56 5 21 2120 1 17 5 7 21

1 H KL K L K

EVWUDFW
7KLV SDSHU H DPLQHV WKH LPSDFW RI H FKDQJH UDWH ULVN RQ HFRQRPLF SHUIRUPDQFHV RI
9LHWQDP LQ SHULR E VLQJ WRUHJUHVVL H LVWULE WH /DJ 5 / DSSURDFK
7KH HPSLULFDO UHV OWV VKR WKDW WKHUH LV H LVWHQFH RI ORQJ U Q UHODWLRQVKLSV EHW HHQ 5((5
volatility and output. Impact of exchange rate risk is statistically signi cant on economic
SHUIRUPDQFHV LQ 9LHWQDP Q LQFUHDVH LQ H FKDQJH UDWH ULVN R O K UW HFRQRPLF JUR WK
7KH OH HO UHODWLRQVKLSV DOVR VKR WKDW JUR WK PD EH LPSUR H KHQ 5((5 HSUHFLDWHV
.H R GV ( FKDQJH UDWH ULVN 5 / DSSURDFK 2 WS W 9LHWQDP
DWH RI V EPLVVLR WK
S LO DWH RI H LVLR WK
S LO DWH RI DSS R DO WK
S LO

,1752 8&7,21 DLPLQJ WR UHIRUP HFRQRPLF VWUXFWXUH DWWUDFW

8QGHUVWDQGLQJ WKH OLQN EHWZHHQ H[FKDQJH IRUHLJQ LQYHVWPHQW HVWDEOLVK JUDGXDOO D
UDWH ULVN DQG HFRQRPLF SHUIRUPDQFHV VXFK QHZ OHJDO IUDPHZRUN GLUHFWLQJ WKH HFRQRP
DV HFRQRPLF JURZWK LQWHUQDWLRQDO WUDGH DQG WRZDUGV PDUNHW HFRQRP DQG WR LQWHJUDWH
LQYHVWPHQW LV RI SDUWLFXODU FRQFHUQ JLYHQ VWHS E VWHS LQWR WKH UHJLRQDO DQG JOREDO
WKDW PRUH HFRQRPLHV LQFOXGLQJ WUDQVLWLRQ HFRQRP )URP WKDW SRLQW RI YLHZ 9LHWQDP
DQG GHYHORSLQJ HFRQRPLHV KDYH DGRSWHG D JRYHUQPHQW KDV FRQWLQXHG WR GHUHJXODWH PDQ
more exible exchange rate arrangement. It VHFWRUV DQG FKDQJHG IURP GLUHFW FRQWURO WR
LQGLUHFW FRQWURO RI LWV HFRQRP LQFOXGLQJ
LV LPSRUWDQW WR XQGHUVWDQG WKH LPSOLFDWLRQV RI
H[FKDQJH UDWH V VWHP +RZHYHU ZKLOH
VXFK D PRYH IRU HFRQRPLF DFWLYLW ZKHQ WKH
Vietnam’s nancial market is still weak, the
transition to exible arrangement of exchange
UHOD[DWLRQ RQ WKH FRQWURO RI H[FKDQJH UDWH
UDWH KDV EHHQ DFFRPSDQLHG E LQFUHDVHV LQ WKH
FRXOG OHDG WR DQ LQFUHDVH LQ WKH H[FKDQJH UDWH
YRODWLOLW RI ERWK QRPLQDO DQG UHDO H[FKDQJH
ULVN XQFHUWDLQW UHVXOWLQJ LQ WKH LQFUHDVH LQ
UDWH &DSRUDOH DQG 3LWWLV +DVDQ DQG
GROODUL]DWLRQ 7KH LQFUHDVH LQ H[FKDQJH UDWH
DOODFH .HQW DQG 1DMD
YRODWLOLW FRXOG DIIHFW QHJDWLYHO WKH HFRQRPLF
2YHU HDUV 9LHWQDP KDV PDGH D VKLIW IURP SHUIRUPDQFHV DQG EULQJV VHULRXV FKDOOHQJHV
D FHQWUDOO SODQQHG HFRQRP WR D PDUNHW IRU 9LHWQDP LQ WKH SHULRG RI LQWHJUDWLQJ LQWR
HFRQRP 7KLV SHULRG LV NQRZQ DV RL 0RL world’s economy. Therefore, it is worthwhile
5HQRYDWLRQ 7KH HFRQRPLF UHIRUP KDV EHHQ WR LQYHVWLJDWH WKH HIIHFW RI H[FKDQJH UDWH
LPSOHPHQWHG WKURXJK D YDULHW RI PHDVXUHV ULVN RQ JURZWK LQ 9LHWQDP FRQWH[W 7KH

PhD, Foreign Trade University , Vietnam; Email: vinhntt@ftu.edu.vn.

1R ( ( 1 / ( 2120 6 ( (
 56 5 21 2120 1 17 5 7 21

ndings may have signi cant implications for WUDGH ZHLJKWHG H[FKDQJH UDWH UHDO HIIHFWLYH
Vietnam to nd an exchange rate arrangement H[FKDQJH UDWH RU 5((5 7KHLU UHVHDUFK
WKDW LV VXLWDEOH IRU LWV SURFHVV RI DFFHVVLQJ nds that the effective ULVN
LQWR WKH ZRUOG HFRQRP ZKLOH HQVXULQJ WKH YDULDEOH SHUIRUPV PXFK EHWWHU WKDQ WKH
FRQWLQXDWLRQ RI UDSLG HFRQRPLF JURZWK ELODWHUDO H[FKDQJH UDWH YRODWLOLW ZLWK
7KHUH DUH WZR DSSDUHQWO RSSRVLWH YLHZV LQ GROODUV 7KHUHIRUH WKLV VWXG LQYHVWLJDWHV WKH
WKH OLWHUDWXUH ZKHQ GHDOLQJ ZLWK WKH HIIHFWV LPSDFW RI UHDO HIIHFWLYH H[FKDQJH UDWH ULVN RQ
RI H[FKDQJH UDWH UHJLPHV DQG YRODWLOLW RQ JURZWK IRU 9LHWQDP DSSO LQJ $XWRUHJUHVVLYH
economic growth. The rst view is called 'LVWULEXWHG /DJ $5'/ ERXQGV WHVWLQJ
D FRVWV RI RODWLOLW ´ DUJXPHQW 0DQ PHWKRGRORJ RI 3HVDUDQ HW DO WR WKH
HFRQRPLVWV EHOLHYH WKDW H[FKDQJH UDWH DQDO VLV RI OHYHO UHODWLRQVKLSV IRU PRGHOV
XQFHUWDLQW UHGXFHV LQWHUQDWLRQDO WUDGH LQFOXGLQJ RXWSXW UHDO HIIHFWLYH H[FKDQJH UDWH
GLVFRXUDJHV LQYHVWPHQW DQG FRPSRXQGV WKH YRODWLOLW DQG WKH UHDO HIIHFWLYH H[FKDQJH
SUREOHPV SHRSOH IDFH LQ LQVXULQJ WKHLU KXPDQ UDWH 7KH HPSLULFDO UHVXOWV VKRZ WKDW WKHUH
FDSLWDO LQ LQFRPSOHWH DVVHW PDUNHWV 2EVWIHOG LV H[LVWHQFH RI ORQJ UXQ UHODWLRQVKLSV
DQG 5RJRII 'H *UDXZH DQG 6FKQDEO EHWZHHQ 5((5 ULVN DQG RXWSXW ,PSDFW RI
HPSKDVL]H WKDW ZKHQ H[FKDQJH UDWH REER volatility is statistically signi cant
PRYHPHQWV DUH DQ LQGHSHQGHQW VRXUFH RI RQ HFRQRPLF SHUIRUPDQFHV LQ 9LHWQDP
YRODWLOLW DQG DUH DOVR GULYHQ E VSHFXODWLYH $Q LQFUHDVH LQ 5((5 YRODWLOLW ZRXOG KXUW
G QDPLFV DQWLFLSDWHG HQWU LQWR PRQHWDU HFRQRPLF JURZWK 7KH OHYHO UHODWLRQVKLSV
XQLRQV PD KHOS VPDOO RSHQ HFRQRPLHV WR DOVR VKRZ WKDW JURZWK PD EH LPSURYHG
DYRLG QHJDWLYH PDFURHFRQRPLF HIIHFWV RI ZKHQ 5((5 GHSUHFLDWHV
H[FKDQJH UDWH YRODWLOLW 6FKQDEO
7KLV SDSHU LV RUJDQL]HG DV IROORZ 6HFWLRQ
also nds a positive association between
SURYLGHV D EULHI RI WKH WKHRUHWLFDO OLWHUDWXUH
H[FKDQJH UDWH VWDELOLW DQG JURZWK LQ
ZKLFK RXWOLQHV WKH UHDVRQLQJ EHKLQG ZK
PRVWO VPDOO RSHQ HFRQRPLHV DW (XURSHDQ
LQFUHDVHG H[FKDQJH UDWH ULVN PLJKW KXUW RU
0RQHWDU 8QLRQ 7KH VHFRQG YLHZ ZKLFK
KHOS WKH HFRQRPLF JURZWK 7KLV VHFWLRQ DOVR
is de ned as “advantage of exibility´
FRYHUV WKH HPSLULFDO OLWHUDWXUH UHYLHZ 1H[W
DUJXPHQW FRQVLGHUV WKDW WHUPV RI WUDGH
shocks are ampli ed in countries with more 6HFWLRQ VKRZV DQ DSSURSULDWH PHDVXUH RI
ULJLG H[FKDQJH UDWH UHJLPHV DQG WKDW DIWHU H[FKDQJH UDWH ULVN IRU 9LHWQDP ,Q 6HFWLRQ DQ
FRQWUROOLQJ IRU RWKHU IDFWRUV FRXQWULHV ZLWK HPSLULFDO VWXG LV FRQGXFWHG XVLQJ D DSSURDFK
exible exchange rate regimes grow faster LQWURGXFHG E 3HVDUDQ HW DO WR DQDO VH
(GZDUGV DQG H DWL 7KH DGYDQWDJH OHYHO UHODWLRQVKLSV &RQFOXGLQJ UHPDUNV DUH
of exibility effect also seems supported by RXWOLQHG LQ 6HFWLRQ
HPSLULFDO HYLGHQFH 7+(25(7,& / )5 0(:25. 1
DJHOOD HW DO VXSSRVH WKDW WKHVH /,7(5 785( 5(9,(:
apparently con icting views can be easily 7KHRUHWLFDO HYLGHQFH FRQFHUQLQJ WKH LPSDFW
UHFRQFLOHG ZKHQ H[FKDQJH UDWH YRODWLOLW RI H[FKDQJH UDWH VWDELOLW RQ JURZWK LV PL[HG
LV SURSHUO PHDVXUHG ZLWK D PXOWLODWHUDO The theoretical arguments in favor of exible

