Impacts of Real Depreciation and Appreciation on Aggregate Output in Taiwan

Abstract

Extending the IS-MP-AS model, this article finds that real depreciation helped to raise real gross domestic product (GDP) during 1999.Q1-2010.Q2 whereas real appreciation helped to increase real GDP during 2010.Q3-2016.Q4. In addition, a lower world real interest rate, a higher stock price, a higher real oil price or a lower expected inflation would increase real GDP. More deficit spending as a percent of GDP does not affect real GDP.

JEL Classification: F41, E62

Keywords

depreciation and appreciation government deficits world interest rates stock prices oil prices

Introduction

Taiwan had a population of 23.5 million and a gross domestic product (GDP) per capita of US$22,561 in 2016. Since 2009, its inflation rate has been below 2%. The unemployment rate of 3.79% is lower than most industrialized countries. Real GDP grew slowly at an annual rate of 1.41% in 2016. It has had trade surpluses for years but also face competition from other countries in the region such as China, Vietnam, India, and so on.

Taiwan is chosen as a case study mainly because international trade plays an important role in Taiwan’s economic development as evidenced by 113.44% of international trade as a percent of GDP. The real exchange rate plays a vital role in determining exports and imports. Due to lack of natural resources, Taiwan imports most energy needs. To create jobs and maintain adequate foreign reserves, exports became a driving engine in its economic development process. The central bank in Taiwan has to find a balance in selecting an exchange rate level which would be conducive to economic growth. Real depreciation is expected to increase exports but may raise import costs and cause a higher domestic inflation. However, real appreciation tends to hurt exports but reduce import costs and domestic inflation. As Taiwan gradually reduces restrictions on international capital flows, currency depreciation tends to reduce net capital flows because the value of foreign investors’ assets declines when it is measured in the foreign currency whereas currency appreciation tends to increase net capital flows because the value of foreign investors’ assets rises when it is measured in the foreign currency.

This article attempts to study the impact of real appreciation and depreciation on aggregate output in Taiwan based on an extended IS-MP-AS model (Romer, 2000). In examining the impact of exchange rate changes on aggregate output, some of previous studies (An, Kim, & Ren, 2014; Bahmani-Oskooee, Chomsisengphet, & Kandil, 2002; Bahmani-Oskooee & Miteza, 2006; Crosby & Otto, 2001; Edwards, 1986; G. Kim, An, & Kim, 2015; Y. Kim & Ying, 2007; Moreno, 1999; Morley, 1992; Narayan & Narayan, 2007; Ratha, 2010) include the money supply in the estimated regression as a proxy for monetary policy. Because many central banks target the policy rate instead of the money supply, the incorporation of a monetary policy function in the model may be more appropriate (Romer, 2000, p. 155).

Whether real depreciation or appreciation of the Taiwan dollar would be beneficial to aggregate output has been studied extensively. Previous findings including Taiwan in the sample are inconclusive. Based on a sample of 48 developing countries, including Taiwan, Nunnenkamp and Schweickert (1990) reject the contractionary devaluation hypothesis and reveal the following findings. Devaluation is expansionary for agricultural exporting countries. The expansionary effect of devaluations is delayed for heavily indebted countries. Contractionary devaluations are found for manufacturing exporting countries when devaluations occurred, but strong growth affects are found in later periods.

Based on a sample of 27 countries, Kamin and Klau (1998) find that there is lack of support for contractionary devaluations in the long run and that there is no evidence to suggest that contractionary devaluations have a stronger effect on output in developing than industrialized countries.

Using a sample of 11 countries, Crosby and Otto (2001) report that based on the whole sample period of 1980.Q1-2000.Q1, real appreciation does not affect real GDP in Taiwan. Y. Kim and Ying (2007) reveal that for the whole sample period, devaluations are expansionary in Taiwan, Singapore, South Korea, and Thailand but contractionary in Chile, Indonesia, Malaysia, Mexico, and the Philippines.

Based on a sample of 128 developing and developed countries during 19,602,008, Kappler, Reisen, Schularick, and Turkisch (2011) show that the negative effect of currency appreciation or revaluation on output is limited and statistically insignificant.

Bahmani-Oskooee and Mohammadian (2016, 2018) and Bahmani-Oskooee, Halicioglu, and Mohammadian (2018) indicate that the impact of currency depreciation shows asymmetric effects, suggesting that currency depreciation and appreciation may have different effects on aggregate output.

