da Sil a, Sab ina Espinele; Fonseca, Simone E angelis a; Roma, Ca olina Magda da
Sil a; Han, Seung Hun; Iquiapaza, Robe Aldo
A icle
In es o sen imen and equi y mu ual und pe o mance
in B azil
Jou nal o Economics, Finance and Adminis a i e Science
P o ided in Coope a ion wi h:
Uni e sidad ESAN, Lima
Sugges ed Ci a ion: da Sil a, Sab ina Espinele; Fonseca, Simone E angelis a; Roma, Ca olina Magda
da Sil a; Han, Seung Hun; Iquiapaza, Robe Aldo (2025) : In es o sen imen and equi y mu ual
und pe o mance in B azil, Jou nal o Economics, Finance and Adminis a i e Science, ISSN
2218-0648, Eme ald Publishing Limi ed, Leeds, Vol. 30, Iss. 59, pp. 189-204,
h ps://doi.o g/10.1108/JEFAS-12-2023-0280
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/319677
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/4.0/
In es o sen imen and equi y mu ual
und pe o mance in B azil
Sab ina Espinele da Sil a
Facul y o Economic Sciences, Fede al Uni e si y o Minas Ge ais,
Belo Ho izon e, B azil
Simone E angelis a Fonseca
Ins i u e o Applied Social Sciences, Fede al Uni e si y o Ou o P e o,
Ma iana, B azil
Ca olina Magda da Sil a Roma
Ins i u e o Economic, Adminis a i e, and Accoun ing Sciences,
Fede al Uni e si y o Rio G ande, Rio G ande, B azil
Seung Hun Han
School o Business and Technology Managemen ,
Ko ea Ad anced Ins i u e o Science and Technology,
Daejeon, Republic o Ko ea, and
Robe Aldo Iquiapaza
Facul y o Economic Sciences, Fede al Uni e si y o Minas Ge ais,
Belo Ho izon e, B azil
Abs ac
Pu pose –Focusing on he B azilian equi y mu ual und indus y, his s udy analyzes whe he including he
in es o sen imen index in asse p icing models is impo an o explaining und alpha.
Design/me hodology/app oach –The in es o sen imen index and isk ac o s in he Fama and F ench (1993)
and Ca ha (1997) models we e es ima ed, he isk-adjus ed pe o mance o a sample o equi y mu ual unds in
B azil was e alua ed, and a Uni ed S a es (US) sample was included o a complemen a y pe spec i e. The
sample pe iod spans 2010–2019 o B azil and 2010–2018 o he US.
Findings –The esul s con as ed wi h hose e idenced in he US, whe e he sen imen index was an impo an
ac o in explaining he p obabili y o alpha occu ence, especially in he case o winne unds, de ined as hose
exhibi ing a posi i e and s a is ically signi ican alpha a he 5% le el. O e all, he indings sugges ha , in he
B azilian ma ke , p icing models inco po a ing in es o sen imen as an addi ional ac o ail o adequa ely
cap u e he ou pe o mance p obabili y o equi y mu ual unds. These esul s sugges ha he ac o s in luencing
und pe o mance may di e be ween he wo coun ies and highligh he ele ance o de eloping mo e sui able
in es o sen imen indica o s o eme ging ma ke s.
O iginali y/ alue –This s udy examines he impac o he sen imen index on he pe o mance o equi y mu ual
unds in B azil, speci ically i s in luence on alpha gene a ion.
Keywo ds Sen imen index, Asse p icing models, Equi y unds, Fund pe o mance, B azil
Pape ype Resea ch pape
Jou nal o
Economics,
Finance and
Adminis a i e
Science
189
© Sab ina Espinele da Sil a, Simone E angelis a Fonseca, Ca olina Magda da Sil a Roma, Seung Hun
Han and Robe Aldo Iquiapaza. Published in Jou nal o Economics, Finance and Adminis a i e Science.
Published by Eme ald Publishing Limi ed. This a icle is published unde he C ea i e Commons
A ibu ion (CC BY 4.0) licence. Anyone may ep oduce, dis ibu e, ansla e and c ea e de i a i e wo ks
o his a icle ( o bo h comme cial and non-comme cial pu poses), subjec o ull a ibu ion o he
o iginal publica ion and au ho s. The ull e ms o his licence maybe seen a h p://c ea i ecommons.o g/
licences/by/4.0/legalcode
This s udy was inanced in pa by he Coo denaç~
ao de Ape eiçoamen o de Pessoal de N�
ı el Supe io
– B asil (CAPES) – Finance Code 001, by CNPq – B azil Na ional Resea ch Council, G an 406954/
2021-6, and Resea ch P oduc i i y G an – PQ.
