The Impact of Yen Fluctuation on the Trade between. Taiwan and Japan

Transcription

1 The Impac of Yen Flucuaion on he Trade beween Taiwan and Japan Bao-bao Wang Chih-Ching Yang Pei-Chun Kao Ting-Yu Hsu Yi-Ting Liu Deparmen of Inernaional Business, Naional Chengchi Universiy Advisor Prof. Kun-Ming Chen Deparmen of Inernaional Business, Naional Chengchi Universiy Absrac The paper deals wih he possible facors which may bring impacs on he rade beween Taiwan and Japan, and how he yen flucuaion will affec Taiwanese expor indusry. We firs derive he funcion by employing he model proposed by Dr. Chen in his former paper work. We hen conclude ha here are hree major causes, which are he real exchange of yen, he producion cos, and he hird counry effec ha may influence he bilaeral rade beween Taiwan and Japan. To go a sep furher, we discuss he possible impacs of hese facors posed on respecive indusries. To simplify he research, we pick up he elecrical machinery and he machinery indusry o represen he Taiwanese expors, and we find ha he yen flucuaion can generae posiive or negaive effecs on he indusry depends on is srucure. 1

2 1. INTRODUCTION 1.1 Moivaion and Background Las year, Abe Shinzo again became he presiden of LDP (Liberal Democraic Pary) on 26 h Sepember. He proposed and implemened a series of economic policies so called Abenomics. He waned o conquer he problems of deflaion and he high Japanese yen exchange rae. Hence he financial measures below Abenomics led o he depreciaion of Japanese yen. This is seen in Figure 1, which presens Japanese yen flucuaion from May I had been 29% depreciaion of Japanese yen from 26 h Sepember 2012 o 31 s May Figure 1: The Flucuaion of Japanese yen o Dollar since May 2012 Daa from Cenral Bank of ROC: hp:// Figure 2 shows more evidences of depreciaion of Japanese yen. Taiwan and China are main rade counries o Japan. The exchange rae o hese wo also shows he rend of depreciaion. Alhough Japanese yen o hese currencies all shows depreciaion, he degree of each one is differen. 2

3 Figure 2: The Flucuaion of Japanese yen o TWD and CNY since July 2012 JPY/TWD JPY/CNY Daa from X-RATES: /7~2013/7 From figure 3 and 4, we can find ha Japan is Taiwan s hird larges exporing counry and is Taiwan s bigges imporing counry. I means ha Japan is Taiwan s imporan rade parner. Figure 3: Taiwan s Top 5 Exporing Counries 16% 7% 1% China 31% 45% Unied Saes Japan Daa from ROC Bureau of Foreign Trade: hp://cus93.rade.gov.w/fsci/ Figure 4: Taiwan s Top 5 Imporing Counries 11% 21% 8% 21% 39% Japan Unied Saes China Korea Saudi Arabia Daa from ROC Bureau of Foreign Trade: hp://cus93.rade.gov.w/fsci/ 3

4 1.2 Purpose of Sudy and Research Procedures Because Taiwan and Japan have close relaion on rade, we wan o figure ou wheher he flucuaion of Japanese yen affecs he rade beween Taiwan and Japan. Our research procedures are divided ino hree pars. Firs, we design a model and ry o find ou if he flucuaion of Japanese yen exchange rae has significan influence on Taiwan s oal expors o Japan. Second, we analyze he aggregae regression resul. Third, according o he World Trade Alas (WTA), he mos op wo goods Japan impors from Taiwan are elecrical machinery and machinery. Therefore, we respecively analyze he regression resul of hese wo indusries. 2. The Third Counry Effec There is an imporan effec called he hird counry effec which can be overlooked when esimaing he impac of he exchange rae flucuaion of yen on Taiwan s expor volume o Japan. The hird counry effec is composed of 2 facors: One is he cos of impored inermediae goods; he oher is he compeior s price. The following is he explanaion of he meaning of hese wo facors. The firs one is he cos of impored inermediae goods. Firs, we sugges ha, in order o expor goods o Japan, Taiwanese companies mainly impor maerials from counry J. Since we impor goods from counry J, he price of hese goods will become pars of our oal cos. Consequenly, if he goods impored from counry J have lower price, hen we will have a lower cos, and his brings o he case ha Taiwan s expor producs have lower exporing price in Japanese marke. From he inference above, we figured ou ha he cos of impored inermediae goods will affec he Taiwan s oal expor volume o Japan. Hence, he cos of impored inermediae goods is one of he facors of he hird counry effec. 4