( ( 1 / ( 2120 6 ( ( 1R
 56 5 21 2120 1 17 5 7 21

H[FKDQJH UDWHV DUH PDLQO RI PDFURHFRQRPLF VDPSOHV VXFK DV E *KRVK HW DO WKHUH

nature, as exible exchange rates allow for an LV ZHDN HYLGHQFH WKDW H[FKDQJH UDWH VWDELOLW
HDVLHU DGMXVWPHQW LQ UHVSRQVH WR DV PPHWULF DIIHFWV JURZWK LQ D SRVLWLYH RU QHJDWLYH ZD
country speci c real shocks. In contrast, +RZHYHU 6FKQDEO LQYHVWLJDWHV WKH
there are also macroeconomic bene ts of LPSDFW RI H[FKDQJH UDWH VWDELOLW RQ JURZWK IRU
xed exchange rates as they contribute to D VDPSOH RI PRVWO VPDOO RSHQ HFRQRPLHV
PDFURHFRQRPLF VWDELOLW DQG KHOS WR DYRLG DW WKH (08 SHULSKHU DQG UHYHDOV D UREXVW
EHJJDU WK QHLJKERXU´ GHSUHFLDWLRQV LQ KLJKO QHJDWLYH UHODWLRQVKLS EHWZHHQ H[FKDQJH UDWH
LQWHJUDWHG HFRQRPLF UHJLRQV )XUWKHUPRUH LW YRODWLOLW DQG JURZWK IRU FRXQWULHV LQ WKH
LV DUJXHG WKDW IRU VPDOO RSHQ FRXQWULHV LQ WKH HFRQRPLF FDWFK XS SURFHVV ZLWK RSHQ FDSLWDO
economic catch-up process xed exchange DFFRXQWV
UDWHV SURYLGH D PRUH VWDEOH HQYLURQPHQW IRU 7KH SUHYLRXV ZRUN LQYHVWLJDWLQJ HIIHFW RI
WKH DGMXVWPHQW RI DVVHW DQG ODERU PDUNHWV H[FKDQJH UDWH XQFHUWDLQW RYHU WLPH RQ JURZWK
0F.LQQRQ 6PDOO RSHQ HFRQRPLHV
ZLWKLQ D VLQJOH FRXQWU LV YHU IHZ $PXHGR
with exible exchange rate regimes are argued
'RUDQWHV DQG 3R]R D FDQQRW GLVFHUQ
WR KDYH KLJKHU ULVN SUHPLXPV RQ LQWHUHVW UDWHV
any signi cant impact of real exchange
DV XQFHUWDLQW LQ DVVHW PDUNHWV LV LQFUHDVLQJ
UDWH XQFHUWDLQW RQ WKH SURGXFWLRQ LQGH[HV
6FKQDEO )URP D PLFURHFRQRPLF
of four countries: Korea, Singapore, Chile
SHUVSHFWLYH ORZ H[FKDQJH UDWH YRODWLOLW FDQ
DQG 0H[LFR $]LG HW DO LQYHVWLJDWH
EH DVVRFLDWHG ZLWK ORZHU WUDQVDFWLRQ FRVWV IRU
LPSDFW RI H[FKDQJH UDWH YRODWLOLW RQ JURZWK
international trade and capital ows thereby
in Pakistan and nd that this relationship is
FRQWULEXWLQJ WR KLJKHU JURZWK
positive but insigni cant.
The empirical literature tends to nd weak
HYLGHQFH LQ IDYRU RI D QHJDWLYH LPSDFW RI (;&+ 1 ( 5 7( 5,6. 0( 685(0(17
H[FKDQJH UDWH YRODWLOLW RQ JURZWK 7KH SDQHO ,Q WKH OLWHUDWXUH ZH VDZ WKDW PRVW SDSHUV
HVWLPDWLRQV IRU FRXQWULHV E (GZDUGV DQG PHDVXULQJ H[FKDQJH UDWH ULVN XVHG WKH
/HY H DWL RYHUWKH SHULRG YRODWLOLW RI ELODWHUDO H[FKDQJH UDWH H DUJXH
nd evidence that countries with more exible WKDW WKLV LV QRW WKH PRVW DFFXUDWH PHDVXUH RI D
H[FKDQJH UDWHV JURZ IDVWHU WKDQ FRXQWULHV ZLWK country’s exchange risk for the case of using
xed exchange rate. Eichengreen and Leblang DJJUHJDWH GDWD DJHOOD HW DO VKRZ
UHYHDO D VWURQJ QHJDWLYH UHODWLRQVKLS DGYDQWDJHV RI UHDO HIIHFWLYH H[FKDQJH UDWH
EHWZHHQ H[FKDQJH UDWH VWDELOLW DQG JURZWK 5((5 YRODWLOLW FRPSDULQJ ZLWK ELODWHUDO
IRU FRXQWULHV RYHU D SHULRG RI HDUV exchange rate volatility and nd that this
They nd that less exible exchange rate YDULDEOH SHUIRUPV PXFK EHWWHU WKDQ WKH ELODWHUDO
UHJLPHV DUH DVVRFLDWHG ZLWK VORZHU JURZWK H[FKDQJH UDWH YRODWLOLW )LUVW LW LQFOXGHV
and suggest that a more exible exchange trade partners’ externalities in the evaluation
rate is especially bene cial for low-income RI WKH HIIHFWV RI H[FKDQJH UDWH YRODWLOLW RQ
FRXQWULHV 7KH FRQFOXGH WKDW WKH UHVXOWV RI HFRQRPLF SHUIRUPDQFHV 7KLV LQFOXVLRQ LV
VXFK HVWLPDWLRQV VWURQJO GHSHQG RQ WKH WLPH IXQGDPHQWDO EHFDXVH D FRXQWU PLJKW KDYH
SHULRG DQG WKH VDPSOH )RU ODUJH FRXQWU ORZ ELODWHUDO H[FKDQJH UDWH YRODWLOLW ZLWK D

