Gu ié ez, Emiliano; Vi dis, Juan Ma celo; Melle , Leand o; Domínguez, Diego
Leand o
A icle
P ice s a egies and economic unce ain y: Case-s udy
o he A gen inian pha maceu ical sec o using machine
lea ning
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
P o ided in Coope a ion wi h:
Uni e sidad Pablo de Ola ide, Se illa
Sugges ed Ci a ion: Gu ié ez, Emiliano; Vi dis, Juan Ma celo; Melle , Leand o; Domínguez, Diego
Leand o (2024) : P ice s a egies and economic unce ain y: Case-s udy o he A gen inian
pha maceu ical sec o using machine lea ning, Re is a de Mé odos Cuan i a i os pa a la Economía
y la Emp esa, ISSN 1886-516X, Uni e sidad Pablo de Ola ide, Se illa, Vol. 38, pp. 1-16,
h ps://doi.o g/10.46661/ e .me odoscuan .econ.emp esa.6407
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/314469
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by-sa/4.0/
ARTÍCULOS 1
Cómo ci a : Melle , L., Vi dis, J., Gu ié ez, E., & Domínguez, D. Es a egias de p ecios e ince idumb e económica: Un es udio de caso pa a la indus ia
a macéu ica a gen ina usando ap endizaje au oma izado. Re is a De Mé odos Cuan i a i os Pa a La Economía Y La
Emp esa, 38. h ps://doi.o g/10.46661/ e .me odoscuan .econ.emp esa.6407
Uni e sidad Pablo de Ola ide (España)
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
núme o 38, 2024
ISSN: 1886-516X
DOI: 10.46661/ e me odoscuan econemp esa.6407
Sección: A ículos
Recibido: 23-12-2021
Acep ado: 05-04-2024
Publicado: 03-12-2024
Páginas: 1-16
Es a egias de p ecios e ince idumb e económica: Un caso
de es udio aplicado al sec o a macéu ico a gen ino
usando ap endizaje au oma izado
P ice s a egies and economic unce ain y: case-s udy o
he A gen inian pha maceu ical sec o using machine
lea ning
Emiliano Gu ié ez
Uni e sidad Nacional del Cen o de la P o incia de Buenos Ai es (A gen ina)
h ps://o cid.o g/0000-0002-6424-996X
emiliano.gu ie e[email p o ec ed]
Juan Ma celo Vi dis
Uni e sidad Nacional del Su (A gen ina)
h ps://o cid.o g/ 0000-0001-7118-9259
[email p o ec ed]
Leand o Melle
Ins i u o de In es igaciones Económicas y Sociales del Su (A gen ina)
h ps://o cid.o g/ 0000-0003-1246-4198
[email p o ec ed]
Diego Leand o Domínguez
Uni e sidad Nacional del Su (A gen ina)
dieguini[email p o ec ed]
RESUMEN
In oducción: En agos o de 2019, un esul ado inespe ado en las elecciones
p esidenciales gene ó una a iación en el ipo de cambio y la in lación espe ados. El
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
N. 38, 2024
–
ISSN: 1886-516X
–
DOI: 10.46661/ e me odoscuan econemp esa.6407
–
[Págs. 1-16]
Es a egias de p ecios e ince idumb e económica: Un caso de es udio aplicado al sec o a macéu ico a gen ino
usando ap endizaje au oma izado
Emiliano Gu ié ez, Juan Ma celo Vi dis, Leand o Melle , Diego Leand o Domínguez
ARTÍCULOS
2
obje i o de es e es udio es analiza la elación en e la pa icipación de me cado y la
decisión de inc emen a los p ecios en la indus ia a macéu ica en A gen ina.
Mé odos: Se ob u ie on da os semanales en línea sob e las a iaciones de los p ecios
de algunos medicamen os median e écnicas de web sc apping, y luego se aplica on
algo i mos de clasi icación (Random Fo es s, G adien Boos ing Machine y eg esión
logís ica).
Resul ados: Los esul ados ue on dispa es. Se encon ó que la pa icipación de
me cado es impo an e de acue do a los mé odos basados en á boles (Random
Fo es s y G adien Boos ing Machine). Sin emba go, en la eg esión logís ica, dicha
a iable no e a signi ica i a.
Conclusiones: La ola ilidad en el ipo de cambio que siguió al esul ado de la elección
causó a ios cambios en los p ecios espe ados, y la es uc u a del me cado
a macéu ico in luyó sob e las eacciones de p ecios esul an es. Los labo a o ios que
enían una mayo pa icipación de me cado inc emen a on sus p ecios p ime o.
