Carbon emissions and crash risk: US evidence from firm-level data

Escola de Economia e Ges ão
Sa a Filipa Ribei o Rod igues
Ca bon Emissions and C ash Risk:
US e idence om i m-le el da a
dezemb o de 2024
Uni e sidade do Minho
Uni e sidade do Minho
Escola de Economia e Ges ão
Sa a Filipa Ribei o Rod igues
Ca bon Emissions and C ash Risk:
US e idence om i m-le el da a
Disse ação de Mes ado
Mes ado em Finanças
T abalho e e uado sob a o ien ação do
P o esso Dou o Nelson Manuel Pinho B andão
dezemb o de 2024
Cos a A eal
i
DIREITOS DE AUTOR E CONDIÇÕES DE UTILIZAÇÃO DO TRABALHO POR
TERCEIROS
Es e é um abalho académico que pode se u ilizado po e cei os desde que espei adas as eg as e boas
p á icas in e nacionalmen e acei es, no que conce ne aos di ei os de au o e di ei os conexos.
Assim, o p esen e abalho pode se u ilizado nos e mos p e is os na licença abaixo indicada.
Caso o u ilizado necessi e de pe missão pa a pode aze um uso do abalho em condições não p e is as no
licenciamen o indicado, de e á con ac a o au o , a a és do Reposi ó iUM da Uni e sidade do Minho.
Licenc

a concedida aos u ilizado es des e abalho
A ibuico-NoCome cial-SemDe i aces
CC BY-NC-ND
h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/
ii
Acknowledgmen s
Fi s and o emos , I would like o hank my supe iso , P o esso Nelson A eal, o b inging he weigh o his
conside able expe ience and knowledge o his disse a ion and o his guidance h oughou his jou ney.
A hea el hank you o my amily, especially my pa en s and my b o he , and close iends o hei lo e and
encou agemen .
A special hanks o my boy iend o his unwa e ing suppo and s imula ing discussions. His p esence has been
a signi ican sou ce o mo i a ion.
Thank you all.

iii
S a emen o In eg i y
I he eby decla e ha ing conduc ed his academic wo k wi h in eg i y. I con i m ha I ha e no used plagia ism
o any o m o undue use o in o ma ion o alsi ica ion o esul s along he p ocess leading o i s elabo a ion.
I u he decla e ha I ha e ully acknowledged he Code o E hical Conduc o he Uni e si y o Minho.
i
Resumo
Num mundo cada ez mais conscien e das ques ões climá icas, comp eende as implicações inancei as das
emissões de ca bono o nou-se c ucial. À medida que in es ido es e go e nado es lidam com iscos associados
a a o es ambien ais, é i al examina de que o ma as emissões de ca bono in luenciam a es abilidade
inancei a e a dinâmica dos me cados. Es a disse ação in es iga a elação en e as emissões de ca bono e o
isco de queda do p eço das ações a a és de uma análise de 1,279 emp esas lis adas nos EUA, no pe íodo
de 1999 a 2022. O es udo baseia-se no enquad amen o eó ico p opos o po Jin and Mye s (2006), que suge e
que a assime ia de in o mação le a a colapsos nos p eços das ações quando in o mações ocul as são
e eladas. Pa indo da p emissa de que emp esas com maio es emissões de ca bono epo adas são mais
anspa en es e genuínas, o obje i o cen al é a alia se es as emp esas es ão associadas a um meno isco de
colapso. Duas medidas de isco de queda são u ilizadas – Assime ia Condicional Nega i a (𝑁𝐶𝑆𝐾𝐸𝑊) e
Vola ilidade (𝐷𝑈𝑉𝑂𝐿) – pa a a alia es a elação. Os esul ados e elam uma associação não signi ica i a en e
as emissões de ca bono e o isco de queda, medido pela 𝑁𝐶𝑆𝐾𝐸𝑊 e pela 𝐷𝑈𝑉𝑂𝐿, após con ola pa a os
a o es que p e eem o isco de queda e pa a os e ei os ixos de ano e emp esa. Es udos u u os de e iam
expandi a análise pa a uma amos a mais ampla e di e en es con ex os geog á icos de o ma a e i a
compa ações e esul ados mais obus os.
Pala as-cha e: Assime ia Condicional Nega i a, Assime ia de In o mação, Emissões de Ca bono, Risco de
Queda do P eço das Ações, Vola ilidade.
Abs ac
In oday’s inc easingly clima e-conscious wo ld, unde s anding he inancial implica ions o ca bon emissions
has become c ucial. As in es o s and policymake s g apple wi h he isks associa ed wi h en i onmen al ac o s,
i is i al o examine how ca bon emissions in luence inancial s abili y and ma ke dynamics. This disse a ion
in es iga es he ela ionship be ween ca bon emissions and s ock p ice c ash isk h ough an analysis o 1,279
U.S. lis ed i ms o e he pe iod om 1999 o 2022. The s udy is g ounded in he heo e ical amewo k
p oposed by Jin and Mye s (2006), which sugges s ha in o ma ion asymme y leads o s ock p ice c ashes
when hidden in o ma ion is e ealed. Based on he p emise ha companies wi h highe epo ed ca bon
emissions a e mo e anspa en and genuine, he main pu pose is o e alua e whe he hese companies a e
associa ed wi h a lowe isk o s ock p ice collapse. Two measu es o c ash isk – Nega i e Condi ional Skewness
(𝑁𝐶𝑆𝐾𝐸𝑊) and Down- o-Up Vola ili y (𝐷𝑈𝑉𝑂𝐿) – a e employed o assess his ela ionship. The indings e eal
no s a is ically signi ican associa ion be ween ca bon emissions and c ash isk as measu ed by 𝑁𝐶𝑆𝐾𝐸𝑊 and
𝐷𝑈𝑉𝑂𝐿, a e con olling o p edic o s ac o s o c ash isk and yea , and i m ixed e ec s. Fu he s udies
should ex end he analysis o a b oade sample and di e en geog aphical con ex s o enhance compa a i e
insigh s.
Keywo ds: Ca bon Emissions, Down- o-Up Vola ili y, In o ma ion Asymme y, Nega i e Condi ional Skewness,
S ock P ice C ash Risk.