1R ( ( 1 / ( 2120 6 ( (
 56 5 21 2120 1 17 5 7 21

OHDGLQJ FXUUHQF IRU H[DPSOH WKH 86 GROODU DQG &DQDGD (DFK VHOHFWHG WUDGLQJ SDUWQHU
IRU FDVH RI 9LHWQDP EXW LPSRUW LQVWDELOLW accounted for at least 0.5 percent of Vietnam’s
YLD YDULDELOLW RI JRYHUQDQFH DQG HFRQRPLF WRWDO IRUHLJQ WUDGH GXULQJ WKH ODVW WHQ HDUV
SROLFLHV RI LWV WUDGH SDUWQHUV ,QGLYLGXDO 7KH EDVNHW FRYHUHG DERXW SHUFHQW RI WRWDO
country stability is therefore insuf cient LI LW WUDGH LQ HYHU HDU VLQFH 7KH WUDGH
LV QRW IUDPHG LQWR UHJLRQDO VWDELOLW DQG WKLV ZHLJKWV DUH UH FDOFXODWHG DQQXDOO H MW LV WKH
LV ZK WKH 5((5 YRODWLOLW YDULDEOH LV PRUH QRPLQDO ELODWHUDO H[FKDQJH UDWH UHODWLYH WR
OLNHO WR PHDVXUH WKH FRVWV RI PLVVLQJ UHJLRQDO FXUUHQF M PHDVXUHG DV WKH QXPEHU RI XQLWV
LQWHJUDWLRQ $ VHFRQG LPSRUWDQW DGYDQWDJH RI RI FXUUHQF M SHU XQLW RI WKH GRPHVWLF FXUUHQF
WKLV PHDVXUH LV WKDW IDYRUDEOH DQG XQIDYRUDEOH DQG H[SUHVVHG DV DQ LQGH[ MW LV WKH ZHLJKW
H[FKDQJH UDWH PRYHPHQWV ZLWK GLIIHUHQW DVVLJQHG WR FXUUHQF M DW WLPH W, re ecting the
SDUWQHUV PD FRPSHQVDWH HDFK RWKHU WKHUHE FRQWULEXWLRQ RI WKH FRXQWU RI FXUUHQF M WR WKH
GDPSHQLQJ WKH QHJDWLYH HIIHFWV RI LQGLYLGXDO Vietnam’s foreign trade; 3W LV WKH GRPHVWLF
ELODWHUDO H[FKDQJH UDWH YRODWLOLW RQ HFRQRPLF SULFH LQGH[ DW WLPH W DQG 3MW LV WKH SULFH LQGH[
SHUIRUPDQFHV 4LDQ DQG 9DUDQJLV 7KLV RI IRUHLJQ FRXQWU M DW WLPH W
HIIHFW LV LQFRUSRUDWHG LQWR 5((5 YRODWLOLW )LJXUH GHVFULEHV WKH PRYHPHQW RI 5((5
PHDVXUH ZKLFK FRQYHQLHQWO WDNHV LQWR DQG ELODWHUDO UDWH RI 9LHWQDP 'RQJ DJDLQVW
account the potential impact of diversi cation 86 'ROODU LQ ERWK UHDO DQG QRPLQDO WHUP IRU
RI WUDGH DQG LQYHVWPHQW SHULRG -DQXDU 0DUFK )URP
7KHQ ZH XVH D PHDVXUH RI UHDO H[FKDQJH UDWH )LJXUH LW FDQ EH VHHQ FOHDUO WKDW 9LHWQDPHVH
YRODWLOLW ZKHUH ELODWHUDO H[FKDQJH UDWHV DUH DXWKRULWLHV ZHUH LQFOLQHG WR PDLQWDLQ D VWDEOH
ZHLJKHG IRU WKHLU UHODWLYH WUDGH VKDUH 7KH QRPLQDO 86' 91' H[FKDQJH UDWH ZLWK
real effective exchange rate index is de ned VWURQJ GHSUHFLDWLRQ LQ EHFDXVH
LQ IRUHLJQ FXUUHQF WHUPV DQ LQFUHDVH LQ LWV of applying a oating exchange rate system
YDOXH LQGLFDWHV DQ DSSUHFLDWLRQ RI 9LHWQDPHVH and in 1997-1998 due to the Asian nancial
FXUUHQF DQG LV HVWLPDWHG E WKH JHRPHWULF FULVLV 1RPLQDO H[FKDQJH UDWH DOVR VOLJKWO
mean method as the following formula: GHSUHFLDWHG LQ SHULRG DIWHU WKH
Q
3 JRYHUQPHQW LQWURGXFHG D QHZ SULQFLSOH IRU
5((5W = ∏ H MW W MW

VHWWLQJ WKH H[FKDQJH UDWH 5((5 DQG 5(5

M= 3MW
WHQG WR DSSUHFLDWH GXULQJ DV WKH
ZKHUH 5((5W LV WKH UHDO HIIHFWLYH H[FKDQJH in ation differential became unfavorable
UDWH RI 9LHWQDP DW WLPH W Q LV WKH QXPEHU RI to Vietnam. When the Asian nancial
WUDGLQJ SDUWQHU FXUUHQFLHV LQ WKH WUDGH EDVNHW FULVLV EURNH RXW LQ 5((5 RI 9LHWQDP
7KH FXUUHQF EDVNHW LQFOXGHV WZHQW RI FRQWLQXHG WR DSSUHFLDWH VKDUSO EHFDXVH RI
Vietnam’s largest trading partners during the WKH SUHFLSLWRXV IDOOV LQ WKH YDOXHV RI D QXPEHU
ODVW HDUV -DSDQ 6LQJDSRUH &KLQD 8QLWHG of Asian trading partner’s currencies. After
6DWHV 6RXWK .RUHD $XVWUDOLD 7KDLODQG VXFK ODUJH FKDQJHV WKH 5((5 PD KDYH
*HUPDQ +RQJ .RQJ 0DOD VLD )UDQFH DWWUDFWHG JUHDWHU DWWHQWLRQ WKDQ EHIRUH 7KH
,QGRQHVLD 8QLWHG .LQJGRP 1HWKHUODQGV 86' 91' UDWH XQGHUZHQW VHYHUDO ODUJH
3KLOLSSLQHV ,WDO 6ZLW]HUODQG ,QGLD 6SDLQ scale devaluations in 1998, suf cient to

( ( 1 / ( 2120 6 ( ( 1R
 56 5 21 2120 1 17 5 7 21

EULQJ DERXW D FRQVLGHUDEOH GHSUHFLDWLRQ LQ H[FKDQJH UDWH XQFHUWDLQW IRU WLPH VHULHV
5((5 DQG 5(5 HVSHFLDOO ZKHQ $VLDQ DQDO VLV ,Q DQRWKHU VWXG $PXHGR 'RUDQWHV
FXUUHQFLHV EHJDQ WR VWDELOL]H DJDLQVW WKH 86' DQG 3R]R E LQYHVWLJDWH WKH UHODWLRQVKLS
with domestic in ation being kept not too far EHWZHHQ H[FKDQJH UDWH YRODWLOLW DQG )',
RXW RI OLQH ZLWK WUDGLQJ SDUWQHUV in the United States and nd that ARCH
WKH 5((5 KDG IDOOHQ EDFN WR D OHYHO ODVW measure is signi cant. In addition, Seabra
prevailing before the Asian nancial crisis. WHVWV VHYHUDO PHDVXUHV RI H[FKDQJH UDWH
)URP WKH 91' ZDV KROGLQJ UHODWLYHO XQFHUWDLQW WKDW DUH SUHYDOHQW LQ WKH OLWHUDWXUH
VWHDG DJDLQVW WKH 86' DQG WKH 86' LQ KLV VWXG RI SXUFKDVLQJ SRZHU SDULW LQ
LWVHOI ZDV ZHDNHQLQJ DJDLQVW PDQ RWKHU /DWLQ $PHULFDQ FRXQWULHV +H FRQFOXGHV WKDW
FXUUHQFLHV WKHQ 91' GHSUHFLDWHG VOLJKWO ARCH-based methods are most ef cient.
in nominal terms. However, its high in ation 7KLV PHDVXUH LV DOVR XVHG LQ PRVW RI SUHYLRXV
UDWH ZDV PXFK PRUH WKDQ RIIVHW WKLV QRPLQDO VWXGLHV 7KHUHIRUH WR GHWHUPLQH WKH YRODWLOLW
GHSUHFLDWLRQ WHQGHQF VR WKDW WKH 91' RI UHDO H[FKDQJH UDWH ZH XVH WKH FRQGLWLRQDO
WHQGHG WR DSSUHFLDWH LQ UHDO WHUPV 5((5 DW YDULDQFH RI WKH ORJ RI UHDO H[FKDQJH UDWH
WKH HQG RI ZDV DERXW SHU FHQW KLJKHU 7KH FRQGLWLRQDO YDULDQFH K RI WKH ORJ RI
WKDQ DW WKH EHJLQQLQJ RI UHDO H[FKDQJH UDWH LV XVHG WR PHDVXUH WKH ULVN
)LJXUH 5((5 1RPLQDO DQG 5HDO ( FKDQJH XQFHUWDLQW DVVRFLDWHG ZLWK WKH UHDO HIIHFWLYH
5DWH 86 91 H[FKDQJH UDWH RU ELODWHUDO UHDO H[FKDQJH UDWH
ZLWK 86 GROODU 9RODWLOLW YDULDEOHV DUH GHULYHG
IURP D *$5&+ PRGHO ZLWK VSHFLI LQJ D
*$5&+ PRGHO EHFDXVH LW LV WKH PRVW
UREXVW RI WKH IDPLO RI YRODWLOLW PRGHOV
(QJOH
7DEOH ) 7HVWV IRU 6WDWLRQDU RI 5((5
/HYHO )LUVW GLIIHUHQFH
9DULDEOH FRQVWDQW FRQVWDQW
FRQVWDQW FRQVWDQW
WUHQ WUHQ
5 5 5 5 86 91 5((5
Note: At the 5% level of signi cance, the critical value for each
7KH PHWKRGV RI PHDVXULQJ YRODWLOLW KDYH WHVW ZLWK FRQVWDQW LV DQG ZLWK FRQVWDQW WUHQ LV
1XPEHU RI ODJV LV VKRZQ LQ SDUHQWKHVHV
evolved over time to re ect new advances in
HFRQRPHWULF WHFKQLTXHV 1RQHWKHOHVV D FOHDUO 7DEOH VKRZV WKDW 5((5 VHULHV KDYH , WKHQ
GRPLQDQW DSSUR[LPDWLRQ IRU XQFHUWDLQW KDV QRW themodelfor themean ofeach seriesis speci ed
HW HPHUJHG DKPDQL 2VNRRHH DQG +HJHUW ZLWK DQ $5,0$ PRGHO /DJUDQJH 0XOWLSOLHU
7KH PRVW FRPPRQ LV VRPH PHDVXUHV WHVW LV XVHG WR WHVW IRU KHWHURVNHGDVWLFLW LQ
RI YDULDQFH EXW WKH H[DFW FRQVWUXFWLRQ RI WKLV the errors of mean equations and to nd the
PHDVXUH GLIIHUV IURP VWXG WR VWXG $PXHGR best tting autoregressive model to measure
'RUDQWHV DQG 3R]R D DUJXH WKDW WKH WKH IRUHFDVW RI XQFHUWDLQW 7KH PRGHOV IRU WKH
$5&+ DSSURDFK DOORZV XV WR EHWWHU FDSWXUH PHDRUWHG LQ 7DEOH