The Model

Romer (2000) suggests that the monetary policy function in the IS-MP-AS model should replace the money market equilbrium (LM) function in the traditional IS-LM-AS model because many central banks including the central bank in Taiwan targets the policy rate instead of the money supply as the IS-LM-AS model suggests. Romer (2000) does not include the stock price, the world real interest rate, and the real oil price in his original model. This paper incorporates several new variables such as the world real interest rate, the stock price, and the real oil price to consider the roles of effect of U.S. monetary policy on Taiwan’s monetary policy, the financial market, and supply shocks in the model.

Extending Romer (2000), we can express the goods market equilibrium (IS), the monetary policy (MP) and the aggregate supply (AS) functions as the following:

Y = W (Y, G, T, L, S, ε)

(1)

R = X (π - π^{*}, Y - Y^{p}, ε, R^{w})

(2)

π = Z (π^{e}, Y - Y^{p}, ε, E)

(3)

L = F (R)

(4)

where

Y = real GDP in Taiwan,

G = government spending,

T = government tax revenue,

L = the real lending rate,

S = the stock price,

$ε$ = the real effective exchange rate (An increase means real appreciation.),

R = the real policy rate,

$π$ = the inflation rate,

$π^{*}$ = the inflation target,

Y^p = potential real GDP,

R^w = the world real interest rate,

$π^{e}$ = the expected inflation rate, and

E = the real oil price.

Note that Equation (2) is an extended Taylor rule (Taylor, 1993, 1999) and that Equation (3) is an extended expectations-augmented aggregate supply function. Assume that the inflation target and potential real GDP are constants in the short run. Solving for the three endogenous variables, we can express equilibrium real GDP as:

Y^{*} = H (ε, G - T, R^{w}, S, E, π^{e})

(5)

The Jacobian for the endogenous variables can be written as:

| J | = [(1 - W_{Y}) - W_{L} F_{R} X_{π} Z_{Y} - W_{L} F_{R} X_{Y}] > 0

(6)

The partial derivative of equilibrium real GDP with respect to the real effective exchange rate is given by:

\partial Y^{*} / \partial ε = (W_{ε} + W_{L} F_{R} X_{π} Z_{ε} + W_{L} F_{R} X_{ε}) / | J | > o r < 0

(7)

Note that the first term in the numerator tends to be negative and that the second and third terms in the numerator tend to be positive. Hence, the impact of real appreciation on equilibrium real GDP is unclear.

The sign of G—T or E is unclear. Barro (1974, 1989) proposes the Ricardian equivalence hypothesis that deficit- or debt-financed government spending has a neutral effect on aggregate output in the long run. Cebula (1997), Cebula and Cuellar (2010), Cebula (2014a, 2014b), and Cebula, Angjellari-Dajci, and Foley (2014) find that more government deficit spending raises the real interest rate and would crowd out consumption and investment spending. Aisen and Hauner (2013) show that the budget deficit would affect the interest rate significantly only if a country exhibits high-budget deficit and high domestic debt. A higher real oil price is expected to reduce short-run aggregate supply and real GDP. If a higher real oil price is demand driven, the impact on real GDP may be positive in the short run (Kilian, 2008a, 2008b).

An analysis of the data in Figure 1 shows that the relationship between real GDP and the real effective exchange rate seem to have changed after 2010.Q3. Hence, an interactive binary variable and an intercept binary variable are included in the estimated equation. A possible cause for the structural change may be due to a change in the exchange rate policy to place more emphasis on lowering domestic inflation and drawing more international capital inflows:

Y^{*} = A (ε, ε \times B, B, G - T, R^{w}, S, E, π^{e})

(8)

where B = 0 during 1999.Q1-2010.Q2 and B = 1 during 2010.Q3-2016.Q4. The partial derivative of $Y^{*}$ with respect to $ε$ can be expressed as $\partial Y^{*} / \partial ε = θ_{1} + θ_{2} B$ where $θ_{1}$ and $θ_{2}$ they are the estimated coefficients for $ε$ and $ε \times B$ , respectively. Thus, $\partial Y^{*} / \partial ε = θ_{1}$ during 1999.Q1-2010.Q2 when the binary variable has a value of zero, and $\partial Y^{*} / \partial ε = θ_{1} + θ_{2}$ during 2010.Q3-2016.Q4 when the binary variable has a value of 1. The scatter diagram in Figure 2 seems to suggest that real GDP and the government deficit as a percent of GDP may not have a clear positive or negative relationship.