5 The second facor is he compeior s price. Firs, we sugges ha Taiwan s main compeior in Japan is counry K. If he produc price of K is lower, i will affec Taiwan s expor producs compeiive power. The influence on Taiwan s expor producs compeiive power will affec Taiwan s oal expor volume o Japan in he end. Forasmuch, he compeior s price is he second facor of he hird counry effec. The hird counry effec exiss since boh he cos of impored inermediae goods and he compeior s price will affec Taiwan s oal expor volume o Japan. This effec will also be considered in our model when discussing wheher he yen s flucuaion will have significan impac on Taiwan s expor o Japan. 3. Model Designing and Explanaion 3.1 Expor Supply Since he opic deals wih he possible impac ha he yen flucuaion may bring o Taiwanese expors o Japan, observing he expor volume is a necessary process. To go a sep furher, he expor volume is deermined by he supply-demand curve. As a resul, he model is designed based on such a concep, and he supply side will be inroduced in his secion firs. As he supply side is originaed from he producion cos, we derive he funcion a he beginning. The Taiwanese suppliers oal producion cos, which is priced a New Taiwanese Dollar (NTD) is lised here: α 1 α TC = [( PD ) ( PM ) ] Q NTD NTD NTD (1) PD!"#!"#! and PM! represen he price he uni cos for which he Taiwanese suppliers spend in he domesic and foreign environmen o expor goods o Japan, respecively. The price for which he Taiwanese suppliers spend in he foreign counries, for insance, is he cos of imporing inermediae goods for reproducion. 5

6 These coss ime he expor quaniy Q! hen become he Taiwanese suppliers oal cos for exporing goods o Japan, which is TC!"!!. In addiion, o simplify he analysis, we assume ha he goods from foreign counries are differen, while he goods made by he domesic producers are all he same. Suppose ha here is a large amoun of suppliers in Taiwanese marke and herefore facing a perfec compeiion. Then according o he principle of marginal cos pricing, he price which he Taiwanese suppliers pursue o maximize heir expor profi will be PX!"#, which is he derivaive from formula (1). In oher words, he price equals he marginal cos: PX NTD = ( PD NTD ) ( PM ) α NTD 1 α (2) Nex, we compare he price of Taiwanese expors, which is he marginal cos, wih he one of Japanese producs. The raio is he real exchange rae of Japanese yen. NTD 1 PX px = ( ) (3) JP NTD JP P E!"# Formula (1) indicaes ha he nominal exchange rae E!"! now is one Japanese yen o E! unis of New Taiwanese Dollar. As here are PX! unis of NTD in oal, PX!!"# divided by E!!"#!" is he nominal value of Japanese yen. Then again do he division of dividing his value by Japanese Wholesale Price Index (WPI), he final saisics is he real exchange rae of Japanese yen, compared wih NTD. 1 NTD NTD NTD j j NTD PD α PM 1 α 1 PD α 1 PM E [( ) ( ) ] = [ ( )] [ ( )] 1 JP NTD JP NTD JP JP NTD JP JP NTD JP P E E P E P E 1 = [ P JP PD ( E NTD NTD JP α 1 )] [ P JP PM ( j E j JP )] α (4) And we compare he exchange rae of he counries where Taiwan impor inermediae goods wih Japanese yen, preformed as (!"!!!!"). Afer deflaing he!! Taiwanese domesic and foreign inpu price by Japanese WPI, he value a he end is: 6

7 α 1 α ( pd ) ( pm ) (5) Where pd! and pm! sand for he raio of Taiwanese domesic and foreign inpu price o Japanese naional produc price, respecively. 3.2 Expor Demand The expor demand of a counry is usually se o be he funcion of expor price, impor counry s income, and impor counry s domesic price. Therefore, Japan s demand funcion for Taiwan s producs can be se like: X D = X JP P = f! JP TW K Y JP JP P, PX P JP TW /JP E, PX $ JP # P JP k/jp & = y " E % ( ) β 1 ( px ) β 2 k ( px ) β 3 Where X!! sands for Japan s demand for Taiwan s produc, X!!" (6) for Taiwan s expor volume o Japan, and P!!" for Japanese domesic produc price, which we use Japan s wholesale price index (WPI) o calculae. The firs variable in he bracke, Y!!" represens Japanese income, which we use Japanese real GDP o calculae. The second variable shows he raio of Taiwan s produc price o Japan s, where we also use Taiwan s WPI o calculae PX!!" and use exchange rae, E!!"/!", o calculae he relaive price. PX!! in he hird variable represens oher exporer s produc price in Japanese marke, where we also use exchange rae, E!!/!" o show he relaiviy. Noe ha in his formula, oher exporer, also called he hird counry, can be in a complemenary relaionship or in a compeiive relaionship wih Taiwan. Move on o exponens,, and. Firs of all, represens income elasiciy. When Japanese real GDP increases, we expec Japan s demand for Taiwan s producs also increase since i has greaer purchasing power. Therefore, we assume ha is greaer han zero. Nex, means expor price elasiciy. If Taiwan s expor price rises relaive o Japan s, hen Taiwan will expor fewer producs o Japan as he laer s demand declines. Thus, we expec smaller han zero. Thirdly, can be 7