1R ( ( 1 / ( 2120 6 ( (
 56 5 21 2120 1 17 5 7 21

7DEOH 0HDQ DQG 9DULDQFH (TXDWLRQV IRU 5((5 0RYHPHQWV

5((5W − 5((5W − = + 5((5W − − 5((5W − + HW
H H
KW = + HW − + KW −
H
4 4
4 4 /0
Estimation Period: 1990M1 – 2009M3
Note: REERrepresents real effective exchange rate. htrepresents the conditional variance, the variance of the error term
IRU WKH PHDQ HTXDWLRQ DERYH LW 4 DQG 4 represent the Ljung–Box statistic for up to twelve and twenty four orders
DXWRFRUUHODWLRQ LQ WKH UHVLGXDOV UHVSHFWLYHO 4 DQG 4 represent the Ljung–Box statistic for up to twelve and twenty
four orders autocorrelation in the squared residuals, respectively.LM(1) represents Engle’s statistic for an additional ARCH term
LQ WKH FRQGLWLRQDO YDULDQFH HTXDWLRQ

)LJXUH GHVFULEHV WKH 5((5 YRODWLOLW LQ and to +/-3% in November) that has never
9LHWQDP IRU HDUV HIRUH WKH KDSSHQHG EHIRUH 7KHVH DGMXVWPHQW LQFUHDVHG
H[FKDQJH UDWH YRODWLOLW LV YHU KLJK DV 9LHWQDP exchange rate uctuation.
adopted a oating exchange rate regime aiming )LJXUH 9RODWLOLW RI 9LHWQDPHVH 5((5
WR GHYDOXH WKH 9LHWQDPHVH FXUUHQF WKDW ZDV -DQXDU 0DUFK
RYHUYDOXHG LQ SUHYLRXV HDUV EHFDXVH RI WKH
H[LVWHQFH RI WKH FHQWUDOO SODQQHG HFRQRP
IRU D ORQJ WLPH $IWHU WKDW LW FDQ EH VHHQ WKDW
9LHWQDP KDV XVHG WKH 86' DV D NH QRPLQDO
DQFKRU WKHUHIRUH WKH YRODWLOLW RI UHDO H[FKDQJH
UDWH LV YHU VPDOO ,W RQO LQFUHDVHV LI WKHUH LV DQ
DGMXVWPHQW LQ QRPLQDO H[FKDQJH UDWH RU WUDGLQJ
EDQG $V GLVFXVVHG DERYH WKH 6WDWH DQN RI
9LHWQDP 6 9 KDG WR GHYDOXH WKUHH WLPHV WKH 5 5 9ROD LOL

9LHWQDPHVH 'RQJ E SHU FHQW DQG ZLGHQ

trading band from +/- 1% to +/-10% because (03,5,& / ,19(67, 7,216
of the Asian nancial crisis 1997-1998. These (FR RPHW LF 0RGHO D G DWD HVF LSWLR V
DFWLRQV LQFUHDVHG WKH UHDO HIIHFWLYH H[FKDQJH
(PSLULFDO VWXGLHV VKRZ WKDW PRVW RI WKH WLPH
UDWH YRODWLOLW $IWHU LQWURGXFLQJ D QHZ SULQFLSOH
VHULHV DUH QRW VWDWLRQDU 6LQFH IDFLQJ D VSXULRXV
IRU VHWWLQJ WKH H[FKDQJH UDWH WKH YRODWLOLW
UHJUHVVLRQ SUREOHP DPRQJ WKHVH VHULHV ZKLFK
ZDV YHU VPDOO H[FHSWLRQ RI WKH YRODWLOLW LQ
ZKHQ 8QLWHG 6WDWH LPSOHPHQWHG D ZHDN LQFOXGH D XQLW URRW VRPH PHWKRGV DUH VXJJHVWHG
GROODU SROLF DQG WUDGLQJ EDQG RI 86' 91' WR VROYH WKLV SUREOHP 2QH RI WKHP LV WDNLQJ WKH
H[FKDQJH UDWH ZDV ZLGHQHG ,W VHHPV WKDW WKHUH GLIIHUHQFHV RI WKH VHULHV DQG WKHQ SXWWLQJ WKHP
LV OLWWOH YRODWLOLW RI H[FKDQJH UDWH IURP WR LQWR UHJUHVVLRQV 7KLV PHWKRG OHDGV WR WKH ORVV
2007 as the of cial exchange rate was almost RI LQIRUPDWLRQ WKDW LV LPSRUWDQW IRU WKH OHYHO
XQFKDQJHG ,Q 6 9 LQFUHDVHG WUDGLQJ relationships. As long as the rst differences of
band four times (from +/- 0.5% to +/-0.75% WKH YDULDEOHV DUH XVHG GHWHUPLQLQJ D SRWHQWLDO
in January, +/- 1% in March, +/-2% in June, ORQJ UXQ UHODWLRQVKLS EHWZHHQ WKHVH YDULDEOHV

( ( 1 / ( 2120 6 ( ( 1R
 56 5 21 2120 1 17 5 7 21

EHFRPHV LPSRVVLEOH 7KLV LV WKH SRLQW RI RULJLQ 0RWDOOHEL DVVDP

RI FR LQWHJUDWLRQ DQDO VLV 3HVDUDQ HW DO VXJJHVW WKHLU PHWKRG
7KH FR LQWHJUDWLRQ DSSURDFK GHYHORSHG E EDVHG RQ $XWRUHJUHVVLYH 'LVWULEXWHG /DJ
(QJOH DQG *UDQJHU RYHUFDPH WKLV $5'/ DSSURDFK $5'/ PRGHO LV FKDQJHG WR
SUREOHP $FFRUGLQJ WR WKLV DSSURDFK WLPH error correction model like below:
VHULHV ZKLFK DUH QRW VWDWLRQDU DW OHYHOV EXW N
∆ W = F + FW +π + ∑π L
stationary in the rst difference can be modeled W−
L=
L W−