PALABRAS CLAVE
Me cado a macéu ico; ince idumb e; ap endizaje au oma izado.
ABSTRACT
In oduc ion: In Augus 2019 an unexpec ed p esiden ial elec ion esul caused a
change in expec ed exchange and in la ion a es. The objec i e o his s udy is o
analyze he ela ion be ween ma ke sha e and he decision o inc easing p ices in
he pha maceu ical indus y in A gen ina.
Me hods: Online weekly da a on a ia ions o some medicine’s p ices we e ob ained
using web sc aping, and hen classi ica ion algo i hms (Random Fo es s, G adien
Boos ing Machine and logis ic eg ession) we e applied.
Resul s: The esul s we e mixed: ma ke sha e was ound o ha e high impo ance
in ee-based me hods. (Random Fo es s and G adien Boos ing Machine).
Howe e , in logis ic eg ession, his a iable wasn’ signi ican .
Conclusions: Exchange a e ola ili y a e he elec ion esul caused se e al
changes on p ice expec a ions, and pha maceu ical ma ke s uc u e in luenced
he esul ing p ice eac ions. Labo a o ies which owned a highe ma ke sha e ose
hei p ices i s .
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
N. 38, 2024
–
ISSN: 1886-516X
–
DOI: 10.46661/ e me odoscuan econemp esa.6407
–
[Págs. 1-16]
Es a egias de p ecios e ince idumb e económica: Un caso de es udio aplicado al sec o a macéu ico a gen ino
usando ap endizaje au oma izado
Emiliano Gu ié ez, Juan Ma celo Vi dis, Leand o Melle , Diego Leand o Domínguez
ARTÍCULOS
3
KEYWORDS
Pha maceu ical ma ke ; unce ain y; machine lea ning.
JEL classi ica ion: C89, I10.
MSC2010: 62P20, 91B26.
INTRODUCTION
The ela ionship be ween ma ke sha e and d ug p ices has been well-documen ed in p io
li e a u e (Lu and Comano , 1998). Companies wielding g ea e ma ke powe possess he
capabili y o se highe p ices, bo h du ing he ini ial launch o a new d ug and in he ensuing
yea s. In high in la iona y en i onmen s, companies may ind i necessa y o aise p ices in o de
o sus ain p o i s agains he backd op o escala ing cos s. Conside ing he a iabili y in ma ke
sha es among companies, and hus hei di e ing le els o ma ke powe , i becomes a subjec
o inqui y whe he all companies can uni o mly adjus p ices, o i hose wi h g ea e ma ke
powe can sus ain o augmen hei p o i s.
In his pape we examine he ela ion be ween ma ke sha e and he abili y o inc easing p ices
a e an exogenous mac oeconomic dis up ion caused by an unp edic ed ou come in he
p esiden ial elec ion which ook place on Augus 11 h, 2019. A e he elec ion, he exchange a e
changed d ama ically, and he A gen inian Peso (ARS) los 25% o i s alue in ou days. Also, he
expec ed exchange a e o he end o he yea inc eased signi ican ly om 50 o 66,70 ARS
(Cen al Bank o A gen ine Republic [BCRA o i s ac onym in Spanish], 2019). I is wo h no ing
ha he e is e idence o inc ease in in la ion a e a e de alua ion in A gen ina (Ba be is, 2021;
Cas iglione, 2017; O e o e al., 2005).
These changes in cu ency p ices and expec a ions migh igge p ice s a egies among
economic agen s. In his ega d, he pha maceu ical sec o ep esen s a signi ican case, as he
cos s o medicines could pose a ba ie o he ealiza ion o he igh o heal h, which is a key
objec i e o public policy (Gu man and La a ello, 2011; Pe ehudo e al., 2019). This s udy aims
o de e mine he in luence o exchange a e inc ease and pha maceu ical sec o concen a ion
on p ices a e he elec ion esul on Augus 11 h 2019. The es o he pape is o ganized as ollows.
The second sec ion p esen s p e ious heo e ical and empi ical s udies abou he ela ion
be ween unce ain y, exchange a e changes and indus ial concen a ion, and p ice
adjus men s. The hi d sec ion desc ibes me hodology and da a. The ou h sec ion shows he
esul s o his in es iga ion. The i h sec ion p esen s a discussion o he esul s. Finally, he six h
sec ion con ains concluding commen s.