8 greaer or smaller han zero, depending on he hird counry s relaionship wih Taiwan. If is produc is compeiive wih Taiwan s produc, hen will be greaer han zero because when Taiwan s compeior s expor price relaively rises, Taiwan s produc is more compeiive and aracive in Japanese marke. Conversely, if hey are in a complemenary relaionship, hen is smaller han zero. Now we subsiue px! wih (pd! )! (pm)!!!!, which is formula (5) we derive from supply side par. The formula becomes: x JP β1 β ( ) ( ) ( ) ( ) ( ) ( ) ( ) 2 k β3 JP β1 αβ2 1 α β β 2 k y px px = y pd pm ( px ) 3 = (7) 4. Regression Analysis 4.1 Taiwan Toal Expor o Japan Regression Model: logy = C + log x + x (8) 1 + log x2 + log x3 log 4 Y: Taiwan s oal expor o Japan from he firs quarer of 1999 o he second quarer of 2012 (uni: Japan s yen) x! : The exchange rae beween Japan and Taiwan from he firs quarer of 1999 o he second quarer of 2012 (Uni: JPY/NTD) x! : Japan s real GDP (uni: Japan s yen) x! : Taiwan s produc cos x! : Taiwan s main compeior s produc cos Y is he represenaive for Taiwan s oal expor o Japan from he firs quarer of 1999 o he second quarer of We derive he number (able 1) from he bureau of foreign rade, and hen we divide he nominal expors by Taiwan s WPI o be real expors. Moreover, we use he Japan s yen o be he uni. x! is he represenaive for he exchange rae beween Japan and Taiwan from he firs quarer of 1999 o he second quarer of 2012 (Uni: JPY/NTD) 8

9 The coefficien of logx! is posiive, which means ha here is a posiive relaionship beween Taiwan s expor o Japan and exchange raes. When he value of x! becomes higher, 1 NT dollar can exchange more yens (in oher words, Japanese yen depreciaes) and Taiwan s expor o Japan will be higher, oo; When he value of x! become lower, 1 NT dollar exchange less yens (yen appreciaes) and Taiwan s expor o Japan will be lower. x! is he represenaive for Japan s real GDP. Firsly, we derive he daa of he nominal GDP (able 2) from OECD, and hen we divide he nominal GDP by Japan s CPI o be real GDP. Secondly, o make he regression model more accurae, we use yen o be he uni. The coefficien of logx! is posiive, which means ha here is a posiive relaionship beween Taiwan s expor o Japan and Japan s real GDP. The increase of he value of x! means ha Japan s real GDP is on he increase; he decrease of he value of x! means ha Japan s real GDP is on he decrease. Table 3 presens he resuls for he oal expors model. The coefficiens of logx! and logx! are posiive, and i means when New Taiwan dollar appreciaes, Taiwan s expor o Japan will be on he increase, and when Japan s real GDP becomes higher, Taiwan s expor o Japan will also be higher. In addiion, as can be seen from able 5, Japan is Taiwan s Top1 impor counry. As yen depreciaes, he cos of goods impored from Japan will be lower and herefore he prices of goods can be lower and compeiive. We can deermine ha he rade relaionship beween he wo counries is he parner. GDP includes household consumpion expendiures and personal consumpion expendiures (C), gross privae domesic invesmen (I), governmen consumpion and gross invesmen expendiures (G), and ne expor (X-M). As Japan s real GDP is on he increase, in erms of household consumpion expendiures and personal consumpion expendiures(c), we assume ha Japanese s 9