ZLWK WKHLU OHYHO VWDWHV ,Q WKLV ZD ORVV RI S− N S−

+ ∑λM∆ + ∑∑ γ WL− M ∆ L
+
LQIRUPDWLRQ LQ WKH ORQJ UXQ FDQ EH SUHYHQWHG M=
W− M
L= M=
W− M W

+RZHYHU WKLV DSSURDFK EHFRPHV LQYDOLG LI ZKHUH N LV WKH QXPEHU RI LQGHSHQGHQW YDULDEOHV
WKHUH DUH PRUH WKDQ RQH FR LQWHJUDWLRQ YHFWRUV
WKH GLVWXUEDQFHV W DUH VHULDOO XQFRUUHODWHG
0RYLQJ IURP WKLV SRLQW ZLWK WKH KHOS RI WKH
DSSURDFK GHYHORSHG E -RKDQVHQ LW 7KH $5'/ DSSURDFK XVHV WZR VWHSV WR HVWLPDWH
LV SRVVLEOH WR WHVW KRZ PDQ FR LQWHJUDWLRQ WKH OHYHO UHODWLRQVKLS )LUVW VWHS LV WR GHWHUPLQH
YHFWRUV WKHUH DUH DPRQJ WKH YDULDEOHV E XVLQJ ZKHWKHU D OHYHO UHODWLRQVKLS H[LVW EHWZHHQ WKH
WKH 9$5 PRGHO LQ ZKLFK DOO WKH YDULDEOHV DUH YDULDEOHV LQ HTXDWLRQV 7KH QXOO K SRWKHVLV
DFFHSWHG DV HQGRJHQRXV 7KHUHIRUH XQOLNH RI QR OHYHO UHODWLRQVKLS DPRQJ YDULDEOHV LV
WKH (QJOH *UDQJHU PHWKRG D PRUH UHDOLVWLF WHVWHG L H + : π = π L = L = N E XVLQJ
H[DPLQDWLRQ LV SURYLGHG ZLWKRXW OLPLWLQJ WKH the F-test for the joint signi cant of the
WHVW LQ RQH FR LQWHJUDWLRQ YHFWRU H[SHFWDWLRQ lagged levels coef cient. Two sets of critical
+RZHYHU LQ RUGHU WR SHUIRUP WKHVH WHVWV YDOXHV DUH JHQHUDWHG 2QH VHW UHIHUV WR ,
GHYHORSHG E (QJOH DQG *UDQJHU VHULHV DQG WKH RWKHU IRU , VHULHV +HUH WKH
-RKDQVHQ DQG -RKDQVHQ DQG -XVHOLXV FULWLFDO YDOXHV IRU , VHULHV DUH UHIHUUHG WR
WKH FRQGLWLRQ PXVW EH PHW WKDW DOO DV WKH XSSHU ERXQG FULWLFDO YDOXHV ZKLOH WKH
VHULHV VKRXOG QRW EH VWDWLRQDU DW WKH OHYHOV FULWLFDO YDOXHV IRU , VHULHV DUH UHIHUUHG
DQG WKH VKRXOG EHFRPH VWDWLRQDU ZKHQ WKH WR DV WKH ORZHU ERXQG FULWLFDO YDOXHV ,I WKH
VDPH GLIIHUHQFHV DUH WDNHQ ,I RQH RU PRUH RI HVWLPDWHG ) VWDWLVWLFV LV JUHDWHU WKDQ WKH XSSHU
WKH VHULHV DUH VWDWLRQDU DW OHYHOV WKDW LV WR VD ERXQG FULWLFDO YDOXHV ZH FRQFOXGH WKDW WKH
, WKH FR LQWHJUDWLRQ UHODWLRQVKLS FDQQRW YDULDEOHV LQ TXHVWLRQ DUH FRLQWHJUDWHG $OVR
EH H[DPLQHG ZLWK WKHVH WHVWV RXQGV WHVW LI WKH HVWLPDWHG ) VWDWLVWLFV IDOOV EHWZHHQ WKH
DSSURDFK GHYHORSHG E 3HVDUDQ HW DO ORZHU DQG WKH XSSHU ERXQG FULWLFDO YDOXHV
UHPRYHV WKLV SUREOHP $FFRUGLQJ WR WKHLU WKH GHFLVLRQ DERXW FRLQWHJUDWLRQ DPRQJ WKH
DSSURDFK WKH H[LVWHQFH RI D FR LQWHJUDWLRQ YDULDEOHV LQYROYHG LV LQFRQFOXVLYH $QG LI
UHODWLRQVKLS FDQ EH H[DPLQHG EHWZHHQ WKH WKH HVWLPDWHG ) VWDWLVWLFV LV OHVV WKDQ WKH
VHULHV UHJDUGOHVV RI ZKHWKHU WKH DUH , RU ORZHU FULWLFDO YDOXHV WKH QXOO K SRWKHVLV RI
, XQGHU WKH FLUFXPVWDQFH WKDW GHSHQGHQW QR FRLQWHJUDWLRQ FDQQRW EH UHMHFWHG 6HFRQG
YDULDEOH LV , DQG WKH LQGHSHQGHQW YDULDEOHV VWHS LV WR HVWLPDWH WKH ORQJ UXQ DQG WKH VKRUW
DUH HLWKHU , RU , 7KLV SRLQW LV WKH JUHDWHVW run coef cients by using the ARDL approach
DGYDQWDJH RI WKH ERXQGV WHVW DPRQJ DOO WKH LI WKH ORQJ UXQ UHODWLRQVKLS LV HVWDEOLVKHG
FR LQWHJUDWLRQ WHVWV 0RUHRYHU WKLV DSSURDFK EHWZHHQ WKH YDULDEOHV 7KH ODJ RUGHUV RI WKH
FDQ GLVWLQJXLVK GHSHQGHQW DQG LQGHSHQGHQW YDULDEOHV DUH FKRVHQ XVLQJ VWDWLVWLF FULWHULD
YDULDEOHV DQG LV PRUH VXLWDEOH WKDQ DQRWKHU H J $NDLNH ,QIRUPDWLRQ &ULWHULD $,&
PHWKRG IRU VPDOO VDPSOH VL]H *KRUEDQL DQG 6FKZDU] ,QIRUPDWLRQ &ULWHULD 6,&

1R ( ( 1 / ( 2120 6 ( (
 56 5 21 2120 1 17 5 7 21

,Q WKLV VWXG ZH LQYHVWLJDWH WKH LPSDFW RI 5HV OWV D G LVF VVLR V

H[FKDQJH UDWH YRODWLOLW RQ JURZWK H[SRUW DQG QLW URRW WHVW
)', SHUIRUPDQFHV XVLQJ ERXQGV WHVW DSSURDFK
HIRUH FRQVWUXFWLQJ RXU PRGHOV ZH H[DPLQH
GHYHORSHG E 3HVDUDQ HW DO 6LQFH ZH
WKH VHULHV LQ RUGHU WR GHWHUPLQH ZKHWKHU WKH
DUH PRVW LQWHUHVWHG LQ WKH LPSDFW RI WKH VKRFNV
are stationary or not. Augmented Dickey–Fuller
WR WKH UHDO H[FKDQJH UDWH DQG WR UHDO H[FKDQJH
$') XQLW URRW WHVWV DUH FRQGXFWHG LQFOXGLQJ D
UDWH YRODWLOLW RQ RXWSXW RI 9LHWQDP WKHQ ZH GULIW WHUP DQG ERWK ZLWK DQG ZLWKRXW D WUHQG 7DEOH
HVWLPDWH D $5'/ PRGHO ZLWK WKUHH YDULDEOHV FRQWDLQV WKH UHVXOWV IURP WKH XQLW URRW WHVWV
ZKLFK LV DSSOLHG E $PXHGR 'RUDQWHV DQG
7DEOH ) 7HVWV IRU 8QLW 5RRW
3R]R D IRU VLQJOH FRXQWU &KLOH .RUHD
0H[LFR 6LQJDSRUH ,Q DGGLWLRQ ZH LQFOXGH D /HYHO )LUVW GLIIHUHQFH
9DULDEOH FRQVWDQW FRQVWDQW
GXPP YDULDEOH 6 WR DFFRXQW IRU WKH HIIHFW FRQVWDQW FRQVWDQW
WUHQ WUHQ
RI WKH FKDQJH LQ H[FKDQJH UDWH V VWHP RQ
92/B5
HFRQRPLF SHUIRUPDQFH 7KHQ WKH HTXDWLRQV
,2
IRU HVWLPDWLQJ DUH IROORZ
1ote: At the 5% level of signi cance, the critical value for
∆,2W = D + D W + D ,2W − + D 5((5W − + D 92/ B 5W −
HDFK WHVW ZLWK FRQVWDQW LV DQG ZLWK FRQVWDQW WUHQ LV
O− O− O−
+ ∑ D M ∆,2W − M + ∑ D M ∆5((5W − M + ∑ D M ∆92/ B 5W − M 1XPEHU RI ODJV LV VKRZQ LQ SDUHQWKHVHV
M= M= M=

