On the asymmetric effects of exchange‐rate volatility on trade flows: Evidence from US–UK Commodity Trade
| Date | 01 February 2021 |
| Published date | 01 February 2021 |
| Author | Mohsen Bahmani‐Oskooee,Toan Luu Duc Huynh,Muhammad Ali Nasir |
| DOI | http://doi.org/10.1111/sjpe.12257 |
Scott J Polit Econ . 2021;68:51–102. wileyonlinelibrary.com/journal/sjpe
|
51
© 2020 Scottis h Economic Societ y
Accepted: 20 May 2020
DOI: 10 .1111/sjpe.1 2257
ORIGINAL ARTICLE
On the asymmetric effects of exchange-rate
volatility on trade flows: Evidence from US–UK
Commodity Trade
Mohsen Bahmani-Oskooee1 | Toan Luu Duc Huynh2,3 |
Muhammad Ali Nasir3,4
1Center for Resea rch on Internation al
Economic, Department of Economics,
University of Wisconsin-Milwaukee,
Milwaukee, WI, USA
2Chair of Behavi oral Finance, WHU - O tto
Beisheim Sch ool of Management, Val lendar,
Germany
3School of Bank ing, Universit y of Economics
Ho Chi Minh Cit y, Ho Chi Minh City,
Vietnam
4Leeds Business School, Leeds Beckett
University, Leeds, UK
Correspondence
Mohsen Bahmani-Oskooee, The Center for
Research on International Economic, The
Departm ent of Economics The Un iversity
of Wisconsin-Milwaukee, Milwaukee, WI
53201, USA.
Email: bahmani@uwm.edu
Abstract
We consider the response of ea ch of the 67 industries that
trade betwee n the United States and United Kingd om to the
volatility of the real d ollar–pound exchange rate. When we
follow previous resea rch and estimate a linear ARDL model
for each industr y, we find short-run effects of volat ility in 22
US exporting in dustries to the United Kingdom t hat last into
the long run only in nin e industries. A s for the UK expor ts
to the United States, we f ind short-run effec ts in 18 indus-
tries that last into th e long run in 15 industr ies. However,
when we estimate a nonl inear model for each i ndustry, we
find short-run effe cts of volatility o n 41 US exporting in-
dustries and on 4 3 UK exporting in dustries, all in an a sym-
metric manner. Short-run asy mmetric effec ts lasted into
long-run asymmetr ic effects in 24 US exp orting indust ries
to the United Kingdom an d in 33 UK exportin g industries
to the United States. Wh ile total trade shar es of industries
from the linear mod els were negligible, th ose of the indus-
tries from the nonl inear models were signi ficant in size, in
the tune of one-third of the tr ade.
KEYWORDS
US-UK Trade, 67 Industr ies, dollar–pound volatility, asym metry
JEL CLASSIFICATION
F14; F31
52
|
BAHMANI-OSKOOEE Et Al .
1 | INTRODUCTION
Most of the stud ies that have tested the impa ct of exchange rate uncer tainty on trade flow s in the past have con-
centrated on ind ustrial countries, mostl y due to availability of the data. E specially in the early years af ter 1973,
when the intern ational moneta ry system sh ifted from fixe d to a relatively mo re flexible excha nge rate system.
Most studies u sed data from the G7 countri es to test whether flexib le rates are detrimenta l to trade. Theoretica l
development s supported by empiric al work aim at explaining th e fact that trade could b e affected in either di rec-
tion by exchange rate u ncertainty or vol atility. Clearly, trader s who are risk averse, will t rade less to avoid any loss .
On the other han d, those who can tolerate risk , will trade more to cover any loss of inco me in the future due to
exchange rate unce rtainty.1 Indeed, emp irical studies rev iewed by Bahmani-Oskooe e and Hegerty (20 07) support
both effect s and more importa ntly, it shows that each indust rial country has its own l iterature and our countr y of
concern, the Un ited Kingdom is no exception . In what follows, we review the U K-related literature in order to se e
in which direction we are extending the literature.