Figure 1.

Scatter diagram between REALGDP and the REER.

Figure 2.

Scatter diagram between REALGDP and the government DEFICITGDP.

Empirical Results

The data were collected from the Central Bank of the Republic of China, Taiwan’s Bureau of National Statistics, and the Federal Reserve Bank of St. Louis. Real GDP is measured in million Taiwan dollars. The real effective exchange rate is a trade-weighted index. An increase means real appreciation of the Taiwan dollar. The government deficit is measured as a percent of GDP. The world real interest rate is represented by the U.S. federal funds rate minus the U.S. inflation rate. The lagged value of the U.S. real federal funds rate is used as the central bank in Taiwan may respond to a change in the real federal funds rate with a time lag. The stock price is represented by the share price index with 2010 as the base year. The real oil price is measured in the Taiwan dollar and adjusted by the consumer price index. The expected inflation rate is estimated as a weighted average inflation rate of the past four quarters. The weights of 0.4, 0.3, 0.2, and 0.1 are assigned for the first, second, third, and fourth lagged quarters, respectively. The sample ranges from 1999.Q1 to 2016.Q4. The data for the government deficit are not available before 1999.Q1.

The ADF test shows that each of the variables has a unit root in level and is stationary in first difference at the 1% level. The Augmented Dickey-Fuller (ADF) test on regression residuals shows that the test statistic of −3.5451 is greater than the critical value of −3.5256 in absolute values at the 1% level. Therefore, these time series variables have a long-term stable relationship.

Table 1 reports the estimated regression and relevant statistics. The GARCH process addresses potential autoregressive conditional heteroscedasticity that may exist in time series data. The significant coefficient in the variation equation shows that autoregressive conditional heteroscedasticity exists. The ADF test uses the residuals from the GARCH estimate. Approximately 96.56% of the change in real GDP can be explained by the exogenous variables with significant coefficients. Except for the coefficient of the government deficit as a percent of GDP, other coefficients are significant at the 1% or 2.5% level. Real GDP is positively associated with real appreciation of the Taiwan dollar during 2010.Q3-2016.Q4, the stock price and the real oil price and negatively influenced by real appreciation during 1999.Q1-2010.Q2, the lagged real federal funds rate and the expected inflation rate. The mean absolute percent error of 3.26% suggests that the forecast error is relatively small. It seems that during 1999.Q1-2010.Q2, potential positive impacts of real depreciation such as increased net exports overwhelmed potential negative impacts of real depreciation such as higher import prices and inflation and capital outflows and that during 2010.Q3-2016.Q4, potential positive impacts of real appreciation such as lower domestic price or inflation and capital inflows outweighed potential negative impacts of real appreciation such as decreased net exports.

Table 1.

Estimated Regression of Log (Real GDP) in Taiwan.

Variable	Coefficient	z statistic	p value
Constant	18.21954	239.0253	.0000
Log(real effective exchange rate)	−0.975401	−48.11048	.0000
Log(real effective exchange rate)*Binary variable	3.283382	10.40207	.0000
Binary variable	−15.02000	−10.28270	.0000
Deficit as a percent of GDP	0.000801	0.803761	.4215
Lagged U.S. real federal funds rate	−0.005051	−2.214020	.0268
Log(stock price)	0.133418	6.883315	.0000
Log(real oil price)	0.087289	8.773614	.0000
Expected inflation rate	−0.007193	−3.887218	.0001
R ²	.955,574
Adjusted R²	.949,933
Akaike information criterion	−3.645084
Schwarz criterion	−3.297260
Sample period	1999.Q1-2016.Q4
Number of observations	72
Methodology	GARCH
MAPE	3.262556%

Note. Binary variable = 0 during 1999.Q1-2010.Q2, and binary variable = 1 during 2010.Q3-2016.Q4; GDP = Gross Domestic Product; GARCH = generalized autoregressive conditional heteroskedasticity; MAPE = mean absolute percent error.