The cu en issue and ull ex a chi e o his jou nal is a ailable on Eme ald Insigh a :
h ps://www.eme ald.com/insigh /2077-1886.h m
Recei ed 6 Feb ua y 2024
Re ised 15 Oc obe 2024
25 No embe 2024
12 Decembe 2024
Accep ed 12 Decembe 2024
Jou nal o Economics, Finance and
Adminis a i e Science
Vol. 30 No. 59, 2025
pp. 189-204
Eme ald Publishing Limi ed
e-ISSN: 2218-0648
p-ISSN: 2077-1886
DOI 10.1108/JEFAS-12-2023-0280
1. In oduc ion
Fo decades, one o he mos widesp ead ideas in he inance li e a u e has been ha in es o s
a e a ional, and he e o e, s ock ma ke p ices should e lec he undamen al alue o u u e
cash lows. Howe e , se e al s udies sugges ha hese assump ions do no i he eal wo ld
and ha ma ke p ices a e indeed a ec ed by in es o emo ions and expec a ions (Lee e al.,
1991;Bake and Wu gle , 2006,2007). Mo eo e , a se ies o s udies indica e ha in es o
sen imen a ec s he capi al ma ke (B own and Cli , 2004;Bake and Wu gle , 2006,2007;
Bandopadhyaya and Jones, 2006;Yoshinaga and Cas o Junio , 2012;Co edo e al., 2013;
Fi h e al., 2015;Bu, 2020a,b;Bi encou and Iquiapaza, 2024).
In es o sen imen can be de ined as he expec a ions o asse e u ns ha hei
undamen als canno explain (Lee e al., 1991). I is also ela ed o pessimism o op imism
ega ding he s ock ma ke in gene al and he p opensi y o specula e (Bake and Wu gle ,
2006,2007). In his sense, sen imen has been widely ecognized o a ec he o e all inancial
ma ke (Cheong e al., 2017;Pa aboni e al., 2018). Ne e heless, only ecen s udies ha e
ocused on he possible e ec o in es o sen imen in he mu ual und indus y, speci ically, on
how i can a ec manage s’ s a egies and he pe o mance deli e ed o sha eholde s
(Bu, 2020a,b;Wang e al., 2020,2021).
In addi ion, mos o hese s udies show he in luence o sen imen on he Chinese and US
ma ke s. Recen ly, some esea che s ha e in es iga ed he es ima ion and in luence o he
in es o sen imen index on he B azilian ma ke (Yoshinaga and Cas o Junio , 2012;Xa ie and
Machado, 2017;Mi anda and Machado, 2018;San ana e al., 2020;Bi encou and Iquiapaza,
2024). Howe e , none o hese s udies ha e conside ed he mu ual und indus y, despi e i s
impo ance in he B azilian ma ke . Addi ionally, no wi hs anding his e idence, B azilian s udies
a e con inuing o p og ess owa d he de elopmen o he sen imen index and i s e ec s on he
ma ke (Xa ie and Machado, 2017;Mi anda and Machado, 2018;San ana e al., 2020).
In his sense, he objec i e o his s udy is o in es iga e he e ec o in es o sen imen on
mu ual und pe o mance (occu ence o und alpha) in he B azilian mu ual equi y und
indus y. Acco ding o benchma k models, in pa icula , he Capi al Asse P icing Model
(CAPM) and he Fama and F ench (1993) and Ca ha (1997) models, a posi i e alpha implies
ha he und has ou pe o med he ma ke , as alpha is measu ed by he di e ence be ween a
und’s ac ual e u n and i s expec ed pe o mance acco ding o a benchma k model (Jensen,
1968;Bu, 2020b). By de ini ion, a posi i e und alpha should occu only occasionally because o
he e icien ma ke hypo hesis, which s a es ha i is no possible o bea he ma ke consis en ly
(Fama, 1970). Howe e , posi i e und alphas a e egula ly documen ed (Bu, 2020a,b),
sugges ing ha some impo an ac o s migh be missing in s anda d benchma k models (Bu,
2020a). The e o e, in es o sen imen is used as one o he possible missing ac o s, ollowing
he app oach o Bu (2020b) and adap ed i o an eme ging ma ke , such as B azil.
Bake and Wu gle (2006,2007) s a e ha no pe ec p oxy o in es o sen imen exis s, bu
p opose a p ac ical app oach o i s es ima ion. This app oach was adap ed by Yoshinaga and
Cas o Junio (2012), and o he s udies p oposed an in es o sen imen index o B azil, which
was used in his s udy. Fo example, Mi anda e al. (2018) es ima e he in es o sen imen index
o B azil wi hou conside ing u no e because i is a ec ed by he equency o nego ia ions.
San ana e al. (2020), who a gue ha sen imen is associa ed wi h inancial decisions and
accoun ing choices, sugges ha i in luences i ms’ acc uals, which, in u n, in luences sha e
p icing in he B azilian ma ke . Consequen ly, discussions abou he e ec s o sen imen on he
B azilian ma ke ha e gained inc easing impo ance in esea ch on asse p icing models.
Equi y mu ual unds in B azil we e analyzed adop ing a compa a i e app oach and
inco po a ing da a om he US as a e e ence poin . Following he me hodology es ablished
by Bu (2020b), he analysis spans a decade, co e ing he pe iod om 2010 o 2019 o
B azilian unds and om 2010 o 2018 o US unds. The e minal da e o he US da a is
dic a ed by he las a ailable yea o he Bake and Wu gle sen imen index, which se es as a
c i ical me ic in his assessmen . The ob ained esul s we e he same as Bu (2020b) conce ning
he US; in o he wo ds, he sen imen index was ound o be a key ac o in cap u ing he
JEFAS
30,59
190
ou pe o mance p obabili y o equi y unds. Howe e , he main esul s o B azil sugges ha
benchma k models accoun ing o in es o sen imen as an addi ional ac o do no adequa ely
cap u e he ou pe o mance p obabili y o equi y mu ual unds. In addi ion, he esul s we e
consis en o es ima ing he B azilian sen imen index, al hough i is impo an o no e ha a
s anda dized se o a iables o his es ima ion has no ye been es ablished.