We classify all UK-relate d studies and tho se that have included t he United Kingdom i n their sample of coun tries into
five categorie s. The first includes stu dies that have relied upon only cr oss-sectional data. The lis t includes Bahmani-
Oskooee and Ltai fa (1992) who included 19 developed a nd 67 developing countries i n their sample. They s howed that,
indeed, exchan ge rate volatility has an a dverse impact on re al exports. Th is was the outcome for devel oped and devel-
oping countrie s separately as well as combine d. Clearly, the results suf fer from aggregation bias in t hat what may be
true for all count ries together, may not be true f or at least some countri es in their sample, su ch as the United Kingdom .
The second grou p includes studies t hat add a time dimension to t he above cross-sectiona l study and use panel
models. The lis t includes Thursby and Thurs by (1985), Sauer and Bohara (20 01), and Tenreyro (20 07). All three
studies reveal t hat in the sample of develope d countries, which includ e the United Kingdom, exchan ge rate vola-
tility has no sig nificant impact o n trade flows. Aga in, these studies s uffer from aggreg ation bias in that what is tr ue
for one cross-sec tional unit in the panel, m ay not be true for another unit .2
To address the aggreg ation bias in the second group, th e third category includes s tudies that have used only
time-series m odels and included United Kin gdom in their samples. These s tudies have provided mixed findi ngs.
While Arize (1995, 1996, 1997), Arize and Shwif f (1998), Asseery and Peel (1991), Chowdhury (1993), Kenen a nd
Rodrick (1986), Peree and Steinh err (1989), Qian and Varangis (1994), found a negative impac t of volatility on the
United Kingdom's t rade flows, Bailey, Tavlas, and Ulan (1987), Kro ner and Lastrape s (1993), Holly (1995), found no
effects . These studies have most ly used aggregate expor ts of the United Kingdom wi th the rest of the world and
still suffer f rom another aggregati on bias in that trade flows of th e United Kingdom with one t rading partner may
react diffe rently to exchange rate volatil ity than her trade flow s with another partne r.
Thus, the four th group investigate s the impact of exchange r ate volatility on bilate ral trade flows of the U nited
Kingdom with ea ch partner to reduce the agg regation bias. Thursby and Thu rsby (1987) assesse d the effects of
exchange rate volat ility on the export v alue of the UK to each of its 17 par tners using a gravit y model and showed
that the link is neg ative. Since no quantit y data and no price dat a were available on a bilate ral basis, they used t heir
product, t hat is, export value. Th erefore, it is not clear if the ne gative relation is betwee n exchange rate volatility
and export p rices or betwee n exchange rate volat ility and expo rt quantitie s. Furtherm ore, since the es timated
model was a panel m odel, again it suffers f rom aggregation bias. Ind eed, when Aristotelous (2 002) included only
the UK expor ts to the United States in a tim e-series model and use d data from 1989 to 1999, he found no signifi-
cant effec ts of exchange rate volatilit y.3
1See Hooper a nd Kohlhagen (1978), De Gr auwe (1988) and Pere e and Steinherr (1989) for t heoretica l development s and the above arg uments.
2Such an aggre gation bias is als o recognized bet ween Emergin g Market Economi es and other deve loping countri es when they are p ooled togethe r
(Hall et al. 2010) .
3Note that Pozo (1992) use d an old set of data ov er the period 190 0–1940 and found negati ve effects of e xchange rate vola tility on Brit ish exports
to the United St ates Similar ad verse effec ts are also repo rted by Cushma n (1988).
|
53
BAHMANI-OS KOOEE Et Al.