i
Lis o Con en s
1. In oduc ion ................................................................................................................................... 1
2. Li e a u e Re iew ........................................................................................................................... 3
3. Me hodology .................................................................................................................................. 5
3.1. The Sample................................................................................................................................. 5
3.2. Ca bon Emissions Measu e ......................................................................................................... 5
3.3. C ash Risk Measu es ................................................................................................................... 5
3.4. Con ol Va iables ......................................................................................................................... 7
3.5. Empi ical Model .......................................................................................................................... 9
4. Da a ............................................................................................................................................10
4.1. Ini ial S ock Da a .......................................................................................................................10
4.2. Da a Fil e s ...............................................................................................................................10
4.2.1. S ock il e s based on s a ic in o ma ion .............................................................................11
4.2.2. S ock and s ockday il e s based on e u n index in o ma ion ..............................................13
4.2.3. S ock il e s based on ca bon emissions measu e and indus y classi ica ion ......................17
4.3. Financial Accoun ing Va iables ...................................................................................................17
4.3.1. Sha e Tu no e ..................................................................................................................19
4.3.2. Book Value o Equi y ..........................................................................................................19
5. Empi ical Resul s .........................................................................................................................20
5.1. Desc ip i e S a is ics .................................................................................................................20
5.2. Pea son Co ela ion Ma ix ........................................................................................................23
5.3. Findings ....................................................................................................................................26
6. Conclusions .................................................................................................................................28
Re e ences............................................................................................................................................29
Appendices ...........................................................................................................................................33
A. Da a Va iables Desc ip ion: Da as eam Da a ypes ...................................................................33
B. Da a Va iables Desc ip ion: ESG Da a ypes ..............................................................................35
C. Da a Va iables Desc ip ion: Wo ldscope Da a ypes ...................................................................36
D. Indus y Classi ica ion o Sample Fi m .....................................................................................38
5
3. Me hodology
3.1. The Sample
My sample comp ises 1,279 lis ed i ms in he Uni ed S a es spanning om Decembe 1999 o Decembe
2022, encompassing bo h ac i ely aded s ocks and s ocks delis ed du ing he analysis pe iod. The sample
and ime-pe iod selec ion will be explo ed u he in Sec ion 4. Appendix D displays he b eakdown o sample
i ms by indus y, e ealing a p edominan p esence o companies in he Business Se ices indus y
(15.95%), wi h Elec onic Equipmen (6.96%), Re ail (6.33%), and Pe oleum and Na u al Gas (6.10%)
ollowing closely behind.
3.2. Ca bon Emissions Measu e
To e ec i ely measu e ca bon emissions, i is c ucial o acqui e a comp ehensi e unde s anding o he
di e se ca ego ies o emissions a company may p oduce, along wi h me hods o no malize his me ic. This
no maliza ion is essen ial o ensu e uni o mi y and compa abili y ac oss he di e en companies wi hin my
sample.
Acco ding o he Uni ed S a es En i onmen al P o ec ion Agency (EPA), scope 1 emissions comp ise di ec
emissions om sou ces unde he o ganiza ion’s con ol. Scope 2 includes indi ec emissions linked o he
acquisi ion o elec ici y, s eam, hea , o cooling se ices. Con e sely, scope 3 emissions ex end beyond
di ec con ol o owne ship, encompassing emissions om ac i i ies associa ed wi h asse s no owned o
con olled by he epo ing en i y bu indi ec ly in luenced wi hin i s alue chain.
Building upon exis ing li e a u e and acknowledging he limi ed a ailabili y o da a in Da as eam, his s udy
will na ow i s ocus o scope 1 ca bon emissions, ecognized as an indispensable pa o co po a e ca bon
esponsibili y and managemen . These emissions will be scaled by he i m’s ne sales o e enues a he
end o each yea , p o iding insigh s in o he ca bon emissions (in ons o di ec CO2 and CO2 equi alen )
pe US dolla s o e enues o ne sales o each i m (Aljughaiman e al., 2024; Qian & Schal egge , 2017).
The a iable 𝐶𝐴𝑅𝐵𝑂𝑁 will se e his pu pose in he empi ical model, enabling o unde s and he
ela ionship be ween he le el o ca bon emissions and he isk o a s ock p ice c ash.
3.3. C ash Risk Measu es
Following he p io li e a u e (Chen e al., 2001; Jin & Mye s, 2006), wo measu es o i m-speci ic c ash
isk will be employed. These measu es a e based on i m-speci ic daily e u ns, 𝑅𝑗,𝜏. This ensu es ha ou
c ash isk measu es indica e i m-speci ic ac o s ins ead o b oad ma ke mo emen s (Kim e al., 2014).
𝑅𝑗,𝜏 is compu ed as he na u al loga i hmic o one plus he esidual e u n, 𝑅𝑗,𝜏 =ln(1 + 𝜀𝑗,𝜏), o he
ollowing ex ended ma ke model:

6
𝑟
𝑗,𝜏 = 𝛼𝑗 + 𝛽1,𝑗𝑟𝑚,𝜏−2 + 𝛽2,𝑗𝑟𝑚,𝜏−1 + 𝛽3,𝑗𝑟𝑚,𝜏 + 𝛽4,𝑗𝑟𝑚,𝜏+1 + 𝛽5,𝑗𝑟𝑚,𝜏+2 + 𝜀𝑗,𝜏
(1)
whe e,
𝑟
𝑗,𝜏 = e u n on s ock 𝑗 on day 𝜏,
𝑟𝑚,𝜏 = e u n on he alue-weigh ed ma ke po olio 𝑚 on day 𝜏.
The e u n on s ock 𝑗 on day 𝜏,
𝑟
𝑗,𝜏, is compu ed based on he daily changes in he To al Re u n Index (RI)
1
alue e ie ed om Da as eam.
The e u n on he alue-weigh ed ma ke po olio 𝑚 on day 𝜏, is sou ced om he Kenne h R. F ench
Lib a y and p o ides he alue-weigh ed e u n o all CRSP (Cen e o Resea ch in Secu i y P ices) i ms
inco po a ed in he US and lis ed on he NYSE, AMEX, o NASDAQ (Fama & F ench, 2015).
The lagged and leading ma ke e ms a e included o allow o non-synch onous ading as di e en s ocks
ha e di e en ading equencies (Dimson, 1979).
The esiduals, 𝜀𝑗,𝜏, o he ex ended ma ke model a e no e enly dis ibu ed and a e skewed. By compu ing
he i m-speci ic daily e u ns, 𝑅𝑗,𝜏, one can easily iden i y bo h c ashes and posi i e jumps in he da a
symme ically as he ans o ma ion helps o e en ou he dis ibu ion (Hu on e al., 2009).
The i s measu e o c ash isk is he Nega i e Condi ional Skewness, 𝑁𝐶𝑆𝐾𝐸𝑊, which con ines he
asymme y o he e u n dis ibu ion (Hunj a e al., 2020). I is calcula ed by aking he nega i e o he hi d
momen o each s ock’s i m-speci ic daily e u ns o each yea and no malizing i by he cubed s anda d
de ia ion o i m-speci ic daily e u ns. Thus, o each i m j in yea ,
𝑁𝐶𝑆𝐾𝐸𝑊
𝑗,𝑡 = −[𝑛(𝑛 − 1)3/2 ∑𝑅3𝑗,𝜏] /[(𝑛 − 1)(𝑛 − 2)(∑𝑅2𝑗,𝜏)3/2]
(2)
whe e,
𝑅𝑗,𝜏 = i m-speci ic daily e u n o i m 𝑗 on day 𝜏,
𝑛 = numbe o daily e u ns du ing yea 𝑡.
This measu e is mul iplied by -1 so ha a highe alue co esponds o a s ock being mo e c ash p one, i.e.,
ha ing a mo e le -skewed dis ibu ion.
1
In Appendix A, a de ailed desc ip ion o RI is gi en.