+D + 7KH UHVXOWV LQGLFDWH WKDW ZLWK WKH H[FHSWLRQ RI

6 W
5((5 YRODWLOLW DOO WKH VHULHV DUH QRQ VWDWLRQDU
ZKHUH ,2 LV UHDO LQGXVWULDO RXWSXW 5((5 LV DQG WKH $') WHVW UHVXOWV DUH LQYDULDQW DV WR
UHDO HIIHFWLYH H[FKDQJH UDWH 92/B5 LV UHDO ZKHWKHU WKH XQLW URRW WHVWV DUH FRQGXFWHG ZLWK
HIIHFWLYH H[FKDQJH UDWH YRODWLOLW 7KH GXPP RU ZLWKRXW D OLQHDU WUHQG 7KH QXOO K SRWKHVHV
variable is de ned by 6 = RYHU WKH SHULRG RI GLIIHUHQFH VWDWLRQDU SURFHVVHV IRU HDFK
0DUFK 'HFHPEHU HOVHZKHUH RI WKH QRQVWDWLRQDU VHULHV DUH DFFHSWHG 8QLW
URRW WHVWV VKRZ WKDW GHSHQGHQW YDULDEOHV DUH
'DWD RI ,2 LV FROOHFWHG IURP *HQHUDO 6WDWLVWLFV
, DQG LQGHSHQGHQW YDULDEOHV DUH PL[WXUH RI
Of ce of Vietnam (GSO). The others are
, DQG , 7KHQ $5'/ DSSURDFK LV PRUH
FDOFXODWHG IURP GDWD RI ,0) ,)6 DQG '27 VXLWDEOH WKDQ RWKHU DSSURDFKHV IRU LQYHVWLJDWLQJ
HFDXVH GDWD RI UHDO LQGXVWULDO RXWSXW LV UHODWLRQVKLSV LQ OHYHO RI YDULDEOHV
DYDLODEOH IURP -DQXDU WKHQ WKH RXWSXW
%R Q V 7HVWLQJ IRU /H HO 5HODWLRQVKLSV
HTXDWLRQ LV HVWLPDWHG XVLQJ PRQWKO GDWD IURP
,W LV LPSRUWDQW WKDW WKH ODJ RUGHU RI WKH
-DQXDU WR 0DUFK
XQGHUO LQJ 9$5 LV VHOHFWHG DSSURSULDWHO
$OO GDWD DUH VHDVRQDOO DGMXVWHG 3ULRU WR 7KHUH LV D GHOLFDWH EDODQFH EHWZHHQ FKRRVLQJ
HVWLPDWLRQ DOO WKH YDULDEOHV DUH WUDQVIRUPHG ODJ order suf cienWO ODUJH WR PLWLJDWH WKH
LQWR QDWXUDO ORJDULWKPV UHVLGXDO VHULDO FRUUHODWLRQ SUREOHP DQG DW

Before February 1999, SBV set the of cial exchange rate. After that, the government introduced a new principle for setting
the exchange rate.An average inter-bank rate of exchange between VND against USD would be of cial exchange rate in the
IROORZLQJ GD 7KLV QHZ DUUDQJHPHQW KDV EHHQ YLHZHG DV D WXUQLQJ SRLQW LQ WKH H[FKDQJH UDWH SROLF LQ 9LHWQDP DLPLQJ DW
gradually moving toward greater exchange rate exibility and giving more weight to market forces in exchange rate determi
QDWLRQ +RZHYHU IURP WKH JRYHUQRU RI WKH 6 9 ZDV TXRWHG DV SXEOLFO VWDWLQJ WKDW WKH 91'ZRXOG QRW EH DOORZHG WR
depreciate by more than1 percent per annum. As it turned out, the of cial USD/VND ratewas again almost unchanged from
-DQXDU WR'HFHPEHU

( ( 1 / ( 2120 6 ( ( 1R
 56 5 21 2120 1 17 5 7 21

the same time, suf ciently small so that 7DEOH JLYHV WKH YDOXHV RI WKH ) VWDWLVWLFV IRU
WKH FRQGLWLRQDO (&0V DUH QRW XQGXO RYHU WHVW WKH H[LVWHQFH RI OHYHO UHODWLRQVKLSV XQGHU
SDUDPHWHUL]HG 3DUWLFXODUO LQ YLHZ RI WKH WZR GLIIHUHQW VFHQDULRV 7KH FULWLFDO YDOXH
OLPLWHG WLPH VHULHV GDWD ZKLFK DUH DYDLODEOH ERXQGV RI WKH VWDWLVWLFV ),9 DQG )9 DUH JLYHQ LQ
D GHFLVLRQ PXVW EH PDGH FRQFHUQLQJ WKH 3HVDUDQ HW DO DQG GHSHQG RQ WKH QXPEHU
WLPH WUHQG LQ HDFK HTXDWLRQ DQG ZKHWKHU LWV RI LQGHSHQGHQW YDULDEOHV LQ UHJUHVVLRQV
coef cient should be restricted.
7DEOH ) VWDWLVWLFV IRU 7HVWLQJ WKH
7R GHWHUPLQH WKH DSSURSULDWH ODJ OHQJWK O P
( LVWHQFH RI D /HYHOV
Q DQG ZKHWKHU D GHWHUPLQLVWLF OLQHDU WUHQG LV
UHTXLUHG LQ DGGLWLRQ WR WKH HTXDWLRQ ZH HVWLPDWH O ), ) W
WKH FRQGLWLRQDO PRGHO E OHDVW VTXDUH
ZLWK DQG ZLWKRXW D OLQHDU WLPH WUHQG HFDXVH
coef cients of trend are always signi cant in
WKH UHJUHVVLRQV RI RXWSXW WKHQ ZH RQO FRQVLGHU
ZLWK FDVH RI UHJUHVVLRQV ZLWK GHWHUPLQLVWLF
Note: F,9 LV WKH ) VWDWLVWLF IRU WHVWLQJ D D D LQ
trend. Tables 4 give Akaike’s and Schwarz’s )9 LV WKH ) VWDWLVWLF IRU WHVWLQJLQ D D D W9 LV WKH
,QIRUPDWLRQ &ULWHULD GHQRWHG UHVSHFWLYHO E W VWDWLVWLF IRU WHVWLQJ D = 0 in (3). The 1% critical value bounds
$,& DQG 6,& DQG /DJUDQJH PXOWLSOLHU /0 DUH DQG IRU ),9 DQG )9 UHVSHFWLYHO 7KH
VWDWLVWLFV χ 6& IRU WHVWLQJ WKH K SRWKHVLV RI QR 1% critical value for t9 LV is signi cant of 1%.
UHVLGXDO VHULDO FRUUHODWLRQ DJDLQVW RUGHUV DQG 7KH UHVXOW RI ) VWDWLVWLFV LQGLFDWHV WKDW DW
IRU HVWLPDWLQJ HTXDWLRQV PRQWKO GDWD
signi cance level, null hypothesis of no level
7KH ODJ RUGHUV VHOHFWHG E $,& DQG 6,& DUH UHODWLRQVKLS LV UHMHFWHG IRU DOO FDVHV RI VHOHFWHG
DQG UHVSHFWLYHO ZKLOH WKH χ 6& VWDWLVWLFV ODJ RUGHUV WKH UHVXOWV IURP WKH DSSOLFDWLRQ RI
suggest using a relatively high lag order: 4 or WKH ERXQG W WHVW VXSSRUW WKH UHVXOWV RI ) WHVWV
,Q YLHZ RI WKH LPSRUWDQFH RI DVVXPSWLRQ RI
,Q RWKHU ZRUGV UHVXOWV RI ERXQGV WHVWV VXSSRUW
VHULDOO XQFRUUHODWHG HUURUV IRU WKH YDOLGLW RI
WKH H[LVWHQFH RI OHYHO UHODWLRQVKLSV LQ WKH
WKH ERXQGV WHVW LW VHHPV SUXGHQW WR VHOHFW ODJ
HVWLPDWLQJ HTXDWLRQV
WR EH HLWKHU RU +RZHYHU IRU FRPSOHWHQHVV
ZH UHSRUW WHVW UHVXOWV IRU O DQG /H HO 5HODWLRQVKLSV
)LUVW WKH RUGHUV RI $5'/ PRGHOV ZHUH VHOHFWHG
7DEOH 6WDWLVWLFV IRU 6HOHFWLQJ WKH /DJ
E VHDUFKLQJ DFURVV WKH $5'/ PRGHOV
2UGHU
7KH HVWLPDWH UHVXOWV LQ WKH FKRLFH RI$5'/
O ,& 6,& χ 6& ( ) χ 6& ( ) IRU WKH HVWLPDWLQJ HTXDWLRQ QRUPDOL]LQJ
RQ RXWSXW IURP WKH VHOHFWHG $5'/ PRGHOV WKH
HPSLULFDO UHVXOWV RI WKH UHODWLRQVKLSV LQ OHYHO
DUH SUHVHQWHG LQ HTXDWLRQ
,2W = − 92/ B 5W − 5((5W + + εW