BACKGROUND
Economic policy as a sou ce o unce ain y is cu en ly a ele an opic. This is suppo ed by he
apid g ow h in li e a u e s udying he ela ion be ween unce ain y o economic policy and he
beha io o se e al mic oeconomic and mac oeconomic a iables (Al-Thaqeb and Algha abali,
2019). A la ge numbe o exis ing s udies ha e examined how Economic Policy Unce ain y Index
(EPU) (Bake e al., 2016) in e ela e wi h commodi y p ices (Wang e al., 2015), o consume p ice
index (Jones and Olson, 2013; Balcila e al., 2016a). This app oach would be app op ia e in o de
o achie e his pape objec i e. Howe e , no EPU is a ailable in A gen ina. P e ious empi ical
esea ch highligh s ha EPU and exchange a e ola ili y a e posi i ely co ela ed (K ol, 2014;
Balcila e al., 2016b; Dai e al, 2017; Olanipekun e . al, 2019; Chen e . al, 2020). Simila ly, bu using
mo e equen da a, Ba shch (2019) inds he same posi i e co ela ion. Also, Ku usawa (2016)
inds simila e idence, bu poin s ou ha , in he long e m, his co ela ion is able o change i s
sign.
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
N. 38, 2024
–
ISSN: 1886-516X
–
DOI: 10.46661/ e me odoscuan econemp esa.6407
–
[Págs. 1-16]
Es a egias de p ecios e ince idumb e económica: Un caso de es udio aplicado al sec o a macéu ico a gen ino
usando ap endizaje au oma izado
Emiliano Gu ié ez, Juan Ma celo Vi dis, Leand o Melle , Diego Leand o Domínguez
ARTÍCULOS
4
The in luence o he ma ke s uc u e on p ices has been ecognized by se e al au ho s.
Acco ding o He gue a (1994), he ele ance o ma ke s uc u e began o ise in he i s hal o
he 80s. In his pe iod, he Ame ican dolla was gaining alue bu he cu en accoun and impo
p ices we e una ec ed. This si ua ion mo i a ed Mann’s (1986) and Do nbusch’s (1987)
in es iga ions, which ep esen ed seminal con ibu ions ha inspi ed a la ge numbe o s udies.
These s udies de eloped pa ial equilib ium models in o de o explain why p ices in di e en
sec o s (e en di e en i ms) changed in di e en magni udes as a esponse o exchange a e
a ia ions. To his end, Ki man and Phlips (1996) de eloped a model ha shows he ele ance o
ma ke sha e o he impac o exchange a e a ia ions on consume p ices.
Se e al empi ical in es iga ions also sugges ha p ice a ia ion, as a esponse o exchange
a e changes, is di e en be ween sec o s (Bha acha ya e . al, 2008; Mallick and Ma ques, 2010;
He e . al, 2015; Tho becke and Ka o, 2018). Howe e , only some o hem poin ou he ma ke
cha ac e is ics as he cause o hese di e ences (Bha acha ya e . al, 2008; Mallick and Ma ques,
2010; He e . al, 2015; Tho becke and Ka o, 2018).
Pha maceu ical sec o in A gen ina
The pha maceu ical sec o in A gen ina can be classi ied as an impe ec ly compe i i e ma ke ,
cha ac e ized by p oduc ion and consump ion complexi ies. Fi s , he demand is inelas ic due o
di icul ies o a oiding medicine pu chasing when needed o ees ablish o main ain a heal hy
s a us. Second, on he supply side, ma ke ing s a egies a e implemen ed by i ms, a ec ing he
consume powe o decision, and p o iding minimum in o ma ion abou hei p oduc s
(B amuglia e . al., 2015). Thi d, e en hough all labo a o ies ep esen a low ma ke sha e
(U bi zondo e al., 2013), he wo mos impo an pha maceu ical business chambe s own 79% o
consume sales.