7
The second measu e o c ash isk is he Down- o-Up Vola ili y, 𝐷𝑈𝑉𝑂𝐿. Fo e e y s ock 𝑗 o e a iscal-yea
pe iod 𝑡, i m-speci ic daily e u ns a e sepa a ed in o “down” days, when e u ns a e below he pe iod
mean, and “up” days, when e u ns a e abo e he pe iod mean. The s anda d de ia ion is compu ed o
each o hese subsamples sepa a ely and hen we ake he log o he a io o he s anda d de ia ion on he
down days o he s anda d de ia ion on he up days. Speci ically, 𝐷𝑈𝑉𝑂𝐿 is calcula ed as ollows:
𝐷𝑈𝑉𝑂𝐿𝑗,𝑡 =𝑙𝑛[(𝑛𝑢− 1)∑𝑅2𝑗,𝜏/(𝑛𝑑− 1)∑𝑅2𝑗,𝜏𝑈𝑝𝐷𝑜𝑤𝑛 ]
(3)
whe e,
𝑛𝑢= numbe o up days in yea ,
𝑛𝑑= numbe o down days in yea .
As sugges ed in Chen e al. (2001), his al e na i e measu e does no in ol e hi d momen s, and hence is
less likely o be o e ly a ec ed by ex eme e u ns. Again, he con en ion is ha a highe alue o 𝐷𝑈𝑉𝑂𝐿
indica es g ea e c ash isk.
3.4. Con ol Va iables
Fo his esea ch, I decided o employ se e al con ol a iables which esea che s ha e p e iously ound o
ha e an in luence on u u e c ash isk. Chen e al. (2001) showed ha ading olume, a p oxy o he
in ensi y o di e ences o opinion among in es o s, is a p edic o o s ock p ice c ash isk. This is explained
by he Hong and S ein (1999) model
2
which p edic s ha nega i e skewness in e u ns will be mos
p onounced a ound pe iods o hea y ading olume. Following hei baseline speci ica ion, I used de ended
u no e , 𝐷𝑇𝑈𝑅𝑁, o accoun o his impac by compu ing he a e age mon hly sha e u no e in yea 𝑡
minus he a e age mon hly sha e u no e in yea 𝑡 − 1 (Kim e al., 2014; Yildiz & Ka an, 2020). The
eason o de ending is o emo e any u no e componen ha can be conside ed as a ela i ely s able
i m cha ac e is ic, adhe ing o a conse a i e app oach. De ending is hen able o cap u e he in ensi y o
disag eemen s in he ma ke (Mu a a & Hamo i, 2021).
Chen e al. (2001) disco e ed ha his o ical e u ns and ma ke - o-book a ios also play a ole in p edic ing
he isk o ma ke c ashes. These ela ionships a e pe haps mos clea ly sugges ed by models o s ochas ic
bubbles in which high pas e u ns o high ma ke - o-book a ios (glamou s ocks) sugges a p olonged
buildup o he bubble. Consequen ly, when he bubble bu s s and p ices e e o undamen al le els, he
2
The Hong-S ein model elies on di e ences o opinion among in es o s ega ding undamen al alue and he p esence o sho -sales cons ain s
o some. When bea ish in es o s wi h cons ain s a e o ced o sell all sha es due o di e ences o opinion, hei in o ma ion is no ully inco po a ed
in o p ices, c ea ing a co ne solu ion. As mo e-bullish in es o s exi he ma ke , he ini ially bea ish g oup may become suppo buye s, e ealing
mo e abou hei signals. This p ocess esul s in hidden in o ma ion coming o ligh du ing ma ke declines, leading o nega i ely skewed e u ns.
8
decline is mo e p onounced. This heigh ened decline is he unde lying eason o bo h being an icipa ed o
ha e a highe isk o c ashing. I hus con ol o pas e u ns, 𝑅𝐸𝑇, calcula ed as he mean o i m-speci ic
daily e u ns o e he iscal yea imes 100, and o he ma ke - o-book a io, 𝑀𝐵, calcula ed as he ma ke
alue o equi y di ided by he book alue o equi y (Kim e al., 2014).
Fu he mo e, i has been well documen ed in se e al s udies (Chen e al., 2001; Ha ey & Siddique, 2000)
ha skewness is mo e nega i e on a e age o la ge-cap i ms. Speci ically, Chen e al. (2001) de eloped a
disc e iona y-disclosu e hypo hesis g ounded in he no ions ha (i) manage s end o p omp ly e eal
posi i e news bu g adually elease nega i e ones, and (ii) manage s o small companies can hide bad news
mo e easily using his app oach. Hence, we con ol o i m size, 𝑆𝐼𝑍𝐸, calcula ed as he na u al loga i hm
o he ma ke alue o equi y (Kim e al., 2014; Mu a a & Hamo i, 2021).
The nex con ol a iable is s ock ola ili y, 𝑆𝐼𝐺𝑀𝐴, as mo e ola ile s ocks a e likely o be mo e c ash
p one. Founda ional a icles (Chen e al., 2001; Hu on e al., 2009) a e based on a “ ola ili y eedback”
mechanism in which he elease o good news, while posi i e, in oduces ma ke ola ili y, and bad news,
in addi ion o i s di ec nega i e impac , is u he magni ied by an inc eased isk p emium. Howe e , some
s udies ha e ound he opposi e ela ionship be ween s ock ola ili y and c ash isk (Mu a a & Hamo i,
2021).
𝑆𝐼𝐺𝑀𝐴 is calcula ed as he s anda d de ia ion o i m-speci ic daily e u ns o e he iscal yea .
In addi ion, I con ol o inancial le e age, 𝐿𝐸𝑉, calcula ed as o al long- e m deb di ided by o al asse s,
and o p o i abili y measu ed by e u n on asse s, 𝑅𝑂𝐴, (Kim e al., 2014). P io li e a u e used le e age
as a p oxy o de aul isk bu ailed o ind suppo o his p oposi ion (Habib e al., 2018). Fo ins ance,
Hu on e al. (2009) and Kim e al. (2011a, 2011b) ind a nega i e associa ion be ween le e age and c ash
isk, when le e age is expec ed o be posi i ely co ela ed wi h bank up cy isk (Campbell e al., 2008).
Acco ding o Habib e al. (2018), one possible explana ion o his unexpec ed inding is ha highly le e aged
i ms migh ini ially be unde alued by in es o s, educing he likelihood o subsequen p ice c ashes. As
p e iously men ioned, glamou s ocks a e conside ed mo e c ash-p one, he e o e a posi i e ela ionship is
expec ed be ween p o i abili y, 𝑅𝑂𝐴, and c ash isk.
9
3.5. Empi ical Model
Following he empi ical model de eloped by (Kim e al., 2014), in o de o examine he impac o ca bon
emissions on u u e s ock p ice c ash isk, he ollowing eg ession model will be es ima ed:
𝐶𝑅𝐴𝑆𝐻_𝑅𝐼𝑆𝐾𝑗,𝑡 = 𝛽0+ 𝛽1(𝐶𝐴𝑅𝐵𝑂𝑁𝑗,𝑡−1)+ 𝛽2(𝐷𝑇𝑈𝑅𝑁𝑗,𝑡−1)
+ 𝛽3(𝑅𝐸𝑇
𝑗,𝑡−1) + 𝛽4(𝑀𝐵𝑗,𝑡−1) + 𝛽5(𝑆𝐼𝑍𝐸𝑗,𝑡−1)
+𝛽6(𝑆𝐼𝐺𝑀𝐴𝑗,𝑡−1) + 𝛽7(𝐿𝐸𝑉
𝑗,𝑡−1)+ 𝛽8(𝑅𝑂𝐴𝑗,𝑡−1)
+ 𝛽𝑥(𝐷𝑌𝑒𝑎𝑟)+ 𝛽𝑦(𝐷𝐹𝑖𝑟𝑚) + 𝜀𝑗,𝑡 (4)
In Equa ion 4, 𝐶𝑅𝐴𝑆𝐻_𝑅𝐼𝑆𝐾 is he dependen a iable and is a p oxy o 𝑁𝐶𝑆𝐾𝐸𝑊 o
𝐷𝑈𝑉𝑂𝐿.