ZKHUH ε W LV WKH HTXLOLEULXP FRUUHFWLRQ WHUP DQG

1RWH AIC = – 2O 7 N 7 and SIC = – 2O 7 NORJ7 7 ZKHUH WKH VWDQGDUG HUURUV DUH LQ SDUHQWKHVLV
O LV WKH PD[LPL]HG ORJ OLNHOLKRRG YDOXH 7 LV WKH QXPEHU RI
REVHUYDWLRQV DQG N is number of freely estimated coef cients. The level estimate is highly signi cant and
3 YDOXHV DUH LQ EUDFNHWV KDV WKH H[SHFWHG VLJQ 7KH UHVXOW VKRZV WKDW

1R ( ( 1 / ( 2120 6 ( (
 56 5 21 2120 1 17 5 7 21

exchange rate volatility affects signi cantly 7KLV LV HYLGHQFH RI FRLQWHJUDWLRQ UHODWLRQVKLSV
QHJDWLYH RQ 9LHWQDP HFRQRPLF SHUIRUPDQFH DPRQJ YDULDEOHV LQ WKH PRGHO 7KH HVWLPDWH
$ ULVH LQ H[FKDQJH UDWH YRODWLOLW GHFUHDVH YDOXH RI HUURU FRUUHFWLRQ WHUPV LPSO WKDW
RXWSXW 9LHWQDP LV VPDOO RSHQ HFRQRP WKH WKH VSHHG RI DGMXVWPHQW WR WKH ORQJ UXQ
LQWHUQDWLRQDO SULFH OHYHO LV JLYHQ DQG WUDGHG HTXLOLEULXP LQ UHVSRQVH WR WKH GLVHTXLOLEULXP
JRRGV PDNH XS D KLJK VKDUH RI WKH GRPHVWLFDOO FDXVHG E VKRUW UXQ VKRFNV RI WKH SUHYLRXV
FRQVXPHG JRRGV H[FKDQJH UDWH VWDELOLW SHULRG LV SHU FHQW 7KLV QXPEHUV LPSOLHV
HQVXUHV GRPHVWLF SULFH VWDELOLW 7KH ZHOIDUH WKDW WKH DGMXVWPHQW RI WKH RXWSXW WR FKDQJHV
HIIHFW RI VWDEOH H[FKDQJH UDWHV RULJLQDWHV LQ LQ WKH LQGHSHQGHQW YDULDEOHV LV PRGHUDWH DQG
PDFURHFRQRPLF VWDELOLW ZKLFK SURYLGHV D PD WDNH DERXW PRUH WKDQ PRQWKV
IDYRUDEOH HQYLURQPHQW IRU LQYHVWPHQW DQG 7DEOH 7KH (&0 IRU WKH 6HOHFWHG 5 /
consumption. In addition, nancial market in 0RGHO
9LHWQDP LV XQGHUGHYHORSHG WKHQ H[FKDQJH UDWH 5HJUHVVRU 5/
YRODWLOLW LV DVVRFLDWHG ZLWK KLJKHU WUDQVDFWLRQ ΔLIO(-1)
FRVW EHFDXVH XQFHUWDLQW DQG KHGJLQJ ΔVOL_R
H[FKDQJH UDWH ULVN LV FRVWO )URP D VKRUW WHUP ΔREER
SHUVSHFWLYH VWDELOL]DWLRQ LQ H[FKDQJH UDWH FDQ F
foster economic growth by a more ef cient WUHQ

LQWHUQDWLRQDO DOORFDWLRQ RI FDSLWDO ZKHQ 6

(&B,2
transaction costs for capital ows are removed
$GM 5 VTXDUHG
0F.LQQRQ 0RUHRYHU 9LHWQDP LV D
6HULDO &RUUHODWLRQ )
GROODUL]HG HFRQRP IRU WKH SUHLRG RI
(&B ,2 ,2 92/B5 5((5
WKH GHJUHH RI GROODUL]DWLRQ LQ 9LHWQDP LV Note: *, **, and *** are respectively signi cant of 10%, 5%,
DOZD V DERYH SHU FHQW LQ FRPSDULVRQ ZLWK and 1%. The numbers in parentheses are standard errors. The
SHU FHQW LQ RWKHU FRXQWULHV LQ 6RXWKHDVW QXPEHU LQ EUDFNHW LV S YDOXH
$VLD VXFK DV 7KDLODQG 0DOD VLD DQG ,QGRQHVLD &21&/86,21
due to massive ow of remittance and foreign
7KLV VWXG LQYHVWLJDWHV WKH LPSDFW RI H[FKDQJH
LQYHVWPHQW DQG LQFUHDVHG H[SRUW HDUQLQJV RYHU
UDWH YRODWLOLW RQ HFRQRPLF JURZWK LQ 9LHWQDP
the past years (Son, 2009). Then uctuations
IRU SHULRG -DQXDU 0DUFK H
LQ H[FKDQJH UDWH OHYHO FRQVWLWXWH D ULVN IRU
XVH D PXOWLODWHUDO UHDO H[FKDQJH UDWH DQG WKH
JURZWK DV WKH DIIHFW WKH EDODQFH VKHHWV RI
*$5&+ PRGHO WR PHDVXUH H[FKDQJH
EDQNV DQG HQWHUSULVHV ZKHUH IRUHLJQ GHEW
UDWH YRODWLOLW 7KLV PHDVXUHPHQW LV VXSSRVHG
WHQGV WR EH GHQRPLQDWHG LQ IRUHLJQ FXUUHQF
DV WKH EHVW YRODWLOLW SUR[ WR UHFRQFLOH
(LFKHQJUHHQ DQG +DXVPDQQ 7KH OHYHO con icting views of costs and advantages
UHODWLRQVKLS DOVR LQGLFDWHV WKDW DQ DSSUHFLDWLRQ of exible exchange rate system. This study
ZLOO GHFUHDVH HFRQRPLF JURZWK DOVR DSSOLHV D DSSURDFK FDOOHG $5'/ WR WKH
7DEOH SURYLGHV WKH UHVXOW RI WKH HUURU SUREOHP RI WHVWLQJ WKH H[LVWHQFH RI D OHYHO
FRUUHFWLRQ UHSUHVHQWDWLRQ RI HVWLPDWHG $5'/ UHODWLRQVKLS EHWZHHQ D GHSHQGHQW YDULDEOH
PRGHO 7KH UHVXOW LQGLFDWHV WKDW WKH HUURU DQG D VHW RI UHJUHVVRUV ZKHQ UHJUHVVRUV DUH D
FRUUHFWLRQ WHUP (&B,2 KDYH WKH ULJKW PL[WXUH RI , DQG , EHFDXVH PHDVXUH RI
sign (negative) and is statistically signi cant. H[FKDQJH UDWH YRODWLOLW LV VWDWLRQDU ZKLOH