Simila ly, a high concen a ion is able o be ound i he ma ke is di ided by he apeu ic class
(3 d le el classi ica ion o ATC). In 2019, 4.5% o hese di ided ma ke s we e monopolies, while
nea ly a 25% o hem we e cha ac e ized by a He indahl-Hi schman Index (HHI) highe han
7500 (S a emen N° 113 Na ional Commission o Compe i ion De ense [CNDC o i s ac onym in
Spanish], 2019). E en hough a pa o he supply is o e ed by he public sec o , mos o
go e nmen al labo a o ies a e no au ho ized by he Na ional Adminis a ion o Medicines, Food
and Medical Technology (ANMAT o i s ac onym in Spanish) o sell hei p oduc s o he public
(San os and Thomas, 2018; B amuglia e al., 2012). Also, hese labo a o ies a e loca ed in Buenos
Ai es, San a Fé and Có doba, which ep esen s a non-uni o m e i o ial dis ibu ion (San os and
Thomas, 2018).
This sec o in e dependence is ano he ele an ac o , mainly, due o i s ela ion o he heal h
ca e ma ke . This ma ke ’s dis o ions a e able o impac on he pha maceu ical one (Aloma e
al., 2005). Simila ly, medicines a e an impo an componen o many households' expendi u es
which can ep esen a eg essi e income ac o (Apella, 2006; Pe ica a, 2008).
DATA AND METHODS
We ob ained medicine p ices by web sc aping, which is an au oma ized compu a ional
echnique used o ex ac high equency online in o ma ion a a lowe cos han adi ional p ice
su eys. This p ocedu e was applied o he web ca alog o a e ail pha macy chain called
Fa maci y and a o al o 29,034 obse a ions we e e ie ed, one o each week be ween 12 h
Augus and 3 d Sep embe . 46% o hese obse a ions we e classi ied as p ice inc eases, which
ep esen s a balanced sample.
On he o he hand, ma ke pa icipa ion da a, unde s ood as he p oducing labo a o y’s sha e
o e ail sales, we e ob ained om a echnical epo done by he CNDC. These da a a e
p esen ed in G aph 1.
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
N. 38, 2024
–
ISSN: 1886-516X
–
DOI: 10.46661/ e me odoscuan econemp esa.6407
–
[Págs. 1-16]
Es a egias de p ecios e ince idumb e económica: Un caso de es udio aplicado al sec o a macéu ico a gen ino
usando ap endizaje au oma izado
Emiliano Gu ié ez, Juan Ma celo Vi dis, Leand o Melle , Diego Leand o Domínguez
ARTÍCULOS
5
G aph 1. Re ail sales sha e by labo a o y
Sou ce: Own elabo a ion based on a echnical epo o pha maceu ical ma ke s uc u e (S a emen N° 113
CNDC, 2019).
A dependen a iable ( ) was de ined as a dummy equal o one i a medicine p ice,
measu ed in ARS, was highe han he p e ious week. O he wise, was se o ze o. As
independen a iables we used he ma ke sha e o he medicine manu ac u ing labo a o y (
and he di e ence be ween he exchange a e in he p e ious week and he one be o e
. The a iable ep esen s he obse ed exchange a e a ia ion.
The a iable ha ep esen s he exchange a e’s e olu ion is es ablished as he di e ence o he
maximum alues ( ) o he wo p e ious weeks, as exposed in he ollowing equa ion
The exchange a e was ob ained om he sale p ice o he Ame ican Dolla , as published by
BCRA. G aph 2 depic s he daily exchange a es du ing he pe iod o analysis. A sudden inc ease
is obse ed a e Augus 11, 2019, he day o he p esiden ial elec ion.
G aph 2. Daily exchange a e (29 July 2019 - 5 Sep embe 2019)
Sou ce: BCRA (2024).
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
N. 38, 2024
–
ISSN: 1886-516X
–
DOI: 10.46661/ e me odoscuan econemp esa.6407
–
[Págs. 1-16]
Es a egias de p ecios e ince idumb e económica: Un caso de es udio aplicado al sec o a macéu ico a gen ino
usando ap endizaje au oma izado
Emiliano Gu ié ez, Juan Ma celo Vi dis, Leand o Melle , Diego Leand o Domínguez
ARTÍCULOS
6
We ind i impo an o dis inguish be ween pha maceu ical na ional labo a o ies and o eign
coun e pa s. A dicho omous a iable (lab_a g) akes a alue o 1 i he labo a o y p oducing
he medicine is A gen ine, and 0 i i is a o eign labo a o y. This in o ma ion was ex ac ed om
he Indus ial Chambe o A gen ine Pha maceu ical Labo a o ies webpage (CILFA, ac onym in
Spanish).
The choice o whe he o inc ease he p ice o no may be signi ican ly in luenced by he d ug
associa ed wi h each medicine. Fo his eason, we in oduce p incipal d ugs as dicho omous
a iables. We conside only he 15 mos equen ly used d ugs in ou sample.