𝐶𝐴𝑅𝐵𝑂𝑁 is he main independen a iable as i con ains he ca bon emissions measu e. All independen
a iables a e lagged by one yea so ha we can es i he ca bon emissions le el o i m 𝑗 in yea 𝑡–1,
𝐶𝐴𝑅𝐵𝑂𝑁𝑗,𝑡−1, p edic c ash isk in yea 𝑡, 𝐶𝑅𝐴𝑆𝐻_𝑅𝐼𝑆𝐾𝑗,𝑡.
The p e iously ou lined con ol a iables se e he pu pose o accoun ing o i m-speci ic ac o s an icipa ed
o in luence he likelihood o u u e s ock p ice c ashes. The e o e, o each i m 𝑗 in yea 𝑡–1, we
inco po a e con ols o changes in ading olume, 𝐷𝑇𝑈𝑅𝑁𝑗,𝑡−1; pas e u ns, 𝑅𝐸𝑇
𝑗,𝑡−1; ma ke - o-book
a io, 𝑀𝐵𝑗,𝑡−1; i m size, 𝑆𝐼𝑍𝐸𝑗,𝑡−1; s ock ola ili y, 𝑆𝐼𝐺𝑀𝐴𝑗,𝑡−1; inancial le e age, 𝐿𝐸𝑉
𝑗,𝑡−1; and e u n
on asse s, 𝑅𝑂𝐴𝑗,𝑡−1, (Kim e al., 2014; Yildiz & Ka an, 2020). All hese i m-speci ic ac o s, including i m-
speci ic daily e u ns, 𝑅𝑗,𝜏, a e winso ized a he 1s and 99 h pe cen iles, a widely used echnique in he
li e a u e o mi iga e he impac o ou lie s (Zaman e al., 2021). The eg ession model also includes yea
and i m ixed e ec s o accoun o unobse ed he e ogenei y.
10
4. Da a
The Re ini i Eikon Da as eam pla o m se es as he p ima y sou ce o da a in his disse a ion. In
subsec ion 4.1, de ailed insigh s will be p o ided ega ding he ini ial s ock uni e se, while in subsec ion 4.2
I will p esen he subsequen applica ion o il e s o a i e a he inal sample. Adhe ing o he guidelines
ou lined by Landis and Skou as (2021), pa icula ly add essing s a ic and e u n index in o ma ion, was
impe a i e o enhance da a accu acy and mi iga e su i o ship bias. These guidelines, ailo ed o he US
ma ke , ensu e a mo e nuanced app oach o wo king wi h Da as eam da a, he eby con ibu ing o he
o e all quali y o he esul s. Finally, subsec ion 4.3 will del e in o a comp ehensi e p esen a ion o all i m-
ela ed in o ma ion.
4.1. Ini ial S ock Da a
To c ea e he ini ial s ock uni e se, while Landis and Skou as (2021) ad oca ed o an al e na i e app oach,
which in ol ed ex ac ing in o ma ion on all equi ies aded ac oss e e y s ock exchange wi hin each coun y
using TDS’ Na iga o GUI, I decided no o ollow his me hod due o i s ime-consuming na u e and
limi a ions in da a a ailabili y om he Da as eam pla o m. Ins ead, I op ed by elying on ma ke -speci ic
cons i uen lis s p o ided by Da as eam and Wo ldscope, an app oach commonly employed by esea ch
eams wo king wi h da a om hese sou ces, as i is mo e e icien and aligns wi h es ablished p ac ices in
he ield (Ka olyi e al., 2012; Schmid e al., 2019).
Acco dingly, di e en cons i uen lis s we e ex ac ed and me ged, including: (i) he FUSALL* lis s,
comp ising all cu en ly ac i e s ocks, (ii) he DEADUS* lis s, consis ing o delis ed s ocks, and (iii) he
WSUS* lis s p o ided by Wo ldscope. Wo h men ioning ha he s ock uni e se should include no only
s ocks ha we e ac i e a he ime o he s udy, bu also s ocks ha we e inac i e due o me ge s,
acquisi ions, o ailu es, bu show a po ion o he p ice his o y o e he en i e analysis pe iod (Landis &
Skou as, 2021). Employing his me hod yielded a da ase comp ising 116,834 s ocks.
4.2. Da a Fil e s
In his subsec ion, I will p o ide a comp ehensi e o e iew o he il e ing p ocess employed o add ess da a
challenges wi hin he ini ial sample.
The il e ing s a egies encompass wo main ca ego ies: s ock il e s, which in ol e excluding he en i e
his o ical eco d o an ins umen , and s ockday il e s, which a ge speci ic days o indi idual s ocks. These
il e s a e subca ego ized based on he ype o in o ma ion hey use, including s a ic da a and e u n indexes.
Conside ing he ime-consuming na u e o applying all il e s p o ided by Landis and Skou as (2021), I op ed
o selec hose iden i ied as mos essen ial when wo king wi h Da as eam da a and mos adap able o my
esea ch. The au ho s p o ide use s wi h he lexibili y o apply hei il e s in he manne hey deem

11
app op ia e, emphasizing ha hei wo k aims o es ablish bes p ac ices o wo king wi h hese da a bu
acknowledging ha he e is no “sou ce o u h”. This app oach unde sco es he impo ance o exe cising
judgemen in il e selec ion and implemen a ion.
A he end o his subsec ion, I illus a e he need o emo ing speci ic s ocks based on ca bon emissions
da a a ailabili y and indus y classi ica ion.
4.2.1. S ock il e s based on s a ic in o ma ion
Fo each s ock on each cons i uen lis , he ollowing a iables we e e ie ed (Oli ei a, 2023): Type o
Ins umen (TYPE), Da as eam Code (DSCD), Base Da e (BDATE), Expanded Secu i y Name (ENAME),
Da as eam Exchange Mnemonic (EXMNEM), Geog aphical Classi ica ion (GEOGN), ISIN (ISIN), P ima y
Indica o Flag (ISINID), Code Local (LOC), Cu ency (PCUR) and Secu i y Type Code (TRAC). In Appendix A,
he desc ip ion o hese a iables is gi en. Table 1E o ! Re e ence sou ce no ound. p esen s a desc ip ion
o he il e s, ha will be u he explained, o ganized in he o de in which hey a e applied in his
Disse a ion.
Fi s ly, I ocused exclusi ely on ins umen s classi ied as equi ies, ensu ing ha all obse a ions ep esen ed
equi y secu i ies (whe e he da a ype TYPE should be “EQ”). This il e is pe inen because cons i uen lis s
o en include a ious ypes o secu i ies, such as Ame ican Deposi o y Receip s (ADR), Closed-end Funds
(CF), and Global Deposi o y Receip s (GDR), which could o he wise in oduce noise o inaccu acies in o he
analysis.
