ( ti )
K
x1i x 2i ( t )
i
Empirical Equation
log h(t ) h0 (t ) 1 t 2 EX M * t 3 EX E * t 4 Rt 1 + 5 EX M * Rt 1 6 EX E * Rt 1 7 t 8 EX M * t 9 EX E * t 10WAGEt 1 11 PFi ,t i 1MKT 2 FUNDi 3 R & Di 4 SIZEi 5 SIZEi * SIZEi + 6 KLi 7 HTi
Dependent Variable
The dependent variable is defined as the duration from 1987 to the year when the firm invested in China.
– It is because Taiwanese firms were not permitted to invest in China until 1987.
Sample period: 1987-2002.
Sample size: 337 listed companies (4063 observations).
Independent Variables
– Rt-1 : the one-period lagged real exchange rate. – μt : the trend of the real exchange rate. – σt : the volatility of the real exchange rate. – Waget-1: the one-period lagged relative real wage rates of China with respect to Taiwan. – MKTi: marketing intensity, a proxy variable of the sunk costs.
Independent Variables (cont’)
Control variables:
Profits rate (PF) Source of funds (FUND): a dummy variable; parent company (1); otherwise (0). R&D intensity (R&D) Firm’s size (SIZE): firm’s sales Capital-labor ratio (KL) High-tech industry dummy (HT): a dummy variable ; high technology industries (1); otherwise (0).
Two measures for μt and σt
Unconditional measure: Tasy (2002)Modified average and modified standard deviation of the monthly change in the logarithm of exchange rate
Conditional measure: GARCH
Independent Variables (cont’)
EXE : Export-substituting firm
dummy
– EXE =1, for a firm with export ratio greater
than 0.6 and the sales of its subsidiary account for more than 80% of the subsidiary’s total sale in China .
– EXE =0, otherwise.
Independent Variables (cont’)
EXM : Market-seeking firm
dummy
–
sales of the firm’s subsidiary account for more than 80% of its total sale in China.
EXM =1, for a firm with zero exports and the
– EXM =0, otherwise.
Distribution of sample firms by industry
MarketIndustry seeking firms Food & Beverage Chemicals and Plastic Products Electronic & Electric Appliances Other Manufacturing Services Others Total 7 1 1 5 8 1 23 substituting firms 0 2 15 5 0 0 22 Export-
Cox estimation of the Determinants of FDI
Equations Covariates σ t( 1 ) EXM*σt ( 2 ) EXE*σt ( 3 ) Rt-1 ( 4 ) EXM *Rt-1 ( 5 ) EXE *Rt-1 ( 6 ) μt ( 7 ) EXM *μt ( 8 ) EXE *μt ( 9 ) WAGEt-1 ( 10 ) MKTi ( 1 )
1 2
1 3
4 5 4 6
Tsay (2002)
(1) Coefficients (2) Coefficients
GARCH (1,1)
(3) Coefficients (4) Coefficients
4.3830 (1.54)
5.2413c (1.88) -10.1228a (-6.06) 1.7716 (0.66)
23.07 (1.57)
27.29c (1.89) -49.59a (-5.50) 5.8598 (0.44)
-0.3988a (-2.92)
-0.2342 (-1.55) 0.7020a (7.54) -0.0925 (-1.27)
-0.3821a (-3.24) -0.2012 (-1.44) 0.6676a (6.00) -0.1071c (-1.88) -0.8448 (-0.03) -9.9318 (-0.45) 92.91a (4.87) 53.92a (8.10) -0.4276a (-6.69) -0.4349a (-6.91) -1.0352a (-4.39) -3.3328a (-2.86) -22.30a (-2.88) 33.15b (2.04) 0.4564a (2.83) -0.3083b (-2.32) 82.98a (7.71) 43.99 (1.55)
204.90
a
Notes: t-statistics are in parentheses; a , band c denote tstatistics are significant at the 1% ,5% and 10 % confidence levels, respectively.
0.0314 (0.03)
-0.3465 (-0.36) 3.9059 (4.72) 2.0351a (13.8)
a
-0.4220a (-6.69) -0.4306a (-6.99) -1.0365a (-4.37) -3.2747a (-2.96) -4.8815a (3.16) 7.0129b (2.20) 0.4678a (2.81) -0.3266b (-2.23) 3.5594a (9.21) 1.6886b (2.09)
205.66
a
7 8 7 9
Likelihood ratio test
232.37
a
232.79a
Cox estimation: Control Variables
Equations Covariates PFi,t ( 11 ) FUNDi ( 2 ) R&Di ( 3 ) SIZEi ( 4 ) SIZEi*SIZEi ( 5 ) KLi ( 6 ) HTi ( 7 )
Tsay (2002) (1) Coefficients 0.0071 (5.58) 0.7722a (2.64) 0.1004c (1.88) 0.1775a (6.33)
a
GARCH (1,1) (2) (3) Coefficients 0.0071 (5.36) 0.7723a (2.64) 0.1005c (1.89) 0.1775a (6.35)
a
(4) Coefficients 0.0107a (6.44) 0.7545a (4.06) 0.0991 (1.59) 0.0935c (1.97)
Coefficients 0.0108 (6.41) 0.7554a (3.98) 0.0954 (1.59) 0.0914c (1.93)
a
-0.0056a (-8.78) -0.0032b (-2.40) -0.1343a (-8.60) -0.0926a (-5.15) -0.6209 (-0.89) -0.6624 (-0.91)
-0.0056a (-8.81) -0.0032b (-2.45) -0.1343a (-8.59) -0.0924a (-5.11) -0.6208 (-0.89) -0.6644 (-0.91)
Notes: t-statistics are in parentheses; a , band c denote tstatistics are significant at the 1% ,5% and 10 % confidence levels, respectively.
Conclusion
Higher exchange rate volatility tends to delay Taiwan’s FDI into China of marketseeking firms, but accelerate that of export-seeking firms.
A depreciation of RMB to NTD tends to stimulate the Taiwan’s FDI into China of export-substituting firms, but deter that of market-seeking firms.
In general, the empirical results are consistent with the prediction of the theory.
THE END
Comments and Suggestions are Welcome!