10 income will be higher and Japanese will be more willing o buy producs from Taiwan. In erms of governmen consumpion and gross invesmen expendiures, Japan s governmen will be richer and more willing o buy Taiwan s producs. Therefore, Taiwan s expor o Japan will be higher. The las wo variables represen he hird counry effec in he regression model. x! is he represenaive for Taiwan s produc cos. We sugges ha Taiwanese companies impor maerials mainly from counry J, which is he se of he imporers in Taiwan. x! is lierally he raio of se J s weighed average price level o Japanese price level. The following are he seps we calculae he weighed average price level of he se J. Firs, in he able 5, we choose he op few counries on he rank of main imporers of Taiwan. Among hese imporers, we pick up 6 imporers which impor mainly similar indusry producs of Taiwan, which are elecrical machinery and machinery, and delee oher imporers which impor mainly oher unrelaed producs. The 6 imporers we selec o be se J are Japan, Korea, Germany, Singapore, he Neherlands, and Hong Kong. Second, we calculae he weigh of each counry based on he rade volume of each counry from 1999 o The weigh of each counry is he raio of he accumulaed oal impor o Taiwan from each counry o he accumulaed oal impor o Taiwan from hese six counries. (Table 8-1) Lasly, we use he price index of each counry o muliplier is own weigh we figured ou on he second sep; he produc is x!, he weighed average price index of se J. Because x! is he represenaive for Taiwan s produc cos, he negaive coefficien of x! here in our regression model means ha once he cos of Taiwanese produc rises, he oal expor from Taiwan o Japan falls. This regression resul is consisen wih our assumpion presened hereofore; ha is, 10

11 once he cos of Taiwanese producs go lower, he price of Taiwanese producs go lower, and Taiwanese producs would be more compeiive han oher counries producs in Japan s marke, resuling in he growh of Taiwan s oal expor o Japan. x! is he represenaive for Taiwan s main compeior s produc cos. Firs, we sugges ha Taiwan s main compeiors impor maerials mainly from counry K, which is he se of he imporers of Taiwan s main compeiors. x! is lierally he raio of se K s weighed average price level o Japanese price level. The following are he seps we calculae he weighed average price level of he se K. Firs, we search on he World Trade Alas (WTA) o selec 6 counries in Japan s op imporers whose Expor Similariy Index (ESI) beween hem and Taiwan are more han 30%. (Table 7) The 6 imporers we selec o be se K are America, Souh Korea, Malaysia, Germany, Thailand, and Singapore. Second, we calculae he weigh of each counry based on he rade volume of each counry from 1999 o The weigh of each counry is he raio of he accumulaed oal impor o Japan from each counry o he accumulaed oal impor o Japan from hese six counries. (Table 8-2) Las, we use he price index or WPI (able 9) of each counry o muliplier is own weigh we figured ou on he second sep; he produc is x!, he weighed average price index of se K. The value answers wheher Taiwan has he srengh over is compeiors in Japanese marke or no. The regression resul is consisen wih our assumpion presened hereofore; ha is, since he coefficien is posiive, i claims ha once he prices of Taiwan s compeiors producs rise, Taiwanese producs may be relaively cheaper and more aracive, resuling in he growh of Taiwan s oal expor o Japan. However, his resul is no saisically significan since he p-value is somehow big; his is perhaps due o he assumpion incompleeness of ESI. 11

12 To sum up, our discussion, he hird counry effec in hese wo variables is consisen wih our assumpion hereofore. 4.2 Elecrical Machinery According o he World Trade Alas (WTA), he goods ha Japan impors he mos from Taiwan is elecrical machinery in recen years, which shows ha he elecrical machinery can be a represenaive indusry o sudy. Table 12: Japan Impor Elecrical Machinery from Taiwan And since he ypes of goods conained in HS 85 are oo broad and may overlap he nex indusry we wan o discuss, we choose he real volume of elecronic producs, communicaion apparaus and household appliance expored o Japan from quarer one in 2004 o he quarer four in 2012(based on he caegory given by he saisics deparmen from he Minisry of Finance) o be he dependen variable. Keep he independen variables he same as he previous seing, he regression funcion is: logy = log x log x log x log x 4 And he saisical informaion from he sofware eviews is: Table 13: Regression Resul from Eviews of Elecrical Machinery (9) 12

13 From able 13 he R-squared is approximaely 60 percen, and he adjused R-squared is more han 50 percen as well, which saes ha he regression funcion is efficien in explaining he dependen variable. In addiion, since he p-values of x! and x! are oo big for us o believe ha hey are accurae enough o explain he dependen variable, he following demonsraion will be focused on x! and x!, which have relaively low p-values. x! is he real exchange rae of Japanese yen. And is coefficien is negaive, which means ha Taiwan will expor less elecrical machinery o Japan as long as he value of x! becomes bigger. To describe his in anoher way, he depreciaion of yen pus a pressure on Taiwanese expors, since he elecrical machinery from Japan will be cheaper and hus more compeiive han ours. x! is he raio of Japanese imporers expor price level o Japanese WPI. The value answers wheher Taiwan has he srengh over is compeiors in Japanese marke or no. Since he coefficien is posiive, i claims ha once he prices of elecrical machinery from Taiwanese compeiors rise, Taiwanese goods may be relaively cheaper and more aracive, and Taiwan will expor more hings o Japan as a resul. In ligh of he informaion provided above, he rade relaionship beween Taiwan and Japan is compeiion in he aspec of elecrical machinery. Take he compuer 13