Nex , I na owed down he da ase o s ocks ha a e aded on US exchanges (EXMNEM). Acco ding o
Landis and Skou as (2021), some esea che s p e e o exclude s ocks lis ed on seconda y exchanges
because hey assume hese s ocks a e likely o be small o aded o e - he-coun e (OTC). Howe e , using
his app oach wi h Da as eam da a can lead o a p oblem as his da abase only p o ides in o ma ion on
he cu en exchange whe e a s ock is lis ed. So, i esea che s exclude seconda y exchanges o a oid issues
ela ed o small s ocks, hey migh also exclude s ocks ha go mo ed om main s ock exchanges due o
poo pe o mance, which can c ea e bias in he sample. Following Oli ei a (2023), he US has 12 exchanges
plus he OTC ma ke s: NYSE (NYS), NYSE MKT (ASE), NYSE ARCA (XC), NASDAQ (NAS), NASDAQ/NMS
(NMS), OTC Bulle in (XBQ), Non-Nasdaq OTC (OTC), Bos on (BOS), Chicago (MID), Paci ic (PSE),
Philadelphia (PHL), and BARS (E1). The e o e, I will only conside s ocks aded on hese exchanges.
To e ine he selec ion u he , I included only ins umen s ha could be classi ied as common s ocks based
on bo h he TRAC and ENAME a iables. As no ed by Landis and Skou as (2021), he classi ica ion o a
s ock as common emains consis en o e ime, hus using his il e helps p e en su i o ship bias. Column
3 o Table 1 p o ides in o ma ion o he alid secu i y ype codes and coun y-speci ic ex s ings ha should
12
no be p esen in he ex ended names o s ocks. Addi ionally, s ocks wi h expanded names con aining
coun y-speci ic local lis ing iden i ie s we e excluded, indica ing po en ial lis ings in o he coun ies. This
addi ional il e aids in iden i ying c oss-lis ed s ocks and ensu es hei exclusion om he sample.
Mo eo e , I ensu ed ha he da ase only consis ed o companies domiciled in he Uni ed S a es (GEOGN)
and secu i ies aded in he US dolla (PCUR).
To p e en any duplica es om a ec ing my analysis, I emo ed any edundan en ies based on bo h LOC
and DSCD codes. The da a ype LOC ep esen s a code assigned o an ins umen by he exchange whe e
i is aded. When Da as eam designa es a s ock as p ima y (da a ype ISIND equals “P”), he au ho s
no iced ha all o he s ocks wi h he same local code a e no common s ocks. The e o e, i is ecommended
o exclude any s ocks wi h a non-unique local code and whose ISINID is no “P”, as long as he e is a leas
one s ock wi h his local code ha does ha e ISINID = “P”.
A e applying hese il e s, he inal sample consis ed o 29,968 obse a ions.
Table 1: Fil e s based on s a ic in o ma ion.
This able ou lines all he il e s applied o he s a ic a iables lis ed in he i s column. I explains he a ionale o
each il e and speci ies he accep ed alues used in he sample cons uc ion.
Va iable
Pu pose
Accep ed Values
Type o Ins umen (TYPE)
To gua an ee ha all obse a ions
a e Equi ies.
‘EQ’.
Da aS eam Exchange
Mnemonic (EXMNEM)
To exclude s ocks ha a e no
aded on US exchanges.
‘NAS’, ‘NYS’, ‘OTC’, ‘ASE’, ‘XSQ’, ‘XBQ’,
‘NMS’, ‘BOS’, ‘MID’, ‘PSE’, ‘PHL’, ‘E1’.
Secu i y Type Code
(TRAC)
To exclude s ocks ha canno be
classi ied as common s ocks.
‘ORD’, ‘ORDSUBR’, ‘FULLPAID’, ‘UNKNOWN’,
‘UNKNOW’, ‘KNOW’, ‘NA’.
Expanded Secu i y Name
(ENAME)
To exclude s ocks iden i iable as
non-common s ocks based on
speci ic ex s ings wi hin hei
ex ended names.
Do no con ain:
'TRUST','REPR','RIGHT','SERIES','NV','IV
TST','REAL ESTATE
INVESTMENT','REALTY','RLTY','ROYALTY
INVESTMENT','ASSET INVESTMENT','CAPITAL
INVESTMENT','ASSET
MANAGEMENT','CAPITAL
MANAGEMENT','INVESTMENT
MANAGEMENT','VENTURE
CAPITAL','FINANCIAL SHBI','PROPERTY
INVESTORS','INCOME
PROPERTY','UNITS','UNIT','LIMITED
PARTENERSHIP','FUND','EQUITY
PARTNERS','LIMITED VOTING','SUB
VOTING','TIER ONE SUB','VARIABLE
13
VOTING','NON VOTINGREIT','RESIDENTIAL','R
E I T','BENEFICIAL','BENEFICIARY','BENEFIT
INTEREST','BEN INTEREST','SH BEN
INT','WARRANT','WRTS','L P','L P
INTEREST','LP UT','HOLDINGS
LP','PARTENERS UNIT','PART INT','UNIT
PARTENERSHIP','UNIT
LIMITED','MORTGAGE','REAL
ESTATE','CERTIFICATE','NO PAR
VALUE','HOLDING UNIT','BACKED','ST
MIN','CORTS','TORPS','TOPRS','SECURITIES
TRUPS','QUIPS','STRATS HIGH YIELD','TOTAL
RETURN','DIVERSIFIED
HOLDINGS','(SICAV)','DEPOSITARY',
'DEPOSITOR','RECEIPT','REP &
SHARES','GLOBAL
SHARES','ADR','GDR','EXPD.','EXPIRED','DUPLI
CATE','CONVERTIBLE','CNVRT.','CONVRT.','EX
CH.','DEBANTURE','(DEB)','NIL
PAID','STRUCTURED
ASSET','CALLABLE','FLOATING
RATE','ADJUSTABLE','REDEEMABLE','PAIRED
CTF','CONSOLIDATED','INSURED','CAPITAL
SHARES','DEBT
STRATEGIES','LIQUIDATING','LIQUID UNIT','L
UNIT','- LASD','ACQUISITION','CAP
UNIT','INCOME
UNIT','PREFERRED','(NYS)','(NAS)','(ASE)','(OT
C)','(XSQ)','(XQB)'.
Geog aphical
Classi ica ion o company
(GEOGN)
To exclude s ocks ha a e no
domicilia ed in he US.
‘UNITED STATES’.
Cu ency (PCUR)
To emo e all secu i ies aded in
a cu ency o he han he US
dolla .
‘U$’.
Code Local (LOC) and
P ima y Indica o Flag
(ISINID)
To emo e duplica ed
obse a ions and ensu e he
exclusion o any non-common
s ocks om he sample.
No duplica ed codes. I duplica es a e ound,
emo e hem only i ISINID is no equal o ‘P’,
p o ided ha a leas one s ock wi h his LOC
has ISINID equal o ‘P’.
Da as eam Code (DSCD)
To emo e duplica ed
obse a ions.
No duplica ed codes.
4.2.2. S ock and s ockday il e s based on e u n index in o ma ion
Fo he 29,968 s ocks p e iously ound, I ex ac ed he To al Re u n Index (da a ype RI) daily om Decembe
1999 o Decembe 2022. The choice o his ime ame is unde pinned by wo p incipal conside a ions.