The indings unde sco e he impo ance o u he esea ch explo ing a iables ha p edic
in es o sen imen in eme ging ma ke s, as hey sugges he need o de elop mo e app op ia e
sen imen indices o hese ma ke s. This aligns wi h o he s udies ha ha e analyzed in es o
sen imen in he s ock ma ke s o de eloping coun ies (Anand e al., 2021;To e-To es
e al., 2021).
The emainde o his pape is o ganized as ollows: Sec ion 2 discusses he p e ious
s udies ha helped us achie e he s udy goals. Sec ion 3 desc ibes he me hodology used in his
s udy. Sec ion 4 p esen s he s udy esul s. Sec ion 5 p esen s he discussion, and Sec ion 6
o e s he conclusion o he pape .
2. Theo e ical ounda ions and ela ed li e a u e
This sec ion discusses s udies on in es o sen imen and asse p icing models in ela ion o
mu ual unds in bo h he B azilian and o e seas ma ke s.
2.1 In es o sen imen
To de e mine whe he he mu ual und indus y akes ma ke sen imen in o conside a ion,
Massa and Yada (2015) in es iga ed he mu ual und indus y in he US o e he pe iod
1984–2005. They a gued ha mu ual und manage s could use ma ke sen imen s a egically o
imp o e hei pe o mance and indi ec ly a ac mo e lows in o he und. Fu he mo e, hey
a gued ha unds wi h low exposu e o sen imen ended o ollow mo e pa icula in es men
s a egies and hose o ac i e und manage s, speci ically, bold and adi ional s a egies o
alloca ing und esou ces in esponse o op imism and pessimism exp essed by changes in
sen imen du ing he pe iod. Mo eo e , hei esul s showed ha he unds in which manage s
in es ed mo e—low-sen imen -be a s ocks—a ac ed mo e signi ican in es o lows.
Jiang and Yuksel (2019) s udied a sample o ac i ely managed US mu ual unds o e he
pe iod 1993–2014 o check o di e ences in he decision-making p ocess o in es o s
ega ding hei alloca ion o money in some mu ual unds, depending on sen imen pe iods
(high o low). The esul s showed ha he p e e ence o mu ual und in es o s o ce ain und
cha ac e is ics changed depending on he sen imen pe iod. They ound ha mu ual und
in es o s ended o pay mo e a en ion o pas pe o mance, expenses, and und isibili y when
sen imen was high. In addi ion, he e ec o sen imen was mo e p onounced among e ail
und in es o s han among ins i u ional in es o s.
By analyzing he ela ionship be ween in es o sen imen and mu ual und pe o mance,
Bu (2020a,b) discussed how in es o sen imen a ec ed a mu ual und’s alpha. He explained
ha he p obabili y o ou pe o ming unds inc eased wi h in es o sen imen and claimed ha
in es o sen imen was a missing ac o in he benchma k models ha esea che s ended o use
o es ima e und pe o mance.
Wang e al. (2021) obse ed how mu ual und manage s eac ed o ma ke sen imen in he
Chinese mu ual und indus y. They showed ha mu ual unds ollowed he sen imen -ca e ing
s a egy; ha is, hey a ac ed new lows in o unds based on in es o s’ en husiasm o popula
s ocks, and his did no necessa ily p o ide be e pe o mance o he und/in es o . Fu he mo e,
hese unds ended o incu g ea e isks when ebalancing hei po olios. Thus, he e is
e idence o an agency p oblem in mu ual unds, wi h manage s ac ing in hei own in e es s
a he han in line wi h he in es o s’ objec i es. San ana e al. (2020) poin ed ou ha his agency
p oblem has also occu ed in B azil. They in es iga ed he ela ionship be ween he sen imen
index and he disc e iona y acc uals o publicly aded companies and ound, speci ically in a
Jou nal o
Economics,
Finance and
Adminis a i e
Science
191
sepa a e analysis o pe iods o low and high sen imen , ha manage s ended o inc ease and
educe acc uals, espec i ely, a e changing he sen imen index o each ime cou se.
In ano he s udy, Wang e al. (2020), who also s udied Chinese mu ual unds be ween 2009
and 2016, unco e ed o he e idence on he sen imen s a egies o mu ual unds. The main
esul s showed a nega i e ela ionship among pe o mance, und isk- aking, and he
sen imen p oxy (FLOW). They a gued ha und manage s ended o minimize isk- aking
when sen imen was high and con ended ha he e was e idence ega ding he dumb money
e ec . The s udy’s indings indica e ha unds con ibu e o s abilizing he inancial ma ke
because en husiasm o in es ing in unds ends o in ensi y manage s’ isk a e sion, which
educes isk- aking.
2.2 Asse p icing models
Asse p icing models eme ged in he 1960s, s a ing wi h he Capi al Asse P icing Model
(CAPM). This model assesses he pe o mance o in es men asse s based on he ma ke isk.
This isk is he main explana o y ac o o he asse e u ns ela i e o he ma ke e u ns and
he e u ns o he isk- ee asse (Sha pe, 1964;Lin ne , 1965;Mossin, 1966). The CAPM
pos ula es a linea ela ionship be ween e u ns and sys emic isk and is based on he p emises
o in es o a ionali y and ma ke balance.