14 producion indusry, which is a crucial one among he whole elecrical machinery, for example. According o he repor from Taiwan Insiue of Economic Research (Chen, 2013), he suppliers in he compuer producion indusry end o produce he goods in a mass and sandard way o reduce he cos. Such a kind of producion mehod shows ha he indusry is geing maure because of he improvemen of he echnology, and will evenually ransformed ino he complee price compeiion. 4.3 Machinery The second indusry ha we choose o be a represenaive one is machinery indusry. According o WTA, in he recen hree years, among Taiwan s oal expors o Japan, machinery expors ranks afer elecrical machinery expors. As he able shown below, he percenage is abou 7~8%. Table 14: Taiwan s Machinery Expors o Japan in he Recen Three Years Machinery includes meal working machine and power generaing machine ec. Given ha he dependen variable sands for Taiwan s machinery expors o Japan and oher independen variables are se he same as he previous seing, here liss he regression model: logy = logx log x log x log x (10) 4 14

15 Also, here shows he saisical informaion from he sofware eviews: Table 15: Regression Resul from Eviews of Machinery From he above able, firs, by looking a R-squared, we infer ha abou 75% of he dependen variable can be explained by his regression model. Second, we found ha only wo variables have heir p-values less han 0.05, which means ha only X! and X! are saisically significan o he model. Before elaboraing on he resul we found, we have o explain some sligh changes of he samples we picked. Firsly, i s abou X! which involves Taiwan s main machinery imporers. Originally we have Japan, Korea, Germany, Singapore, he Neherlands, and Hong Kong o be he sample counries, bu Hong Kong is deleed because based on he saisic from Naional Saisic ROC websie, we found ha from 2004 o las year, among hose who expor machinery goods o Taiwan, he accumulaed impors from Hong Kong only ranks he welfh. Therefore we don consider i o be one of Taiwan s main imporers. Secondly, i s abou X! which deals wih Japan s main machinery imporers. Here liss our original sample counries: he US, Korea, Malaysia, Germany, Thailand, and Singapore. However, we decided o delee Malaysia and Singapore since based on 15

16 WTA daabase, we found ha in he recen hree years, among hose who expor machinery goods o Japan, boh Malaysia and Singapore have a relaively small proporion (less han 2%) compared wih oher Japan s imporers. Thus, hey are no appropriae samples for our model. Nex, moving back o he regression model, as he elecrical machinery par, we focus on X! and X!. Unlike he previous par, he coefficien of X! is posiive in his par, which indicaes ha when Yen appreciae, Taiwan s machinery expor o Japan decrease. We use wo graphs o illusrae wha happened here. Figure 5: Exchange Rae (NTD/JPY) Figure 6: Taiwan s Machinery Expors o and Impors from Japan (uni: USD) 16

17 As you can see from graph A, unil 2012, Yen appreciaed as a whole. In addiion, graph B shows ha Taiwan s machinery expors o Japan decrease as a whole. On he oher hand, as he red line in graph B shows, Taiwan impors a lo more machinery goods from Japan han Taiwan expors. We herefore infer ha Taiwan impors los of inermediae machinery goods from Japan bu expors far less machinery producs o i. When Yen appreciae, Taiwan has o expend more o buy machinery pars from Japan. Since he cos increases, he price of machinery producs Taiwan sells also increases. Therefore, Japan s demand for Taiwan s goods declines as Taiwan s producs have no advanage in price. In shor, Taiwan depends a lo on Japanese machinery goods and Taiwan s machinery indusry is no so compeiive wih Japans as elecrical machinery indusry is. Tha s why he coefficien of X! is posiive. As for X!, like he explanaion in elecrical par, he coefficien is posiive. When he prices of Taiwan s compeiors machinery producs from rise, Taiwan has greaer chance o expor more o Japan. 5. Conclusion The resul shows ha he depreciaion of Japanese yen will lead o he increase in Taiwan s expor o Japan. And when i comes o separae indusry, i has differen 17