14
Fi s ly, da a eliabili y o US s ocks signi ican ly imp o es a e Decembe 1984, a poin emphasized by
Landis and Skou as (2021) o ensu ing da ase obus ness. Howe e , due o limi a ions in Da as eam and
he eme gence o ca bon emissions disclosu e om companies in he la e 1990s and ea ly 2000s, I op ed
o begin da a ex ac ion om Decembe 1999. Secondly, o include he mos ecen and pe inen
in o ma ion, I ex ended he da ase o co e da a up un il he end o 2022.
Fo he da a ex ac ion p ocess, I used he Da as eam’s DPL unc ion, ensu ing p ecision by e aining up
o six decimal poin s, which is he maximum a ailable numbe o decimal places in he da abase. I is
impo an o no e ha Da as eam does no p o ide di ec e u ns bu ins ead o e s a e u n index ha
moni o s he hypo he ical alue o an in es men wi h ein es ed cash lows, no ably di idends.
Consequen ly, e en small alues in his index can signi ican ly a ec e u n calcula ions due o ounding
issues, emphasizing he impo ance o his s ep.
Fo ce ain s ocks, he e u n index is una ailable o any da e, and as is s anda d in he li e a u e, hese
ins umen s we e consequen ly excluded.
Fo he i s il e ing s ep, I ob ained in o ma ion ega ding each s ock’s s a us (da a ype ESTAT) and i s
co esponding delis ing da e (da a ype TIME). When a company’s s ock is delis ed, i means i is no longe
ac i ely aded on a public exchange. Ini ially, ce ain s ocks we e ound o ha e been delis ed p io o
Decembe 1999, and as a esul , hey we e p omp ly excluded om he sample. Fo s ocks ha unde wen
delis ing a e his da e, i is impo an o ecognize ha he e can be a delay be ween he las da e o which
ading da a is a ailable and he o icial da e when he s ock is ma ked as delis ed in he da abase.
Consequen ly, many s ocks exhibi cons an e u n indexes owa d he conclusion o hei ading his o y
e en when he se ies ha e been unca ed a hei delis ing da es. To add ess his conce n, he i s s ep
in ol ed he emo al o da a ela ed o s ocks showing cons an alues beyond hei delis ing da es.
Addi ionally, a compa ison was made be ween he da e o he las non-ze o e u n and he delis ing da e. I
a di e ence o mo e han 10 days was iden i ied, he en h and subsequen daily obse a ions we e
excluded; o he wise, he delis ing da e e ie ed om Da as eam was used as he end da e o he se ies.
S ocks wi h cons an e u n indexes h oughou he en i e se ies we e also excluded o ensu e da a
eliabili y
3
.
In he second phase o he il e ing p ocess, I s a ed by compu ing he daily pe cen change in he To al
Re u n Index o each day 𝜏, o daily e u n (%), wi hin he ime se ies, aking in o accoun he indi idual
s ock’s ini ial and end da es.
3
Please no e ha I ha e applied he condi ion o excluding s ocks wi h cons an e u n indexes o bo h dead and ac i e companies in my sample.
This ensu ed ha subsequen il e s we e applied smoo hly, a oiding any po en ial issues.
21
he nega i e impac o he Russian-Uk aine wa on he global inancial ma ke which began in Feb ua y
2022 (Assa e al., 2023).
As o he 𝐶𝐴𝑅𝐵𝑂𝑁𝑡−1, i shows ela i ely s able alues o e he yea s, indica ing a g adual inc ease in
epo ing and po en ially in emissions hemsel es, especially a e 2015. This can be linked o he epo ing
obliga ions es ablished by he Pa is Ag eemen in Decembe 2015. Unde i s Enhanced T anspa ency
F amewo k, all Pa ies a e manda ed o epo hei g eenhouse gas (GHG) balance biannually and o
moni o he p og ess o indi idual coun ies in achie ing hei mi iga ion a ge s (Pe ugini e al., 2021).
Ne e heless, a epo de eloped by As You Sow
7
in 2022, e alua ed he e o o 55 majo US co po a ions
in educing GHG emissions in alignmen wi h he Pa is Ag eemen ’s a ge o limi ing global empe a u e
ise o 1.5 deg ees Celsius, aiming o ne ze o emissions by 2050. They ound ha mos companies s ill
lack comp ehensi e 1.5 deg ee-aligned GHG educ ion goals and a e no making su icien p og ess owa d
ne ze o emissions. Mo eo e , e en hough mos companies epo Scope 1 emissions, disclosu e o ca bon
o se s is o en unclea , which unde sco es he c i ical need o exe cising cau ion while looking a ca bon
emissions igu es o e ime. Mo eo e , he e does no appea o be any speci ic pa e n be ween he lagged
ca bon emissions and c ash isk a iables.
Table 4: Mean alues o c ash isk and lagged ca bon emissions measu es (2000-2022).
This able p esen s he a e age alues o c ash isk, 𝑁𝐶𝑆𝐾𝐸𝑊𝑡 and 𝐷𝑈𝑉𝑂𝐿𝑡, and lagged ca bon emissions,
𝐶𝐴𝑅𝐵𝑂𝑁𝑡−1, measu es om 2000 o 2022. Re e o Sec ion 3 o a iables de ini ion.
𝐘𝐄𝐀𝐑𝐭
𝑵𝑪𝑺𝑲𝑬𝑾𝒕
𝑫𝑼𝑽𝑶𝑳𝒕
𝑪𝑨𝑹𝑩𝑶𝑵𝒕−𝟏
2000
-0.2261
0.0028
-
2001
-0.1430
0.0018
-
2002
-0.0125
0.0016
-
2003
-0.1627
0.0007
0.2062
2004
-0.1923
0.0005
0.1496
2005
-0.1801
0.0004
0.1608
2006
-0.1777
0.0003
0.1357
2007
-0.1932
0.0005
0.1240
2008
0.0042
0.0016
0.1294
2009
-0.2204
0.0013
0.1417
2010
-0.1881
0.0004
0.1777
7
As You Sow is a non-p o i o ganiza ion ha p omo es en i onmen al and social co po a e esponsibili y h ough sha eholde ad ocacy, coali ion
building, and inno a i e legal s a egies. I engages in a a ie y o ini ia i es aimed a encou aging companies o adop mo e sus ainable and socially
esponsible p ac ices. These ini ia i es include ad oca ing o co po a e anspa ency, educing GHG emissions, p omo ing ai labo p ac ices, and
minimizing en i onmen al impac s.