The CAPM was es ima ed using a simple linea eg ession ep esen ed by Equa ion (1).
The O dina y Leas Squa es (OLS) we e used, es ima ed om asse e u ns and he di e ence
be ween he a e age ma ke e u ns and he isk- ee asse e u ns. An asse ’s exposu e o
ma ke isk ep esen s he in es men p emium es ima ed acco ding o he ma ke isk.
Ri¼αþβi�Rm�R �(1)
As an al e na i e o he CAPM, in which he isk es ima ion p ocess is mul iple and composed
o mac oeconomic ac o s, Fama and F ench (1993) p oposed a h ee- ac o model. They a gue
ha , in addi ion o ma ke isk, we mus also conside he in luences o size (SMB) and he alue
measu ed in he book- o-ma ke (HML) on asse pe o mance, as shown in Equa ion (2). In
addi ion, o he e alua ion o in es men unds, Ca ha (1997) added a ou h ac o , he
momen ac o (MOM), o he h ee- ac o model. This app oach was based on G inbla and
Ti man (1993), who assessed he pe sis ence o abno mal und e u ns. This ac o exp esses he
pe sis ence o he e u n on asse s o e ime, and he model is shown in Equa ion (3).
Ri¼αþβi�Rm�R �þsiSMB þhiHML (2)
Ri¼αþβi�Rm�R �þsiSMB þhiHML þmiMOM:(3)
Recen ly, Fama and F ench (2015,2016) de ended a i e- ac o model. They a gued ha he
p icing p ocess should consis o he h ee- ac o model and he ac o s o in es men (RMW)
and p o i abili y (CMA), as shown in Equa ion (4). He e Riis he excess o a und’s e u ns
es ima ed by he di e ence be ween he e u ns o und iand he isk- ee asse , Ri−R .
O e all, s udies app aising B azilian und pe o mance equen ly use he CAPM and h ee-
and ou - ac o models o show he s a is ical signi icance o he e ec o hese ac o s on asse
pe o mance (Bo ges and Ma elanc, 2015;Ne as i and Lucinda, 2016;Maes i and
Malaquias, 2018;Fe nandes e al., 2018;Sil a e al., 2018,2020).
Ri¼αþβi�Rm�R �þsiSMB þþhiHML þ iRMW þciCMA (4)
JEFAS
30,59
192
3. Me hodology
The B azilian und sample comp ised 303 ac i e equi y mu ual unds. Ca ego y selec ion
ollowed he new classi ica ion o he B azilian Financial and Capi al Ma ke s Associa ion
(Associaç~
ao B asilei a das En idades dos Me cados Financei o e de Capi ais [ANBIMA]).
Indexed, o eign, and sec o al in es men unds we e no conside ed because o he
pa icula i ies o hese classes.
The sample pe iod spans om Janua y 2010 o Decembe 2019. Only unds wi h comple e
e u n da a we e conside ed, as did Bu (2020b). Fund da a we e collec ed om SI ANBIMA.
The in es o sen imen index and he isk ac o s o he p icing models we e es ima ed
mon hly o he B azilian ma ke . Da a o es ima ing he isk ac o s we e ob ained om
Economa ica, ollowing he p ocedu es ou lined by Fama and F ench (1993) and Ca ha
(1997). Fo he sen imen index, some da a we e collec ed om he B azilian s ock exchange
(B asil, Bolsa, Balc~
ao - B3) and he Secu i ies and Exchange Commission (Comiss~
ao de
Valo es Mobili�
a os [CVM]), while he emaining da a we e sou ced om Economa ica.
The B azilian in es o sen imen index was es ima ed using ou a iables: Numbe o
Ini ial Public O e ings (NIPO), Indi idual In es o Pa icipa ion (Pa In Ind), P emium o
Di idends (PDi ), and he High-Low Ra io (AD). NIPO was measu ed as he a e age o IPOs
and Follow-ons du ing he p e ious wel e mon hs. Pa In Ind e e s o he pe cen age o
indi idual in es o s’ pa icipa ion in B3’s mon hly ansac ions.
Fo PDi , companies we e classi ied as ei he di idend paye s o non-paye s based on whe he
hey paid di idends g ea e han ze o, as epo ed in hei inancial s a emen s. PDi was hen
calcula ed by aking he di e ence in he loga i hms o he ma ke - o-book a io o each g oup.
The AD is he a io o he highes o lowes s ock p ices o e he pas wel e mon hs. These
a iables we e selec ed based on p e ious s udies (Bake and Wu gle , 2006,2007;Yoshinaga
and Cas o Junio , 2012;Xa ie and Machado, 2017;Mi anda and Machado, 2018;San ana e al.,
2020). I is impo an o no e ha he discoun a e a iable, o iginally used in Bake and Wu gle ’s
index, was no included due o i s una ailabili y in B azil (Yoshinaga and Cas o Junio , 2012).
The u no e a iable was also excluded because o he high equency o ins i u ional in es o
ades, which, acco ding o Wu gle ’s websi e, no longe e lec s in es o sen imen .