Addi ionally, he weeks ollowing he elec ion esul s se e as an impo an signal o
unde s anding how he p ice inc ease adjus s o e ime. I is expec ed ha de ec ing a p ice
inc ease becomes less likely in he subsequen weeks a e he e en . The e o e, an o dinal
a iable ( ) is included in he es ima ions.
To ul ill he objec i es o his in es iga ion, we conduc ed es ima ions using logis ic eg ession,
Random Fo es s, and G adien Boos ing Machine. The da ase was spli in o a aining se (90%)
and a es se (10%). The e o e, he aining se was u ilized o pa ame e izing he models, and
he pe o mance o each model was e alua ed using he es se . All es ima ions we e
implemen ed using he R p og amming language (R Co e Team, 2022). Wi h he excep ion o
logis ic eg ession, speci ic packages in R we e employed o execu e he algo i hms.
The Logis ic Reg ession is use ul when he dependen a iable is bina y. A Be noulli dis ibu ion
o he p edic ed e en is assumed, wha in his pape is he inc ease o a medicine p ice (
. The eg ession’s pa ame e s co espond o he independen a iables and he
eg ession is modeled as ollows
whe e is he independen a iable ec o . I he denomina o is simpli ied and he exponen s
a e dele ed, he p e ious equa ion is exp essed as ollows
The p e ious exp ession indica es ha he logis ic model can be s a ed as he quo ien o e en
p obabili ies (odds- a io). Also, i shows how he coe icien s di ec ly impac on his a io (Hai e
al., 2014).
B eiman (2001) has sugges ed he u iliza ion o classi ica ion and eg ession ees (CART) in
o de o gene a e a classi ie algo i hm by applying he mean o a la ge enough numbe o
in e ela ed ees. This algo i hm, called Random Fo es (RF), assumes ha mul iple ee c ea ion
educes he a iabili y in compa ison o any indi idual ee, and minimizes he inal model
o e i ing. This co esponds o an assembly me hodology, because he inal model is composed
o hund eds (o housands) o p edic i e ones, independen ly c ea ed (Has ie, 2009).
The i s s ep consis s o subsampling a pa o he obse a ions by boo s apping (Typically, 2/3
o he aining da a). E e y gene a ed ee will be composed by a numbe o p edic o s
(James e al., 2013). This p ocedu e is pe o med imes, gene a ing a di e en classi ica ion ee
e e y ime.
As a esul , he p edic ion o a highe p ice is decided by o es, which means ha he medicine
inal classi ica ion will depend on how i was classi ied mos o he imes:
In his esea ch, he implemen a ion o RF algo i hm is h ough he packages ca e (Kuhn, 2008)
and Range (Liaw & Wiene , 2002) om R.
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
N. 38, 2024
–
ISSN: 1886-516X
–
DOI: 10.46661/ e me odoscuan econemp esa.6407
–
[Págs. 1-16]
Es a egias de p ecios e ince idumb e económica: Un caso de es udio aplicado al sec o a macéu ico a gen ino
usando ap endizaje au oma izado
Emiliano Gu ié ez, Juan Ma celo Vi dis, Leand o Melle , Diego Leand o Domínguez
ARTÍCULOS
7
In con as o RF, in which e e y ee is independen o he p e ious ones, F iedman (2001) has
sugges ed he cons uc ion o an algo i hm by a sequen ial ee building, which allows using
p e ious ee’s in o ma ion. This me hodology is named G adien Boos ing Machine (GBM), and
i minimizes a loss unc ion ha measu es he di e ence be ween he eal and es ima ed
alue.
Consequen ly, he classi ying me hod p oceeds ecu si ely imes in o de o adjus he
pa ame e using p e ious models (F). The e o e, he las i e a ion o he model is able o be
es ablished as
The alue is called he lea ning a e and, in compa ison o RF, he algo i hm is indi e en abou
he a iables employed in e e y ee, bu i conside s he ami ica ions in each one . The
implemen a ion o GBM applied was wi h he gbm package (G eenwell e al.,2022)
An inhe en challenge in ee-based models such as G adien Boos ing Machine (GBM) and
Random Fo es s (RF) is he loss o in e p e abili y. Unlike linea models like logis ic eg ession,
whe e coe icien s indica e he con ibu ion o each a iable o he ou come, "black-box" models
like RF and GBM lack his anspa ency.