18 consequence. The depreciaion will make he expor of elecrical machinery decrease and make he machinery expor increase. Japan is Taiwan s bigges impor counry, he rade volume beween Taiwan and Japan shows huge defici (Figure 7). Therefore, if Japanese yen depreciaes, in he shor erm, we will expec he rade defici o reduce. Figure 7: Trade Defici beween Taiwan and Japan in million Daa from Cenral Bank of R.O.C The impac of yen depreciaion on differen counries depends on he counry s relaive posiion in Japan s produc supply chain. If Japan is he counry s supplier and he counry is Japan s las producs consumer, he counry will ge benefi from i. On he oher hand, if he counry has he similar expor srucure o Japan, i will become he vicim. Figure 8 shows Asia s counries share of rade o Japan. Taiwan has huge rade defici o Japan. So he depreciaion o Taiwan is beneficial. And Taiwan s companies will increase he invesmens in Japan or buy heir echniques. Alhough Korea has rade defici o Japan, i also has similar expor srucure o Japan. So i can be consider as Japan s compeior. And oher counries in Souh Asia have complemenary expor srucure o Japan. Therefore, hey will ge benefi from i. Figure 8: Asia s Counries Share of Trade o Japan 18

19 18 % Malaysia Singapore Hong Kong Thailand Phillipine Taiwan Indonisia Korea China India Expor Impor Daa from Cenral Bank of R.O.C Table 1: Taiwan s oal expor o Japan Quarer JPY/USD TWD/USD JPY/TWD TW o JP nominal TW o JP expors (USD) nominal expors(jpy) JP- WPI original TW o JP real expors(jpy) 1999Q ,699,579, ,563,063, ,220,493, Q ,850,961, ,687,003, ,829,073, Q ,868,647, ,875,472, ,051,407, Q ,480,893, ,847,331, ,354,362, Q ,685,838, ,856,468, ,504,728, Q ,210,953, ,279,289, ,364,025, Q ,213,788, ,795,509, ,900,165, Q ,488,476, ,942,409, ,370,627, Q ,855,330, ,932,822, ,294,786, Q ,418,222, ,221,050, ,592,440, Q ,697,990, ,404,739, ,766,382, Q ,053,036, ,260,698, ,726,902, Q ,014,025, ,352,341, ,887,124, Q ,283,080, ,261,798, ,064,587, Q ,974,234, ,757,808, ,515,643, Q ,096,431, ,170,481, ,332,048, Q ,989,312, ,611,571, ,991,861, Q ,091,025, ,394,656, ,609,663, Q ,030,866, ,911,623, ,579,074, Q ,318,422, ,110,770, ,317,528, Q ,305,254, ,167,911, ,286,400, Q ,412,503, ,443,767, ,734,904,

20 2004Q ,529,207, ,958,720, ,432,450, Q ,560,684, ,645,628, ,334,876, Q ,615,552, ,980,743, ,298,036, Q ,721,466, ,169,335, ,132,052, Q ,723,621, ,660,824, ,176,987, Q ,049,877, ,746,895, ,166,033, Q ,931,105, ,495,096, ,091,408, Q ,892,754, ,716,519, ,277,071, Q ,219,018, ,477,758, ,376,999, Q ,257,178, ,401,985, ,553,040, Q ,159,062, ,513,033, ,261,224, Q ,809,570, ,188,521, ,640,722, Q ,859,303, ,405,989, ,713,469, Q ,105,378, ,371,624, ,132,366, Q ,131,328, ,764,487, ,537,007, Q ,499,921, ,826,768, ,784,969, Q ,681,690, ,782,685, ,324,550, Q ,242,781, ,650,682, ,435,465, Q ,212,848, ,828,627, ,730,806, Q ,376,855, ,746,997, ,823,910, Q ,805,312, ,846,217, ,951,990, Q ,106,952, ,750,902, ,060,317, Q ,162,740, ,339,951, ,919,438, Q ,400,565, ,314,040, ,317,846, Q ,844,865, ,442,354, ,244,773, Q ,597,469, ,484,367, ,363,798, Q ,415,541, ,213,606, ,152,903, Q ,649,576, ,168,329, ,594,821, Q ,623,228, ,744,485, ,617,019, Q ,539,422, ,006,266, ,830,863, Q ,196,662, ,807,958, ,761,681, Q ,684,868, ,168,977, ,704,718, Daa from Bureau of Foreign Trade of R.O.C and Inernaional Financial Saisics Table 2: Japan s nominal and real GDP Quarer JP- GDP(JPY) JP- CPI org. JP- CPI adj. JP- Real GDP (JPY) 20