In he es ima ion o sen imen , he a iables NIPO and PDi we e conside ed
con empo a y, and he a iables Pa In Ind, and AD lagged. This choice be ween he
a iable and i s lag depended on i s co ela ion wi h he i s componen o he p incipal
componen analysis (PCA), a me hod o es ima ing sen imen . In addi ion, he a iables we e
o hogonalized conce ning he mac oeconomic a iables (in la ion, GDP g ow h,
employmen le el, consump ion le el, and c ises). The a iables we e ob ained om he
B azilian Ins i u e o Geog aphy and S a is ics (IBGE), he Ins i u e o Applied Economic
Resea ch (IPEA), and he Na ional Bu eau o Economic Resea ch (NBER).
Fo o hogonaliza ion, he esiduals o he linea eg ession o each a iable o he
sen imen index wi h he se o mac oeconomic a iables we e used, a common p ocedu e in
he sen imen li e a u e (Bake and Wu gle , 2006,2007;Yoshinaga and Cas o Junio , 2012;
Xa ie and Machado, 2017;Mi anda and Machado, 2018;San ana e al., 2020). All a iables
we e alida ed using PCA, meaning only eigen alues g ea e han one we e alida ed
(Yoshinaga and Cas o Junio , 2012;Fi h e al., 2015;Xa ie and Machado, 2017;Mi anda
and Machado, 2018;San ana e al., 2020). In ac , he pe cen age o a ia ion explained by he
i s componen was 59%, highe han some s udies poin ed ou ; o example, Mi anda and
Machado (2018) calcula ed a p opo ion o 49%, like ha calcula ed by Yoshinaga and Cas o
J�
unio (2012). To es ima e he sen imen index, se en s eps we e ollowed, summa ized as
ollows:
(1) Es ima ion o a iables and lags, 12-mon h lagged a iables.
(2) Valida ion o he da abase s uc u e o PCA using Ba le ’s es and he Kaise -Meye -
Olkin (KMO) es .
Jou nal o
Economics,
Finance and
Adminis a i e
Science
193
(3) PCA and e alua ion o he co ela ion be ween he a iables and he i s p incipal
componen .
(4) Selec ion o con empo a y NIPO and PDi ; Pa In Ind, and AD lagged.
(5) O hogonaliza ion wi h he mac oeconomic a iables o in la ion, GDP g ow h, c isis,
and employmen and consump ion le els.
(6) Valida ion o new da a using Ba le ’s sphe ici y es and he Kaise -Meye -Olkin
(KMO) es .
(7) Es ima ion o he sen imen index using PCA and app op ia e a iables.
Fo da a alida ion, Ba le ’s and he KMO es s we e used, which a e complemen a y
echniques o mul i a ia e s a is ics ha ensu e he use o PCA and he use ulness o he esul s.
Ba le ’s es alida es he co ela ion s uc u e among he a iables, while he KMO index
e alua es he magni ude o his co ela ion (Johnson and Wiche n, 2007;Sh es ha, 2021;
Fonseca, 2022). P io o o hogonaliza ion, Ba le ’s es yielded a es s a is ic o 1,182.50 (wi h
ap- alue o ze o), indica ing he ejec ion o he null hypo hesis ha PCA would no be
app op ia e. The KMO es was 0.69, also conside ed alid, as alues close o one indica e
g ea e app op ia eness o he echnique o he da a. A e o hogonaliza ion, he es s we e
eapplied, and he esul s emained consis en : Ba le ’s es p oduced a s a is ic o 350.89 (wi h
ap- alue o ze o), and he KMO index was 0.60, main aining he p e iously iden i ied adequacy.
In his s udy, he CAPM and h ee- and ou - ac o models a e named CAPM, FF, and FFC,
espec i ely. They we e es ima ed wi h and wi hou including he Sen imen Index (Sen ) wi h
he ime se ies. He e oscedas ici y and au oco ela ion es s we e pe o med using esiduals,
and when necessa y, he esidual co a iance ma ix was co ec ed, as shown by MacKinnon
and Whi e (1985). The eg essions we e es ima ed o all pe iods and h ee subpe iods: Jan.
2010 o Dec. 2012, Jan. 2013 o Dec. 2015, and Jan. 2016 o Dec. 2019. The esul s se ed as a
basis o calcula ing he p obabili y o supe io und pe o mance (Bu, 2020b).
Ou pe o mance p obabili y e lec s he p obabili y ha he alphas a e posi i e and
signi ican in he models di ided by he alpha’s o al. Speci ically, he coe icien s o he
winne unds ( he winne unds included all unds wi h posi i e alphas ha we e s a is ically
signi ican a he 5% le el) we e e alua ed, sen imen was classi ied in o high and low
quin iles o analysis o he und alphas, and he models o each quin ile we e es ima ed.
To complemen he analysis o B azil, a sample o US mu ual unds was also collec ed o
compa ison, co e ing a pe iod simila o he B azilian unds. The US da a we e ga he ed as
ollows: mu ual und in o ma ion was sou ced om he Cen e o Resea ch in Secu i y P ices
(CRSP) da abase, speci ically he su i o -bias- ee (El on e al., 1996) mu ual und da abase.
The sample includes US equi y unds om he EDYG class. Addi ionally, due o he a ailable
yea s o he sen imen index, 2019 was no included. Fo he US sample, he in es o
sen imen index was ob ained om he Wu gle websi e (h p://people.s e n.nyu.edu/jwu gle /
), and he isk ac o s we e ob ained om he Kenne h F ench websi e (h ps://
mba. uck.da mou h.edu/pages/ acul y/ken. ench/da a_lib a y.h ml).