Add essing his issue, an in iguing ques ion a ises: how can we e alua e he impac o ea u es
in such "black-box" models? A aluable app oach o enhance in e p e abili y in machine
lea ning is he u iliza ion o model-agnos ic me hods. The model-agnos ic app oach e ains
om imposing assump ions abou he in insic cha ac e is ics o he model (Molna , 2022).
While a ious me hods exis , we speci ically selec ed he pe mu a ion ea u e impo ance
me hod o assess he in e p e abili y o ou ee-based models. This app oach enhances
explicabili y wi hou elying on assump ions abou he in e nal s uc u e o he model.
In he ea u e impo ance p ocedu e, he con en ional app oach in ol es unning a model wi h
all a iables in he o iginal da a . Subsequen ly, he a iable o in e es in he ec o wi h
he o al independen a iables is pe mu ed, and he model is e un. The esul ing ma ix
dis up s he linkage be ween and he a iable o in e es in p edic ing . The me ic
used o measu e he loss o ea u e impo ance due o pe mu a ion is he c oss en opy (CE),
which quan i ies inco ec p edic ions. This me ic is de ined as:
He e is he p obabili y ha a p ice was highe han he p e ious week, and a low alue o CE
indica es a well- i model. High alues o CE sugges ha he a iable is c ucial o he p edic ion,
as i s absence ad e sely a ec s he model's pe o mance. In con as , he package iml (Molna
e al., 2018) p o ides an in e es ing applica ion o Fea u e Pe mu a ion Impo ance.
RESULTS
The logis ic eg ession es ima ion esul s, as p esen ed in Table 1, indica e a signi ican
ela ionship be ween a p ice inc ease and a ia ions in he exchange a e. Addi ionally, i e eals
a s a is ically signi ican associa ion be ween he decision o aise p ices and whe he he
p oduce is a na ional labo a o y.
The week ollowing he elec ions plays a c ucial ole in he s a egic decision o an icipa e and
adjus he cu en p ices. As expec ed, a g ea e numbe o weeks a e he elec ions nega i ely
and signi ican ly impac s he likelihood o a p ice inc ease.
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
N. 38, 2024
–
ISSN: 1886-516X
–
DOI: 10.46661/ e me odoscuan econemp esa.6407
–
[Págs. 1-16]
Es a egias de p ecios e ince idumb e económica: Un caso de es udio aplicado al sec o a macéu ico a gen ino
usando ap endizaje au oma izado
Emiliano Gu ié ez, Juan Ma celo Vi dis, Leand o Melle , Diego Leand o Domínguez
ARTÍCULOS
8
Table 1. Logis ic eg ession
Coe icien
S anda d
e o
- alue
p- alue
(In e cep )
-0,676
0,055
-12,288
0,000*
-0,145
0,022
-6,704
0,000*
conc
-0,946
0,732
-1,292
0,196
_usd
7,444
0,121
61,548
0,000*
lab_a g
0,313
0,036
8,633
0,000*
ca edilol
-0,065
0,130
-0,500
0,617
me o mina
-0,035
0,132
-0,267
0,789
p egabalina
0,252
0,136
1,847
0,065
osu as a ina
0,203
0,142
1,434
0,152
que iapina
0,232
0,157
1,476
0,140
alp azolam
0,170
0,154
1,107
0,268
enalap il
0,533
0,153
3,480
0,001*
ibup o eno
0,461
0,155
2,984
0,003*
diclo enac
-0,225
0,158
-1,422
0,155
amoxicilina
0,133
0,159
0,836
0,403
a o as a ín
0,518
0,162
3,190
0,001*
sildena il
-0,085
0,164
-0,521
0,603
clonazepam
0,236
0,167
1,416
0,157
ispe idona
-0,152
0,177
-0,859
0,390
cla i omicina
0,026
0,175
0,146
0,884
N = 28807 AIC = 27872 *'=: signi ican (Con idence In e al o 95%)
Sou ce: own elabo a ion.
Ano he no ewo hy inding is he lack o signi icance ega ding he ma ke sha e. Acco ding o
he logis ic eg ession es ima ion, he e seems o be no linea ela ionship be ween he decision
o inc ease he selling p ice and he pe cen age o sales held by he labo a o y ac oss he b and.