21 1999Q1 115,743,915,493, Q2 116,196,774,316, Q3 116,046,679,677, Q4 116,620,484,180, Q1 118,555,890,046, Q2 118,747,558,199, Q3 118,422,444,812, Q4 119,264,014,750, Q1 120,065,011,138, Q2 119,787,096,065, Q3 118,483,342,637, Q4 118,336,622,878, Q1 118,115,730,769, Q2 119,292,088,748, Q3 120,065,786,110, Q4 120,539,419,192, Q1 119,898,442,088, Q2 121,346,940,346, Q3 121,853,322,257, Q4 123,147,901,014, Q1 124,451,404,451, Q2 124,358,407,773, Q3 124,551,950,859, Q4 124,244,561,840, Q1 124,523,001,893, Q2 126,102,595,468, Q3 126,576,853,527, Q4 126,811,270,154, Q1 127,369,925,198, Q2 127,865,757,486, Q3 127,803,834,698, Q4 129,441,876,185, Q1 130,733,505,047, Q2 130,886,199,593, Q3 130,419,466,265, Q4 131,545,976,025, Q1 132,434,969,269,

22 2008Q2 130,830,376,587, Q3 129,458,000,609, Q4 125,253,050,815, Q1 120,240,904,855, Q2 122,262,532,640, Q3 122,296,281,434, Q4 124,543,851,149, Q1 126,354,086,485, Q2 127,622,091,190, Q3 129,350,404,453, Q4 128,878,371,314, Q1 126,552,129,411, Q2 125,469,168,095, Q3 128,658,929,153, Q4 128,787,974,544, Q1 130,722,130,453, Q2 130,429,215,917, Q3 129,217,884,187, Q4 129,267,932,399, Daa from Organizaion for Economic Cooperaion and Developmen (OECD) hp:// Table 3: Toal Expors Model 22

23 Table 4: The Original Daa of Regression Model Year/ Quarer Y:Taiwan s oal expor o Japan (real) X1: JPY/TWD (real) X2:Japan s real GDP (JPY) X3: X4: 1999Q1 307,220,493, ,125,495,627, Q2 337,829,073, ,126,263,779, Q3 318,051,407, ,128,075,580, Q4 355,354,362, ,133,653,464, Q1 383,504,728, ,159,576,684, Q2 438,364,025, ,158,817,684, Q3 442,900,165, ,158,271,477, Q4 485,370,627, ,168,019,650, Q1 450,294,786, ,180,084,326, Q2 417,592,440, ,178,506,292, Q3 328,766,382, ,168,733,037, Q4 381,726,902, ,171,505,344, Q1 405,887,124, ,177,444,205, Q2 424,064,587, ,184,077,157, Q3 362,515,643, ,193,723,368, Q4 389,332,048, ,199,620,086, Q1 364,991,861, ,197,990,238, Q2 377,609,663, ,207,257,786, Q3 366,579,074, ,214,300,804, Q4 372,317,528, ,229,234,764, Q1 363,286,400, ,245,134,028, Q2 381,734,904, ,240,906,602, Q3 392,432,450, ,242,426,331, Q4 380,334,876, ,234,047,947, Q1 381,298,036, ,245,436,730, Q2 401,132,052, ,259,562,562, Q3 412,176,987, ,266,399,107, Q4 471,166,033, ,268,744,441, Q1 454,091,408, ,275,604,719, Q2 437,277,071, ,275,059,895, Q3 476,376,999, ,271,076,434, Q4 488,553,040, ,290,776,736, Q1 484,261,224, ,310,597,949,

24 2007Q2 443,640,722, ,306,043,975, Q3 434,713,469, ,298,805,428, Q4 442,132,366, ,304,847,720, Q1 409,537,007, ,314,964,881, Q2 432,784,969, ,287,995,128, Q3 449,324,550, ,262,110,766, Q4 378,435,465, ,229,872,280, Q1 288,730,806, ,195,464,441, Q2 319,823,910, ,215,563,959, Q3 345,951,990, ,219,521,839, Q4 361,060,317, ,248,131,575, Q1 366,919,438, ,267,116,130, Q2 393,317,846, ,278,130,151, Q3 404,244,773, ,302,802,050, Q4 369,363,798, ,295,449,094, Q1 350,152,903, ,275,897,891, Q2 361,594,821, ,262,022,934, Q3 344,617,019, ,294,106,925, Q4 337,830,863, ,298,439,669, Q1 319,761,681, ,313,982,042, Q2 360,704,718, ,309,290,279, Table 5: Taiwan s Top 20 Impor Counries (1999Q1 o 2012Q2) Rank Counry Impor( million dollar) Percen 1 JAPAN 547,726,206, UNITED STATES 298,083,336, CHINA 281,024,328, KOREA,REPUBLIC OF 159,813,244, SAUDI ARABIA 103,504,740, GERMANY,FEDERAL 84,428,949, REPUBLIC OF 7 MALAYSIA 76,645,858, AUSTRALIA 73,477,242, SINGAPORE 67,814,475, INDONESIA 62,607,094, KUWAIT 54,625,087,