Suspectin g that the studies in the four th group also suffer from ag gregation bias, the fift h category includes
two studies t hat have disaggregated the UK-US tr ade flows by industry. Masku s (1986) disaggre gated the US ex-
ports to the U nited Kingdom by one-digit SITC cate gories and showed that exchange rat e volatility has adverse
impact on four i ndustries, t hat is, machiner y, transpor t, chemicals, and miscellan eous manufac turers. However,
when Bahmani- Oskooee and Kovyrya lova (2008) disaggre gated UK-US trade flows by thre e-digit SITC industries
and included 177 th ree-digit industries in t heir analysis, they found mu ch more support for signif icant effects of
exchange rate volat ility on the US exports to t he United Kingdom and her impo rts from the United Kingd om (or
UK export s to the United States). More pr ecisely, using annual data ove r the period 1971–2003, they found signif-
icant short-run ef fects of volatilit y on exports of 99 indus tries and imports of 109 industries. However, shor t-run
effects l asted into the long run in 86 ind ustries and 62 import ing industries.4
A common feature of a ll studies in all categori es reviewed above is that they have a ll assumed that the effec ts
of exchange rate volat ility on trade flows are sy mmetric. The symmetr y assumption implies that if a n increase in
volatility by one s tandard deviation reduces t rade by x%, a dec rease in volatility by one stand ard deviation will
boost the tra de by x%. However, this need not b e the case. First, B ahmani-Oskooee and A ftab (2017) have argued
and demonstr ated that trad ers’ response to a n increased volat ility could be di fferent from th eir response to a
decreased volat ility, mostly due to changes in th eir expectation about t he ability of the government s to stabilize
the exchange rate. S econd, Bahmani-Oskooee a nd Nouira (2019) argued that trade rs could react asymmet rically
to exchange rate volat ility if they assi gn higher weights f or losses compar ed to gains from hol ding foreign ex-
change to hedge aga inst future uncertain ties. Third, since any asymm etric effect is commonl y attributed to non-
linearity, Belke an d Goecke (2005) point at no nlinearities in the rel ation between total eco nomy employment and
the exchange rate. T hey identify a band of ina ction within which relat ionship between empl oyment and its deter-
minants inclu ding exchange rate u ncertaint y is weaker compare d to action band , hence asymmet ric response.
Finally, if trade fl ows respond asymmet rically to changes in the r eal exchange rate itsel f, we would expect them to
respond asym metrically to exchange ra te volatility too.5
Therefore, our m ain goal in this paper is to revisit the ef fects of the real dollar–pound volat ility on the trade
flows betwee n the United Kingdo m and the United States by engaging in a symmetry a nalysis. We use monthly
data and trad e flows of 66 two-digit industr ies that trade between the t wo countries. The main reaso n for con-
centrating on t wo-digit industries is that th e data come on monthly frequenc y, which is usefu l in increasing the
number of obser vations to gain ef ficiency i n the estimatio n. The rest of the p aper is organize d as follows. In
Section II, w e outline the models an d discuss the estimati on method. The empiri cal results are repo rted in Section
III that is followed by a s ummary in Sec tion IV. Finally, the definition of the var iables and sour ces of the data is
provided in the Appendix.
2 | THE MODELS AND METHODS6
Bahmani-Oskooee and Aftab (2017) followed the literature in formulating their export and import demand mod-
els. General ly, the models include a scal e variable proxied by a measu re of economic activit y, a relative price term
proxied by the real e xchange rate, and a measur e of uncertainty pr oxied by the volatility of t he real exchange rate.
We modify their n otations in order for the mo dels to conform to the US–UK trad e as follows:
4While in most i ndustries th e effects of d ollar-pound volat ility was nega tive, in some ind ustries it was po sitive. Such po sitive effec ts of volatilit y is
also repor ted on many two-d igit industri es that expor t from Ireland to t he United Kingd om by Doyle (2001) b ut not by de vita an d Abbott (200 4)
who looked int o the UK export s to EU14. Indeed, th ey found no signif icant long-ru n effects of ex change rate volat ility either a t bilateral leve l or at
sectoral level.
5On the asymm etric effec ts of exchange rat e changes on the tr ade balance se e Bahmani-Osko oee and Faridit avana (2016), Arize e t al. (2017), and
Nusair (2016).
6This secti on closely foll ows Bahmani-Os kooee and Aft ab (2017).
To continue reading
Request your trial