The con usion ma ix o he logis ic eg ession (Table 2) illus a es he model's pe o mance on
he aining se . The logis ic eg ession accu acy s ands a 59.94%, wi h be e pe o mance in
p edic ing non-inc eased p ices (60.36%) han in p edic ing p ice inc eases (56.52%).
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
N. 38, 2024
–
ISSN: 1886-516X
–
DOI: 10.46661/ e me odoscuan econemp esa.6407
–
[Págs. 1-16]
Es a egias de p ecios e ince idumb e económica: Un caso de es udio aplicado al sec o a macéu ico a gen ino
usando ap endizaje au oma izado
Emiliano Gu ié ez, Juan Ma celo Vi dis, Leand o Melle , Diego Leand o Domínguez
ARTÍCULOS
15
Gu man, G. & La a ello, P. (2011). Fo mas de o ganización de las emp esas bio echnológicas en el sec o
a macéu ico a gen ino. Desa ollo Económico, 51 (201), 81-105. h ps://www.js o .o g/s able/23612337
Hai , J. F., Black, W. C., Babin, B. J., & Ande son, R. E. (2014). Mul i a ia e da a analysis: Pea son new
in e na ional edi ion. Essex: Pea son Educa ion Limi ed.
Has ie, T., Tibshi ani, R., and F iedman, J. (2009). The elemen s o s a is ical lea ning: Da a mining, in e ence,
and p edic ion. Sp inge Science & Business Media.
He, B., Zhu, H., Chen, D., & Shi, Y. (2015). On pass- h ough o RMB exchange a e o p ices o di e en indus ies.
P ocedia Compu e Science, 55, 886-895. h ps://doi.o g/10.1016/j.p ocs.2015.07.146
He gue a, I. (1994). Indus y p ice adjus men o exchange a e luc ua ions in oligopoly: An empi ical s udy
o he pass- h ough ela ionship de e minan s in he Spanish au omobile indus y, 1981-1991 (PhD
Thesis). Eu opean Uni e si y Ins i u e, Flo ence. h ps://op.eu opa.eu/es/publica ion-de ail/-
/publica ion/213 69a9-c5b3-4219-b 1d-a5195 9ab738
James, G., Wi en, D., Has ie, T., & Tibshi ani, R. (2013). An in oduc ion o s a is ical lea ning (Vol. 112, p. 18).
New Yo k: sp inge .
Jones, P. M., & Olson, E. (2013). The ime- a ying co ela ion be ween unce ain y, ou pu and in la ion:
E idence om a DCC-GARCH model. Economics Le e s, 118, 33-37.
h ps://doi.o g/10.1016/j.econle .2012.09.012
Ki man, A., & Phlips, L. (1996). Exchange a e pass- h ough and ma ke s uc u e. Jou nal o Economics, 64
(2), 129-154. h ps://doi.o g/10.1007/BF01250111
K ol, R. (2014). Economic policy unce ain y and exchange a e ola ili y. In e na ional Finance, 17 (2), 241-
255. h ps://doi.o g/10.1111/in i.12049
Ku asawa, K. (2016). Policy unce ain y and o eign exchange a es: The DCC-GARCH model o he
US/Japanese o eign exchange a e. In e na ional Jou nal o Economic Sciences, 5 (4), 1-19.
h ps://doi.o g/10.52950/ES.2016.5.4.001
Kuhn, M. (2008). Building P edic i e Models in R Using he ca e Package. Jou nal o S a is ical So wa e,
28(5), 1–26. h ps://doi.o g/10.18637/jss. 028.i05-
Liaw, A., & Wiene , M. (2002). Classi ica ion and Reg ession by andomFo es . The R Jou nal, 2/3, 18-22.
h ps://jou nal. -p ojec .o g/a icles/RN-2002-022/RN-2002-022.pd
Lu, Z. J., & Comano , W. S. (1998). S a egic P icing o New Pha maceu icals. The Re iew o Economics and
S a is ics, 80(1), 108–118.
Mallick, S., & Ma ques, H. (2010). Da a equency and exchange a e pass- h ough: E idence om India’s
expo s. In e na ional Re iew o Economics and Finance, 19, 13-22. h ps://doi.o g/10.1016/j.i e .2009.02.007
Mann, C. L. (1986). P ices, p o i ma gins, and exchange a es. Fede al Rese e Bulle in, (Jun), 366-379.
h ps:// ase .s louis ed.o g/ iles/docs/publica ions/FRB/pages/1985-1989/31910_1985-1989.pd
Mayo Clinic. (2023). D ugs and Supplemen s. A o as a in (O al Rou e).
h ps://www.mayoclinic.o g/diseases-condi ions/demen ia/symp oms-causes/syc-20352013
Medica ions Commi ee o he Spanish Associa ion o Pedia ics (2015). Pediamécum. Edición 2015.
h ps://www.aeped.es/comi e-medicamen os/pediamecum/enalap il.