22
2011
-0.1036
0.0004
0.1467
2012
-0.0989
0.0003
0.1289
2013
-0.1080
0.0002
0.1291
2014
-0.0657
0.0003
0.1427
2015
-0.0042
0.0004
0.1426
2016
-0.0332
0.0007
0.1614
2017
0.0978
0.0003
0.1786
2018
0.1318
0.0006
0.1912
2019
0.0836
0.0004
0.1915
2020
-0.0719
0.0018
0.1804
2021
-0.0772
0.0007
0.2048
2022
0.0395
0.0006
0.1601
Table 5 p esen s he desc ip i e s a is ics o he sample. Building on he insigh s p o ided ea lie , he mean
alues o c ash isk measu es, 𝑁𝐶𝑆𝐾𝐸𝑊𝑡 and 𝐷𝑈𝑉𝑂𝐿𝑡, a e -0.0724 and 0.0008, espec i ely. These
alues sugges ha , on a e age, he sample i ms ha e mo e igh -skewed i m-speci ic daily e u ns and
sligh ly highe ola ili y in i m-speci ic e u ns on down days compa ed o up days, al hough he di e ence
is small. The medians (-0.0557 o 𝑁𝐶𝑆𝐾𝐸𝑊𝑡 and 0.0001 o 𝐷𝑈𝑉𝑂𝐿𝑡) a e sligh ly lowe han he
espec i e means, indica ing he p esence o a ew ela i ely high alues ha aise he mean, while mos
da a poin s a e clus e ed owa ds he lowe end o he ange. Fo 𝑁𝐶𝑆𝐾𝐸𝑊𝑡, he es ima es a e simila o
hose ound by Chen e al. (2001) and Yildiz and Ka an (2020) bu di e om Kim e al. (2014) and Mu a a
and Hamo i (2021). Fo 𝐷𝑈𝑉𝑂𝐿𝑡, he mean is e y close o ze o, which is lowe han he a e age epo ed
in he exis ing li e a u e, likely due o di e ences in sample pe iods and sample cons uc ion.
The a e age alue o he o al ca bon emission pe US dolla s o e enues o ne sales is 0.1682. The
a e age change in mon hly ading olume (as a pe cen age o sha es ou s anding) is 0.0214. Fu he mo e,
he a e age i m in he sample has a i m-speci ic daily e u n o 13.27%, a ma ke capi aliza ion o $3.24
billion
8
, a ma ke - o-book a io o 0.1650, a daily e u n ola ili y o 0.0209, a le e age o 0.2643, and a
e u n on asse s o 0.0490. 𝑅𝐸𝑇𝑡−1 and
𝑆𝐼𝑍𝐸𝑡−1 ha e he highes s anda d de ia ions, sugges ing (i) high
a iabili y in i m-speci ic daily e u ns, and (ii) di e se se o i ms in he sample om small-cap o la ge-
cap companies.
8
To in e p e he mean alue o he 𝑆𝐼𝑍𝐸 a iable in e ms o ac ual ma ke alue, one mus e e se he log ans o ma ion by exponen ia ing he
mean alue.
23
Table 5: Summa y o desc ip i e s a is ics.
This able p esen s he desc ip i e s a is ics o he a iables used in he disse a ion. The da a comp ises a sample o
1,279 US-lis ed s ocks om 1999 o 2022, including bo h ac i e s ocks and hose ha ha e been delis ed bu e ain
a po ion o hei p ice his o y h oughou he analysis pe iod. Financials (SIC codes 6000-6999) and u ili ies (SIC
codes 4900-4949) a e excluded, based on he Fama-F ench 48 indus y classi ica ions. Re e o Sec ion 3 o a iable
de ini ions.
VARIABLES
MEAN
MEDIAN
STANDARD
DEVIATION
25TH
PERCENTILE
75TH
PERCENTILE
𝑫𝑼𝑽𝑶𝑳𝒕
0.0008
0.0001
0.0030
0.0000
0.0002
𝑵𝑪𝑺𝑲𝑬𝑾𝒕
-0.0724
-0.0557
0.5432
-0.3805
0.2596
𝑪𝑨𝑹𝑩𝑶𝑵𝒕−𝟏
0.1682
0.0132
0.4194
0.0026
0.0848
𝑫𝑻𝑼𝑹𝑵𝒕−𝟏
0.0214
0.0077
0.5948
-0.2087
0.2344
𝑹𝑬𝑻𝒕−𝟏
0.1327
-0.0211
0.8778
-0.0956
0.0543
𝑴𝑩𝒕−𝟏
0.1650
0.1137
0.1662
0.0629
0.2030
𝑺𝑰𝒁𝑬𝒕−𝟏
8.0842
8.0577
1.7676
6.9209
9.2774
𝑺𝑰𝑮𝑴𝑨𝒕−𝟏
0.0209
0.0173
0.0128
0.0124
0.0250
𝑳𝑬𝑽𝒕−𝟏
0.2643
0.2376
0.1855
0.1323
0.3646
𝑹𝑶𝑨𝒕−𝟏
0.0490
0.0636
0.1199
0.0248
0.1035
5.2. Pea son Co ela ion Ma ix
Table 6 epo s he esul s o he Pea son co ela ion ma ix. The highes co ela ions a e obse ed be ween
𝐷𝑈𝑉𝑂𝐿 and 𝑆𝐼𝐺𝑀𝐴 (0.73), and 𝐷𝑈𝑉𝑂𝐿 and 𝑅𝐸𝑇 (0.58), indica ing ha s ocks wi h g ea e ola ili y
and highe pas e u ns a e posi i ely co ela ed wi h inc eased c ash isk, as measu ed by 𝐷𝑈𝑉𝑂𝐿.
Con e sely, he e is a nega i e co ela ion be ween 𝑁𝐶𝑆𝐾𝐸𝑊 and 𝑆𝐼𝐺𝑀𝐴 (-0.22), and 𝑁𝐶𝑆𝐾𝐸𝑊 and
𝑅𝐸𝑇 (-0.43) wi hin my sample. Addi ionally, he e is a s ong nega i e co ela ion be ween 𝑆𝐼𝐺𝑀𝐴 and
bo h 𝑆𝐼𝑍𝐸 (-0.53) and 𝑅𝑂𝐴 (-0.42), indica ing ha smalle i ms and hose wi h lowe p o i abili y end o
expe ience highe ola ili y in i m-speci ic daily e u ns.
Fu he mo e, he posi i e and s a is ically signi ican co ela ion be ween 𝐶𝐴𝑅𝐵𝑂𝑁 and 𝐷𝑈𝑉𝑂𝐿 indica es
a weak posi i e ela ionship (0.1356), sugges ing ha i ms wi h highe ca bon emissions a e associa ed
wi h sligh ly highe down- o-up ola ili y. Con e sely, he nega i e and s a is ically signi ican co ela ion
be ween 𝐶𝐴𝑅𝐵𝑂𝑁 and
𝑁𝐶𝑆𝐾𝐸𝑊
e eals a e y weak nega i e ela ionship, in con as o he posi i e
ela ionship ound wi h 𝐷𝑈𝑉𝑂𝐿, indica ing ha highe ca bon emissions migh be associa ed wi h sligh ly
24
less nega i e skewness in s ock e u ns. Howe e , gi en he weak na u e o hese ela ionships, d awing
obus conclusions is challenging.