Specifically, during 1999.Q1-2010.Q2, a 1% real appreciation of the Taiwan dollar would result in a 0.9754% decrease in real GDP whereas during 2010.Q3-2016.Q4, a 1% real appreciation of the Taiwan dollar would lead to a 2.3080% increase in real GDP. A 1% increase in the stock price is expected to cause real GDP to rise by 0.1334%. The significant positive coefficient of the real oil price suggests that the demand-driven hypothesis applies to Taiwan.

The insignificant coefficient of the government deficit-to-GDP ratio suggests that deficit-financed government spending shifts aggregate demand to the right, but the crowding-out effect tends to reduce private spending, shift aggregate demand to the left, and cancel out the increase in aggregate demand. The negative significant coefficient of the lagged U.S. real federal funds rate indicates that the central bank in Taiwan is expected to raise its real policy rate with a time lag if the real federal funds rate increases. The higher real policy rate causes the real lending rate to rise and hurts private spending. The positive significant coefficient of the stock price implies that a higher stock price increases household wealth, which, in turn, raises consumption spending and aggregate demand.

Following a referee’s suggestion, the lagged federal funds rate is replaced with the current real federal funds rate. The coefficient of the current real federal funds rate is found to be positive and significant, which is opposite to the theoretical sign mainly due to a high degree of multicollinearity between the real federal funds rate and some of the other right-hand side variables. Therefore, the lagged federal funds rate tends to indicate that there may be a time lag in Taiwan’s monetary policy in response to the U.S. monetary policy.

Several other versions are considered. If the interactive and intercept binary variables for the real effective exchange rate are not included in the estimated regression, the coefficient of the real effective exchange rate is estimated to be −1.0869 and significant at the 1% level. This result may be misleading as the positive relationship shown in Figure 1 is overlooked. If the lagged U.S. real lending rate is used to represent the world real interest rate, its estimated coefficient of −0.0048 is significant at the 1% level. Other results are similar. If the expected inflation rate is represented by the simple lagged inflation rate, the estimated coefficient of −0.0056 is significant at the 1% level but has a larger standard error than that reported in Table 1.

Summary and Conclusions

This article has examined the effects of real appreciation or depreciation and other relevant variables on aggregate output in Taiwan. An extended IS-MP-AS model is applied. Real depreciation raised real GDP during 1999.Q1-2010.Q2 whereas real appreciation increased real GDP during 2010.Q3-2016.Q4. More government deficit as a percent of GDP does not affect real GDP. In addition, real GDP has a positive relationship with the stock price and the real oil price and a negative relationship with the U.S. real federal funds rate and the expected inflation rate.

There are several policy implications. The finding that real depreciation raised real GDP during 1999.Q1-2010.Q2 and real appreciation raised real GDP during 2010.Q3-2016.Q4 suggests that the impact of real appreciation or real depreciation on aggregate output could be either positive or negative, depending upon the time period under consideration. The central government may need to be more prudent in engaging in expansionary fiscal policy as a higher government deficit-to-GDP ratio does not raise real GDP. Because Taiwan’s monetary policy is influenced by U.S. monetary policy, the expected increase in the U.S. federal funds rate in the foreseeable future would cause Taiwan’s policy rate and other short-term rates to rise.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Author Biography

Yu Hsing received his PhD in Economics from the University of Tennessee – Knoxville. He has written or coauthored many articles published in Industrial Relations, Economics Bulletin, Applied Economics, Journal of Urban Economics, Journal of Macroeconomics, Information Economics and Policy, Global Economy Journal, Economics of Education Review, Education Economic Contemporary Economic Policy, Energy Economics, American Journal of Economics and Sociology, Cogent Economics & Finance, etc.

References

Aisen

Hauner

(2013). Budget deficits and interest rates: A fresh perspective. Applied Economics, 45, 2501-2510.

Kim

Ren

(2014). Is devaluation expansionary or contractionary: Evidence based on vector autoregression with sign restrictions. Journal of Asian Economics, 34, 27-41.

Bahmani-Oskooee

Chomsisengphet

Kandil

(2002). Are devaluations contractionary in Asia? Journal of Post Keynesian Economics, 25, 69-81.

Bahmani-Oskooee

Halicioglu

Mohammadian

(2018). On the asymmetric effects of exchange rate changes on domestic production in Turkey. Economic Change and Restructuring, 51, 97-112.

Bahmani-Oskooee

Miteza

(2006). Are devaluations contractionary? Evidence from panel cointegration. Economic Issues, 11(Pt. 1), 49-64.