25 12 THAILAND 40,506,567, PHILIPPINES 36,298,617, UNITED ARAB EMIRATES 30,605,946, NETHERLANDS 29,606,073, IRAN(ISLAMIC REPUBLIC OF) 29,088,033, FRANCE 28,670,167, HONG KONG 25,232,974, RUSSIA 25,040,073, UNITED KINGDOM 22,753,014, Daa from Bureau of Foreign Trade of R.O.C Table 6: The Exchange Rae of Each Counry Year TWD/JPY USD/JPY KRW/JPY GBP/JPY EUR/JPY SGD/JPY MYR/JYP PHP/JYP HK/JPY 1999Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q

26 2005Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Daa from Cenral Bank of R.O.C hp:// Table 7: Expor Similariy Index (ESI) from 1999 o 2012 Year TW- US TW- Souh Korea TW- Malaysia TW- Germany TW- Thailand TW- Singapore

27 Average ESI Daa from Bureau of Foreign Trade of R.O.C hp://cus93.rade.gov.w/fsci/ Table 8-1: The weigh of counry J Year Japan expor o Korea expor o Germany expor Singapore expor Neherland Hong Kong expor o TW(USD) TW(USD) o TW(USD) o TW(USD) expor o TW(USD) TW(USD) ,590,096, ,192,677, ,312,433, ,312,032, ,705,675, ,091,879, ,556,954, ,987,938, ,541,894, ,013,648, ,087,287, ,185,324, ,932,813, ,731,817, ,274,025, ,401,625, ,540,922, ,050,054, ,362,757, ,741,326, ,448,077, ,562,201, ,469,377, ,914,646, ,718,847, ,738,052, ,986,096, ,878,975, ,306,962, ,917,038, ,715,710, ,661,956, ,851,121, ,330,669, ,202,540, ,309,052, ,052,601, ,239,023, ,180,054, ,960,609, ,068,687, ,109,605, ,283,618, ,999,435, ,135,030, ,105,539, ,342,514, ,880,519, ,936,063, ,158,216, ,069,816, ,791,649, ,776,776, ,824,800, ,507,203, ,168,228, ,474,068, ,825,113, ,353,495, ,492,713, ,219,354, ,506,681, ,672,734, ,809,159, ,862,737, ,122,506, ,916,721, ,058,628, ,263,905, ,636,054, ,199,873, ,627,580, ,199,118, ,860,156, ,427,438, ,953,091, ,935,999, ,675,422, ,573,055, ,073,110, ,753,884, ,105,749, ,622,888, ,658,796, Toal 571,564,918, ,117,248, ,390,582, ,686,118, ,475,738, ,859,941, Weigh Daa from Bureau of Foreign Trade of R.O.C hp://cus93.rade.gov.w/fsci/ 27

28 Table 8-2: The weigh of counry K Year US expor o Korea expor o Malaysia expor Germany expor Thailand expor Singapore expor Japan(USD) Japan(USD) o Japan(USD) o Japan(USD) o Japan(USD) o Japan(USD) ,320 16,195 10,975 11,538 8,895 5, ,136 20,443 14,484 12,726 10,590 6, ,177 17,212 12,859 12,397 10,373 5, ,848 15,508 11,197 12,440 10,520 5, ,931 17,931 12,599 14,228 11,896 5, ,563 22,068 14,125 17,095 14,118 6, ,199 24,420 14,711 17,886 15,590 6, ,011 27,319 15,472 18,451 16,878 7, ,939 27,290 17,397 19,430 18,307 7, ,667 29,501 23,241 20,888 20,813 7, ,959 21,978 16,727 16,759 16,023 6, ,443 28,659 22,723 19,294 21,041 8, ,485 39,835 30,488 23,335 24,522 8, ,237 40,515 32,865 24,713 23,641 8,761 Toal 939, , , , ,206 94,844 Weigh Daa absraced from World Trade Alas Table 9: The price index or WPI of counry K (base period: 2005) Quarer US- Expor price org. KR- Expor Price org. MS- WPI org. GM- Expor Price org. TL- WPI org. SG- Expor Price org. 1999Q Q Q Q Q Q Q Q Q Q