4. Empi ical indings
4.1 Analysis o desc ip i e s a is ics on und e u ns, sen imen index, and isk ac o s
Table 1 p esen s he desc ip i e s a is ics o he a iables used o es ima e he models. I was
obse ed ha he a e age and median alues o he measu ed s a is ics we e e y close. This
indica es symme y in he dis ibu ion o he alues. In addi ion, conside ing he coe icien o
a ia ion wi h his s anda dized measu e o dis ibu ion, i was obse ed ha he unds
p esen ed g ea a ia ion in mon hly e u ns h oughou he se ies. Simila ly, unds a e
p esen ed as mean and median posi i e e u ns in he sample pe iod. Fu he mo e, he mean
and median o he sen imen index we e nega i e, showing ha B azilian in es o s we e
gene ally pessimis ic abou he B azilian ma ke h oughou he pe iod 2010–2019.
JEFAS
30,59
194
Figu e 1 p esen s he mo emen s o he sen imen index and he mon hly ma ke excess
e u ns (RMRF). The sample pe iod was om Janua y 2010 o Decembe 2019. The sen imen
index was mul iplied by i e o acili a e compa ison.
Figu e 1 shows he mon hly dis ibu ions o he ma ke p emium and in es o sen imen
index in he B azilian ma ke . Acco ding o he O ganiza ion o Economic Co-ope a ion and
De elopmen (OECD), ecessions occu ed in he Ame ican ma ke om May 2012 o Feb ua y
2014, May 2015 o Ap il 2017, and Augus 2018 o Decembe 2019. These pe iods a e
associa ed wi h signi ican luc ua ions in he B azilian ma ke and, like Bu (2020b) a es s o he
US case, do no espond quickly o he a ia ions in he sen imen index. In he case o B azil, he
esponsi eness o he sen imen index o ma ke luc ua ions was de e mined o be oo low.
4.2 Pe o mance and e ec s o he sen imen index in he B azilian ma ke
Fi s , he ou pe o mance p obabili y o unds was es ima ed, ha is o say, he p obabili y o
achie ing a posi i e alpha ha is s a is ically signi ican a he 5% le el, using he s anda d
Table 1. Desc ip i e s a is ics o he a iables, B azil
Mean S anda d de ia ion Coe icien o a ia ion Minimum Median Maximum
Re u n 0.874 4.963 5.677 �33.60 0.743 33.05
Sen �0.440 1.122 �2.549 �2.980 �0.567 3.349
Risk ee 0.008 0.002 0.275 0.004 0.008 0.012
RMRF �0.002 0.057 �30.253 �0.126 �0.004 0.158
SMB �0.001 0.039 �58.946 �0.134 �0.004 0.120
HML �0.017 0.052 �3.045 �0.133 �0.017 0.122
Mom 0.020 0.050 2.487 �0.133 0.021 0.132
No e(s): Table 1 p esen s he desc ip o s o he a iables ha se ed as he basis o he measu emen o p icing
models. RMRF e e s o ðRm−R Þ. The und sample includes 306 mu ual unds. The sample pe iod anges om
Janua y 2010 h ough Decembe 2019 (mon hly e u n da a in %)
Sou ce(s): Au ho s’ own wo k
Figu e 1. Rela ionship be ween ma ke excess e u n (RMRF) and he sen imen index
Jou nal o
Economics,
Finance and
Adminis a i e
Science
195
benchma k models CAPM, Fama and F ench (1993) (FF), and Ca ha (1997) (FFC). The
sample was also di ided in o subpe iods o check o he e ec o he ma ke s a e. Table 2
p esen s he es esul s.
Ou esul s demons a ed ha a highe p obabili y o ou pe o ming occu ed when alphas
we e measu ed based on he CAPM model. As Table 2 shows, his p obabili y is 29.04%
agains he 14.52% o he 3- ac o model and 7.92% o he ou - ac o model o he en i e
pe iod. In con as , he ou pe o mance p obabili ies dec ease sha ply when he es ima ion o
alphas is based on he FFC model, sugges ing ha momen um is an impo an ac o o
conside o mu ual und pe o mance analysis, as no ed by o he au ho s (Ca ha , 1997;
Bo ges and Ma elanc, 2015;Ne as i and Lucinda, 2016;Maes i and Malaquias, 2018;
Fe nandes e al., 2018;Sil a e al., 2020).
By subpe iod, Table 2 shows ha he ou pe o mance p obabili ies a y om 27.06% o
6.60% using he CAPM, om 6.93% o 1.98% using he FF model, and om 3.96% o 1.65%
using he FFC model. These esul s indica e ha he FFC model bes explains und
pe o mance e en o sho e sample pe iods. Fu he mo e, he lowes ou pe o mance
p obabili y was in pe iod 3 (2016–2019). The e o e, all changes in ou pe o mance
p obabili ies ac oss pe iods indica e ha ma ke s a es migh play a ole in he p obabili y
o abno mal und e u ns occu ing.