Bahmani-Oskooee

Mohammadian

(2016). Asymmetry effects of exchange rate changes on domestic production: Evidence from nonlinear ARDL approach. Australian Economic Papers, 55, 181-191.

Bahmani-Oskooee

Mohammadian

(2018). Asymmetry effects of exchange rate changes on domestic production in emerging countries. Emerging Markets Finance & Trade, 54, 1442-1459.

Barro

R. J.

(1974). Are government bonds net wealth? Journal of Political Economy, 82, 1095-1117.

Barro

R. J.

(1989). The ricardian approach to budget deficits. Journal of Economic Perspectives, 3, 37-54.

10.

Cebula

R. J.

(1997). An empirical note on the impact of the federal budget deficit on ex ante real long term interest rates, 1973-1995. Southern Economic Journal, 63, 1094-1099.

11.

Cebula

R. J.

(2014a). An empirical investigation into the impact of US federal government budget deficits on the real interest rate yield on intermediate-term treasury issues, 1972-2012. Applied Economics, 46, 3483-3493.

12.

Cebula

R. J.

(2014b). Impact of federal government budget deficits on the longer-term real interest rate in the US: Evidence using annual and quarterly data, 1960-2013. Applied Economics Quarterly, 60, 23-40.

13.

Cebula

R. J.

Angjellari-Dajci

Foley

(2014). An exploratory empirical inquiry into the impact of federal budget deficits on the ex post real interest rate yield on ten year treasury notes over the last half century. Journal of Economics and Finance, 38, 712-720.

14.

Cebula

R. J.

Cuellar

(2010). Recent evidence on the impact of government budget deficits on the ex-ante real interest rate yield on Moody’s Baa-rated corporate bonds. Journal of Economics and Finance, 34, 301-307.

15.

Crosby

Otto

(2001). Growth and the real exchange rate-evidence from eleven countries. Hong Kong: Hong Kong Institute for Monetary Research.

16.

Edwards

(1986). Are devaluations contractionary? Review of Economics and Statistics, 68, 501-508.

17.

Kamin

S. B.

Klau

(1998, September). Some multi-country evidence on the effects of real exchange rates on output (Working Papers No. 48, Bank for International Settlements). Basel, Switzerland.

18.

Kappler

Reisen

Schularick

Turkisch

(2011). The macroeconomic effects of large exchange rate appreciations (ZEW Discussion Papers, No. 11-016). Mannheim, Germany.

19.

Kilian

(2008a). The economic effects of energy price shocks. Journal of Economic Literature, 46, 871-909.

20.

Kilian

(2008b). Not all oil price shocks are alike: Disentangling demand and supply shocks in the crude oil market (CEPR Discussion Paper No. 5994). London, UK.

21.

Kim

(2015). Exchange rate, capital flow and output: Developed versus developing economies. Atlantic Economic Journal, 43, 195-207.

22.

Kim

Ying

Y. H.

(2007). An empirical assessment of currency devaluation in East Asian countries. Journal of International Money and Finance, 26, 265-283.

23.

Moreno

(1999). Depreciation and recessions in East Asia. Economic Review-Federal Reserve Bank of San Francisco, 3, 27-40.

24.

Morley

S. A.

(1992). On the effect of devaluation during stabilization programs in LDCs. Review of Economics and Statistics, 74, 21-27.

25.

Narayan

P. K.

Narayan

(2007). Is devaluation expansionary or contractionary? Empirical evidence from Fiji. Applied Economics, 39, 2589-2598.

26.

Nunnenkamp

Schweickert

(1990, January). Real exchange rates and economic growth in developing countries: Is devaluation contractionary? (Kiel Working Paper No. 405). Kiel, Germany: Kiel Institute for the World Economy.

27.

Ratha

(2010). Does devaluation work for India? Economics Bulletin, 30, 247-264.

28.

Romer

(2000). Keynesian macroeconomics without the LM curve. Journal of Economic Perspectives, 14, 149-169.

29.

Taylor

J. B.

(1993). Discretion versus policy rules in practice. Carnegie Rochester Conference Series on Public Policy, 39, 195-214.

30.

Taylor

J. B.

(1999). A historical analysis of monetary policy rules. In Taylor

J. B.

(Ed.), Monetary policy rules (pp. 319-348). Chicago, IL: University of Chicago Press.