Molna , C. (2022). In e p e able Machine Lea ning: A Guide o Making Black Box Models Explainable (2nd
ed.). h ps://ch is ophm.gi hub.io/in e p e able-ml-book/
Re is a de Mé odos Cuan i a i os pa a la Economía y la Emp esa
N. 38, 2024
–
ISSN: 1886-516X
–
DOI: 10.46661/ e me odoscuan econemp esa.6407
–
[Págs. 1-16]
Es a egias de p ecios e ince idumb e económica: Un caso de es udio aplicado al sec o a macéu ico a gen ino
usando ap endizaje au oma izado
Emiliano Gu ié ez, Juan Ma celo Vi dis, Leand o Melle , Diego Leand o Domínguez
ARTÍCULOS
16
Molna , C., Casalicchio, G., & Bischl, B. (2018). iml: An R package o In e p e able Machine Lea ning. The
Jou nal o Open Sou ce So wa e, 3 (26), 786. h ps://doi.o g/10.21105/joss.00786
Olanipekun, I. O., Olasehinde-Williams, G., & Güngö , H. (2019). Impac o economic policy unce ain y on
exchange ma ke p essu e. SAGE Open, 9(3), 1-13. h ps://doi.o g/10.1177%2F2158244019876275
O e o, G. A., Cadelli, M. E., Ca bajal, R., y Ce imedo, F. (2005). Explo ando los de e minan es del aspaso de
la de aluación a p ecios: Una explicación del éxi o de alua o io a gen ino de 2002 (documen o de
abajo). G upo de In es igación Económica (GIE) del Minis e io de Economía de la P o incia de Buenos
Ai es. h ps://www.ec.gba.go .a /p ensa/A chi os/Julio2005.pd
Pe ica a, M. (2008). Incidencia de los gas os de bolsillo en salud en sie e países la inoame icanos. San os,
G., and Thomas, H. (2018). P oducción pública de medicamen os. Ciencia, Tecnología y Polí ica, 1(1), 007-
007. h ps://www.cepal.o g/es/publicaciones/6146-incidencia-gas os-bolsillo-salud-sie e-paises-
la inoame icanos
Pe ehudo , S. K., Alexand o , N. V., & Hoge zeil, H. V. (2019). The igh o heal h as he basis o uni e sal heal h
co e age: A c oss-na ional analysis o na ional medicines policies o 71 coun ies. PLOS ONE, 14(6),
e0215577. h ps://doi.o g/10.1371/jou nal.pone.0215577
R Co e Team (2022). R: A language and en i onmen o s a is ica lcompu ing. R Founda ion o S a is ical
Compu ing, Vienna, Aus ia. h ps://www.R-p ojec .o g/
San os, G., y Thomas, H. (2018). P oducción pública de medicamen os: desa íos pa a una polí ica
es a égica en ma e ia de salud, Ciencia, Tecnología y Polí ica, 1 (1), 1-7.
h ps://doi.o g/10.24215/26183188e007
S a emen N° 113/2019. Na ional Commission o Compe i ion De ense, Buenos Ai es, A gen ina, 23 de
Decembe 2019. h ps://www.a gen ina.gob.a /si es/de aul / iles/im._4_-_medicamen os_-
_disposicion_y_anexo.pd
Tho becke, W., & Ka o, A. (2018). Exchange a es and he Swiss economy. Jou nal o Policy Modeling, 40 (6),
1182-1199. h ps://doi.o g/10.1016/j.jpolmod.2018.07.002
U biz ondo, S., Con , W., & Panadei os, M. (2013). La compe encia en el segmen o ups eam de la indus ia
a macéu ica a gen ina. Documen o de abajo, 121.
h p://www. iel.o g/publicaciones/Documen os/DOC_TRAB_1396372329842.pd
Wang, Y., Zhang, B., Diao, X., & Wu, C. (2015). Commodi y p ice changes and he p edic abili y o economic
policy unce ain y. Economics Le e s, 127, 39-42. h ps://doi.o g/10.1016/j.econle .2014.12.030