Subsequen ly, he ou pe o mance p obabili ies o unds we e es ima ed using he s anda d
benchma k models (CAPM, FF, and FFC) adjus ed o in es o sen imen . The esul s
p esen ed in Table 3 sugges ha adding he sen imen index o a benchma k model can
subs an ially inc ease he ou pe o mance p obabili y, excluding he ou - ac o model in he
en i e pe iod, he CAPM model in he i s subpe iod, and all he models in he 2016–2019
subpe iod.
The pe iod om 2016 o 2019 was ma ked by g ea unce ain y in B azil. In 2016, P esiden
Dilma Rousse was impeached and se e al poli icians we e a es ed as a esul o Ope a ion
Table 2. Ou pe o ming p obabili ies based on s anda d benchma k models, B azil
CAPM (%) FF - 3 ac o model (%) FFC - 4 ac o model (%)
Whole pe iod 29.04 14.52 7.92
2010–2012 27.06 1.98 2.64
2013–2015 6.60 6.93 3.96
2016–2019 10.23 2.97 1.65
No e(s): Table 2 p esen s he ou pe o ming p obabili y o he sample unds based on he CAPM, Fama and
F ench (1993), and Ca ha (1997) models. The und sample includes 306 mu ual unds. The sample pe iod
anges om Janua y 2010 h ough Decembe 2019 (mon hly e u n da a)
Sou ce(s): Au ho s’ own wo k
Table 3. Ou pe o ming p obabili ies on sen imen -adjus ed models, B azil
CAPM þsen (%) FF - 3 ac o model þsen (%) FFC - 4 ac o model þsen (%)
Whole pe iod 33.66 15.18 5.94
2010–2012 16.50 3.30 3.96
2013–2015 8.25 15.51 9.57
2016–2019 3.63 2.31 1.65
No e(s): Table 3 p esen s he ou pe o ming p obabili y o he sample unds based on he CAPM, Fama and
F ench (1993), and Ca ha (1997) models adjus ed by in es o sen imen . The und sample includes 306 mu ual
unds. The sample pe iod anges om Janua y 2010 h ough Decembe 2019 (mon hly e u n da a)
Sou ce(s): Au ho s’ own wo k
JEFAS
30,59
196
Fe nandes, A.R.D.J., Fonseca, S.E. and Iquiapaza, R.A. (2018), “Modelos de mensu aç~
ao de
desempenho e sua in lu^
encia na cap aç~
ao l�
ıquida de undos de in es imen o”, Re is a
Con abilidade and Finanças, Vol. 29 No. 78, pp. 435-451, doi: 10.1590/1808-057x201805330.
Fi h, M., Wang, K. and Wong, S.M. (2015), “Co po a e anspa ency and he impac o in es o
sen imen on s ock p ices”, Managemen Science, Vol. 61 No. 7, pp. 1630-1647, doi: 10.1287/
mnsc.2014.1911.
Fonseca, S.E. (2022), “Fundos de in es imen o: ma ke iming, sen imen o do in es ido e ince eza da
pol�
ı ica econ^
omica”, Tese de dou o ado, Uni e sidade Fede al de Minas Ge ais, a ailable a :
h p://hdl.handle.ne /1843/50770
G inbla , M. and Ti man, S. (1993), “Pe o mance measu emen wi hou benchma ks: an examina ion
o mu ual und e u ns”, Jou nal o Business, Vol. 66 No. 1, pp. 47-68, doi: 10.1086/296593,
a ailable a : h p://www.js o .o g/s able/2353341
Hillie , D. and Loncan, T. (2019), “Poli ical unce ain y and s ock e u ns: e idence om he B azilian
poli ical c isis”, Paci ic-Basin Finance Jou nal, Vol. 54, pp. 1-12, ISSN 0927-538X, doi:
10.1016/j.pac in.2019.01.004.
Jensen, M.C. (1968), “The pe o mance o mu ual unds in he pe iod 1945-1964”, The Jou nal o
Finance, Vol. 23 No. 2, pp. 389-416, doi: 10.2307/2325404.
Jiang, G.J. and Y€
uksel, H.Z. (2019), “Sen imen al mu ual und lows”, Financial Re iew, Vol. 54 No. 4,
pp. 709-738, doi: 10.1111/ i e.12201.
Johnson, R.A. and Wiche n, D.W. (2007), Applied Mul i a ia e S a is ical Analysis, Pea son P en ice
Hall, Uppe Saddle Ri e , NJ.
Lee, C.M., Shlei e , A. and Thale , R.H. (1991), “In es o sen imen and he closed-end und puzzle”,
The Jou nal o Finance, Vol. 46 No. 1, pp. 75-109, doi: 10.1111/j.1540-6261.1991. b03746.x.
Lin ne , J. (1965), “Secu i y p ices, isk, and maximal gains om di e si ica ion”, The Jou nal o
Finance, Vol. 20 No. 4, pp. 587-615, doi: 10.2307/2977249.
MacKinnon, J.G. and Whi e, H. (1985), “Some he e oskedas ici y-consis en co a iance ma ix
es ima o s wi h imp o ed ini e sample p ope ies”, Jou nal o Econome ics, Vol. 29 No. 3,
pp. 305-325, doi: 10.1016/0304-4076(85)90158-7.
Maes i, C.O.N.M. and Malaquias, R.F. (2018), “Aspec s o manage , po olio alloca ion, and und
pe o mance in B azil”, Re is a Con abilidade and Finanças, Vol. 29 No. 76, pp. 82-96, doi:
10.1590/1808-057x201804590.