( ( 1 / ( 2120 6 ( ( 1R
 56 5 21 2120 1 17 5 7 21

RWKHU YDULDEOHV LQ PRGHOV DUH QRQ VWDWLRQDU based on the three benchmarks: demand and
7KH HPSLULFDO UHVXOW VKRZV WKDW WKHUH H[LWV VXSSO RI WKH 9LHWQDP GRQJ WKH H[FKDQJH
RI OHYHO UHODWLRQVKLS LQ RXWSXW PRGHO ZKLFK UDWHV IRU D EDVNHW RI HLJKW VWURQJ IRUHLJQ
DFFRXQWV IRU LPSDFW RI UHDO H[FKDQJH UDWH FXUUHQFLHV DQG IRU DQ FKDQJH WR EDODQFH
PDFUR HFRQRPLF QHHGV HVLGHV DQQRXQFLQJ
DQG UHDO H[FKDQJH UDWH YRODWLOLW (FRQRPLF
WKH FHQWUDO UDWH GDLO HDFK ZHHN WKH 6 9 ZLOO
SHUIRUPDQFHV ZLOO EH LQMXUHG E H[FKDQJH
FDOFXODWH WKH FURVV H[FKDQJH UDWH EHWZHHQ WKH
UDWH YRODWLOLW $Q LQFUHDVH LQ H[FKDQJH UDWH
91' DQG WKH HLJKW RWKHU FXUUHQFLHV LQFOXGLQJ
volatility will reduce signi cantly economic
WKH 86' WKH (XUR WKH XDQ WKH -DSDQHVH
JURZWK 7KHUHIRUH D VWDEOH RI H[FKDQJH UDWH
HQ WKH 6LQJDSRUH 'ROODU WKH 6RXWK .RUHDQ
V VWHP ZLWK PDLQ WUDGLQJ SDUWQHUV LV QHHGHG
RQ WKH 7KDL DKW DQG WKH 7DLZDQ 'ROODU
WR HQKDQFH HFRQRPLF JURZWK
7KLV FKDQJH LV WKH ULJKW GLUHFWLRQ WR HQKDQFH
)URP WKH EHJLQQLQJ RI WKH 6 9 HFRQRPLF JURZWK 7KH QHZ UHJLPH ZLOO
DQQRXQFHV D FHQWUDO H[FKDQJH UDWH HYHU GD IDFLOLWDWH VWURQJHU SHUIRUPDQFH LQ WKH IRUHLJQ
IRU WKH 91' 86' ZKLFK ZRXOG EH XVHG E FXUUHQF GHULYDWLYHV PDUNHW PHHWLQJ WKH
nancial institutions authorized to trade in UHTXLUHPHQWV IRU ULVN SUHYHQWLRQ LQ H[FKDQJH
IRUHLJQ FXUUHQFLHV 7KH UDWH LV FDOFXODWHG UDWHV DQG LQFUHDVH OLTXLGLW LQ WKH PDUNHW q

5HIHUHQFHV
$PXHGR 'RUDQWHV & DQG 6 3R]R D 8QFHUWDLQW DQG (FRQRPLF 3HUIRUPDQFHV ´
5HYLHZ RI 'HYHORSPHQW (FRQRPLFV SS
$PXHGR 'RUDQWHV & DQG 6 3R]R E )RUHLJQ ([FKDQJH 5DWHV DQG )RUHLJQ 'LUHFW
,QYHVWPHQW LQ WKH 8QLWHG 6WDWHG ´ 7KH ,QWHUQDWLRQDO 7UDGH -RXUQDO
$]LG 7 0 -DPLO DQG $ .RXVDU ,PSDFW RI ([FKDQJH 5DWH 9RODWLOLW RQ *URZWK DQG
Economic Performance: A Case Study of Pakistan, 1973–2003,” The Pakistan Development
Review 44 (4), Part II (Winter 2005), pp. 749–775.
DJHOOD 0 / HFFKHWWL , +DVDQ 5HDO (IIHFWLYH ([FKDQJH UDWH 9RODWLOLW DQG
Growth: A Framework to Measure Advantages of Flexibility vs. Costs of Volatility “,
Journal of Banking & Finance 30, pp. 1149–1169.
DKPDQL 2VNRRHH 0 DQG 6 +HJHUW ([FKDQJH 5DWH 9RODWLOLW DQG 7UDGH
Flows: a Review Article,” Journal of Economic Studies 34 (3), pp. 211-255.
Bassam AbuAl-Foul (2010), “The Causal Relation between Savings and Economic Growth:
6RPH (YLGHQFH IURP 0(1$ &RXQWULHV ´ 7KH WK 0(($ 0HHWLQJ $WODQWD -DQ
Available at: http://www.meeaweb.org/meeapapers/EconomicGrowth.pdf
&DSRUDOH * 0 DQG 1 3LWWLV 1RPLQDO ([FKDQJH 5DWH 5HJLPHV DQG WKH 6WRFKDVWLF
HKDYLRU RI 5HDO 9DULDEOHV ´ -RXUQDO RI ,QWHUQDWLRQDO 0RQH DQG )LQDQFH SS
'H *UDXZH 3 DQG * 6FKQDEO ([FKDQJH 5DWH 5HJLPHV DQG 0DFURHFRQRPLF
6WDELOLW LQ &HQWUDO DQG (DVWHUQ (XURSH ´ &HVLIR RUNLQJ 3DSHU 1R

1R ( ( 1 / ( 2120 6 ( (
56 5 21 2120 1 17 5 7 21

(GZDUGV 6HEDVWLDQ /HY H DWL (GXDUGR )OH[LEOH ([FKDQJH 5DWHV DV 6KRFN

$EVRUEHUV ´ (XURSHDQ (FRQRPLF 5HYLHZ SS
Eichengreen, B. and D. Leblang (2003), “Exchange Rates and Cohesion: Historical
3HUVSHFWLYHV DQG 3ROLWLFDO (FRQRP &RQVLGHUDWLRQV ´ -RXUQDO RI &RPPRQ 0DUNHW 6WXGLHV
41, pp. 797–822.
(LFKHQJUHHQ DQG 5 +DXVPDQQ ([FKDQJH 5DWHV )LQDQFLDO )UDJLOLW ´ 1 (5
RUNLQJ3DSHU 1R
Engle, R. (2001), “GARCH 101: The Use of ARCH/GARCH Models in Applied
(FRQRPHWULFV ´ -RXUQDO RI (FRQRPLF 3HUVSHFWLYH SS
Engle, R. and C.W.J Granger (1987), “Cointegration and Error Correction: representation,
HVWLPDWLQ DQG WHVWLQJ´ (FRQRPHWULFD SS
*KRUEDQL 0 DQG 0 0RWDOOHEL $SSOLFDWLRQ 3HVDUDQ DQG 6KLQ 0HWKRG IRU
(VWLPDWLQJ ,UDQVC ,PSRUW 'HPDQG )XQFWLRQ ´ -RXUQDO RI $SSOLHG 6FLHQFH SS
Ghosh, A., A. M. Gulde and Wolf H. (2003), Exchange Rate Regimes: Choices and
Consequences. Cambridge, Massachusetts: MIT Press.
-RKDQVHQ 6 6WDWLVWLFDO $QDO VLV RI &RLQWHJUDWLRQ 9HFWRUV´ -RXUQDO RI (FRQRPLF
' QDPLFV DQG &RQWURO SS
-RKDQVHQ 6 DQG . -XVHOLXV 0D[LPXP /LNHOLKRRG (VWLPDWLRQ DQG ,QIHUHQFH RQ
&RLQWHJUDWLRQ LWK $SSOLFDWLRQV WR WKH 'HPDQG )RU 0RQH ´ 2[IRUG XOOHWLQ RI (FRQRPLFV
DQG 6WDWLVWLFV 9RO 1R SS
+DVDQ 6 DQG 0 DOODFH 5HDO ([FKDQJH 5DWH 9RODWLOLW DQG WKH ([FKDQJH 5DWH
Regimes: Evidence from Long term Data,” Economic Letters 52, pp. 67-73.
.HQW & DQG 5 1DMD (IIHFWLYH 5HDO ([FKDQJH 5DWHV DQG ,UUHOHYDQW 1RPLQDO
([FKDQJH 5DWH 5HJLPHV ´ 5HVHDUFK 'LVFXVVLRQ 3DSHU 1R
0F.LQQRQ 5RQDOG 2SWLPXP &XUUHQF $UHDV ´ $PHULFDQ (FRQRPLF 5HYLHZ
pp. 717–725.
0F.LQQRQ 5RQDOG 0RQH DQG &DSLWDO LQ (FRQRPLF 'HYHORSPHQW URRNLQJV
,QVWLWXWLRQ
2EVWIHOG 0 DQG . 5RJRII 7KH 0LUDJH RI )L[HG ([FKDQJH 5DWHV ´ 7KH -RXUQDO RI
Economic Perspectives 9 (4), pp. 73–96.
3HVDUDQ 0 + 6KLQ DQG 5 - 6PLWK RXQGV 7HVWLQJ $SSURDFKHV WR WKH $QDO VLV
of Level Relationships”, Journal of Applied Econometrics 16, pp. 289–326.
4LDQ 9DUDQJLV 3 'RHV ([FKDQJH 5DWH 9RODWLOLW +LQGHU ([SRUW *URZWK" ´
Empirical Economics 19, pp. 371–396.
6FKQDEO * ([FKDQJH 5DWH 9RODWLOLW DQG *URZWK LQ 6PDOO 2SHQ (FRQRPLHV DW WKH
(08 3HULSKHU ´ (FRQRPLF 6 VWHPV SS
6HDEUD ) 6KRUW UXQ ([FKDQJH 5DWH 8QFHUWDLQW LQ /DWLQ $PHULFD ´ $SSOLHG
(FRQRPLFV SS
Son, N. H. (2009), Banking system of Vietnam: Reform Strategies and TransitionAssessment
http://www.trf.or.th/TRFGallery/Upload/Gallery/Documents/Files/1000000015.pdf

( ( 1 / ( 2120 6 ( ( 1R