Massa, M. and Yada , V. (2015), “In es o sen imen and mu ual und s a egies”, Jou nal o Financial
and Quan i a i e Analysis, Vol. 50 No. 4, pp. 699-727, doi: 10.1017/s0022109015000253,
a ailable a : h ps://www.js o .o g/s able/43862270
Mi anda, K.F., Machado, M.A. and Macedo, L.A. (2018), “In es o sen imen and ea nings
managemen : does analys s’ moni o ing ma e ?”, RAM. Re is a de Adminis aç~
ao Mackenzie,
Vol. 19 No. 4, doi: 10.1590/1678-6971/e am 180104.
Mossin, J. (1966), “Equilib ium in a capi al asse ma ke ”, Econome ica: Jou nal o he Econome ic
Socie y, Vol. 34 No. 4, pp. 768-783, doi: 10.2307/1910098.
Ne as i, J.N. and Lucinda, C.R. (2016), “Pe sis ^
encia de desempenho em undos de aç~
oes no B asil”,
B azilian Re iew o Finance, Vol. 14 No. 2, pp. 269-297, doi: 10.12660/ b in. 14n2.2016.57958.
Pa aboni, A.L., Righi, M.B., Viei a, K.M. and Sil ei a, V.G. (2018), “The ela ionship be ween
sen imen and isk in inancial ma ke s”, BAR - B azilian Adminis a ion Re iew, Vol. 15 No. 1,
e170055, doi: 10.1590/1807-7692ba 2018170055.
San ana, C.V.S., San os, L.P.G.D., Ca alho J�
unio , C.V.D.O. and Ma inez, A.L. (2020), “In es o
sen imen and ea nings managemen in B azil”, Re is a Con abilidade and Finanças, Vol. 31
No. 83, pp. 283-301, doi: 10.1590/1808-057x201909130.
Sha pe, W.F. (1964), “Capi al asse p ices: a heo y o ma ke equilib ium unde condi ions o isk”,
The Jou nal o Finance, Vol. 19 No. 3, pp. 425-442, doi: 10.1111/j.1540-6261.1964. b02865.x.
Jou nal o
Economics,
Finance and
Adminis a i e
Science
203
Sh es ha, N. (2021), “Fac o analysis as a ool o su ey analysis”, Ame ican Jou nal o Applied
Ma hema ics and S a is ics, Vol. 9 No. 1, pp. 4-11, doi: 10.12691/ajams-9-1-2, a ailable a :
h ps://pubs.sciepub.com/ajams/9/1/2/index.h ml
Sil a, S.E., Roma, C.M.D.S. and Iquiapaza, R.A. (2018), “Does he managemen ee signal he
pe o mance o equi y in es men unds in B azil?”, Re is a de Educaç~
ao e Pesquisa em
Con abilidade, Vol. 12 No. 3, pp. 275-290, doi: 10.17524/ epec. 12i3.1717.
Sil a, S.E.D., Roma, C.M.D.S. and Iquiapaza, R.A. (2020), “Po olio u no e and pe o mance o
equi y in es men unds in B azil”, Re is a Con abilidade e Finanças, Vol. 31 No. 83, pp.
332-347, doi: 10.1590/1808-057x201909420.
To e-To es, O.V.D.L., Fige oa, E.G. and R�
ıo-Rama, M.D.L.C.D. (2021), “The Fama-F ench
mul i ac o model wi h ma ke and Pandemic news ea sen imen s: a es in he Mexican s ock
ma ke s”, Con adu �
ıa y Adminis aci�
on, Vol. 66 No. 5, p. 303.
Wang, J., Wang, X., Yang, J. and Zhuang, X. (2020), “Impac o in es o sen imen on mu ual und isk
aking and pe o mance: e idence om China”, En e p ise In o ma ion Sys ems, Vol. 14 No. 6,
pp. 833-857, doi: 10.1080/17517575.2020.1758795.
Wang, J., Yi, S., Wang, X., Yang, J. and Jiang, Z. (2021), “How do mu ual unds in China exploi
in es o sen imen ?”, Eme ging Ma ke s Finance and T ade, Vol. 57 No. 14, pp. 1-16, doi:
10.1080/1540496X.2020.1784715.
Xa ie , G.C. and Machado, M.A.V. (2017), “Anomalies and in es o sen imen : empi ical e idences in
he B azilian ma ke ”, BAR - B azilian Adminis a ion Re iew, Vol. 14No No. 3, e170028, doi:
10.1590/1807-7692ba 2017170028.
Yoshinaga, C.E. and Cas o Junio , F.H.F.D. (2012), “The ela ionship be ween ma ke sen imen index
and s ock a es o e u n: a panel da a analysis”, BAR - B azilian Adminis a ion Re iew, Vol. 9
No. 2, pp. 189-210, doi: 10.1590/S1807-76922012000200005.
Co esponding au ho
Sab ina Espinele da Sil a can be con ac ed a : [email p o ec ed]
Fo ins uc ions on how o o de ep in s o his a icle, please isi ou websi e:
www.eme aldg ouppublishing.com/licensing/ ep in s.h m
O con ac us o u he de ails: [email p o ec ed]
JEFAS
30,59
204