25
COLUMN1
𝑫𝑼𝑽𝑶𝑳
𝑵𝑪𝑺𝑲𝑬𝑾
𝑪𝑨𝑹𝑩𝑶𝑵
𝑫𝑻𝑼𝑹𝑵
𝑹𝑬𝑻
𝑴𝑩
𝑺𝑰𝒁𝑬
𝑺𝑰𝑮𝑴𝑨
𝑳𝑬𝑽
𝑹𝑶𝑨
𝑫𝑼𝑽𝑶𝑳
1.0000
𝑵𝑪𝑺𝑲𝑬𝑾
-0.2128***
1.0000
𝑪𝑨𝑹𝑩𝑶𝑵
0.1356***
-0.0407***
1.0000
𝑫𝑻𝑼𝑹𝑵
0.1062***
0.0250***
0.0314***
1.0000
𝑹𝑬𝑻
0.5764***
-0.4294***
0.0363***
-0.0171**
1.0000
𝑴𝑩
0.0300***
-0.0719***
-0.1902***
0.0059
0.0525***
1.0000
𝑺𝑰𝒁𝑬
-0.3222***
0.1142***
-0.1875***
-0.0268***
-0.1399***
0.2407***
1.0000
𝑺𝑰𝑮𝑴𝑨
0.7272***
-0.2182***
0.2176***
0.1624***
0.3701***
-0.0161**
-0.5325***
1.0000
𝑳𝑬𝑽
0.0400***
0.0059
0.0968***
0.0216***
0.0060
-0.1026***
-0.0050
0.0567***
1.0000
𝑹𝑶𝑨
-0.3061***
0.0518***
-0.1198***
-0.0160**
-0.0734***
0.1162***
0.3108***
-0.4177***
-0.0729***
1.0000
Table 6: Pea son Co ela ion Ma ix.
This able p esen s he Pea son co ela ion coe icien s o he a iables used in his disse a ion. The da a comp ises a sample o 1,279 US-lis ed s ocks om
2000 o 2022, including bo h ac i e s ocks and hose ha ha e been delis ed bu e ain a po ion o hei p ice his o y h oughou he analysis pe iod.
Financials (SIC codes 6000-6999) and u ili ies (SIC codes 4900-4949) a e excluded, based on he Fama-F ench 48 indus y classi ica ions. Re e o Sec ion
3 o a iable de ini ions. No e: *, **, and *** indica e signi icance a he 10%, 5% and 1% le els, espec i ely.
26
5.3. Findings
Table 7 p esen s he OLS es ima ion esul s, wi h 𝑁𝐶𝑆𝐾𝐸𝑊𝑡 as he dependen a iable in Panel A, and
𝐷𝑈𝑉𝑂𝐿𝑡 as he dependen a iable in Panel B. Bo h eg essions include ixed e ec s a he yea and i m
le els.
Examining he esul s p esen ed in Panel A, he model shows a nega i e associa ion be ween he le el o
ca bon emissions pe US dolla s o e enues o ne sales, 𝐶𝐴𝑅𝐵𝑂𝑁𝑡−1, and he isk o a s ock p ice c ash
in he upcoming yea , measu ed by 𝑁𝐶𝑆𝐾𝐸𝑊𝑡. This ela ionship seems o be aligned wi h he ini ial
expec a ion ha being mo e genuine and anspa en in epo ing would associa e wi h lowe c ash isk.
Howe e , his ela ionship did no achie e s a is ical signi icance, meaning ha he e is insu icien e idence
o conclude a meaning ul associa ion be ween hese a iables wi hin he con ex o he s udy. The only
a iable ound o be s a is ically signi ican a he 1% le el is 𝑆𝐼𝑍𝐸𝑡−1 wi h a coe icien o 0.3091. This
posi i e and signi ican coe icien sugges s ha la ge i ms a e mo e likely o expe ience highe
𝑁𝐶𝑆𝐾𝐸𝑊𝑡, indica ing an inc eased isk o a s ock p ice c ash. This inding suppo s he disc e iona y-
disclosu e hypo hesis p oposed by Chen e al. (2001), which posi s ha la ge-cap i ms end o exhibi mo e
nega i e skewness on a e age. This ela ionship is u he co obo a ed by ecen s udies (Hunj a e al.,
2020; Kim e al., 2014).
Tu ning o he esul s in Panel B, we obse e a posi i e associa ion be ween he le el o ca bon emissions
pe US dolla s o e enues o ne sales, 𝐶𝐴𝑅𝐵𝑂𝑁𝑡−1, and he isk o a s ock p ice c ash in he upcoming
yea , 𝐷𝑈𝑉𝑂𝐿𝑡. Howe e , his associa ion also lacks s a is ical signi icance. Despi e his, he posi i e
ela ionship appea s coun e in ui i e o my analysis. This disc epancy sugges s ha 𝐷𝑈𝑉𝑂𝐿𝑡 migh be
cap u ing a di e en dimension o ca bon emissions o e ime. This indica es a need o in es iga e whe he
he ela ionship be ween emissions and c ash isk is mo e complex han ini ially assumed. Addi ionally, he
model iden i ies 𝑅𝐸𝑇𝑡−1 (0.0009***) and 𝑀𝐵𝑡−1 (0.0009*) as p edic o s o one-yea ahead c ash isk,
wi h bo h showing posi i e and s a is ically signi ican ela ionships a he 1% and 10% le els, espec i ely.
These associa ions seem o be aligned wi h he heo y behind models o s ochas ic bubbles in which high
pas e u ns and high ma ke - o-book a ios (glamou s ocks) a e associa ed wi h highe isk o s ock p ice
c ash (Chen e al., 2001). Vola ili y, 𝑆𝐼𝐺𝑀𝐴𝑡−1 (0.1475***) and inancial le e age, 𝐿𝐸𝑉𝑡−1, (-0.0005*)
a e also ound o ha e a s a is ically signi ican impac on he isk o a s ock p ice c ash. These esul s a e
also in line wi h he li e a u e, which sugges s ha high ola ili y and lowe le e aged i ms a e linked o
inc eased c ash isk (Chen e al., 2001; Hu on e al., 2009; Kim e al., 2011a, 2011b).

27
Table 7: Ca bon Emissions on C ash Risk - Reg ession Resul s.
This able p esen s he eg ession esul s o he e ec o ca bon emissions on i m-le el s ock p ice c ash isk. The
s anda d e o s clus e ed a he i m and yea le els a e epo ed in pa en heses. Re e o Sec ion 3 o a iable
de ini ions. No e: *, **, and *** indica e signi icance a he 10%, 5% and 1% le els, espec i ely.
DEPENDENT VAR. =
𝑵𝑪𝑺𝑲𝑬𝑾𝒕
PANEL A
𝑫𝑼𝑽𝑶𝑳𝒕
PANEL B
𝑪𝑨𝑹𝑩𝑶𝑵𝒕−𝟏
-0.0633
(0.0606)
0.0005
(0.0003)
𝑫𝑻𝑼𝑹𝑵𝒕−𝟏
-0.0254
(0.0191)
0.0001
(0.0001)
𝑹𝑬𝑻𝒕−𝟏
0.0226
(0.0019)
0.0009***
(0.0003)
𝑴𝑩𝒕−𝟏
0.2876
(0.1841)
0.0009*
(0.0005)
𝑺𝑰𝒁𝑬𝒕−𝟏
0.3091***
(0.0416)
-0.0001
(0.0001)
𝑺𝑰𝑮𝑴𝑨𝒕−𝟏
-1.9829
(2.4056)
0.1475***
(0.0308)
𝑳𝑬𝑽𝒕−𝟏
0.0126
(0.1038)
-0.0005*
(0.0003)
𝑹𝑶𝑨𝒕−𝟏
-0.1247
(0.1881)
-0.0001
(0.0007)
CONSTANT
-0.0701
(0.0961)
-0.0021**
(0.0006)
YEAR FE
Yes
Yes
COMPANY FE
Yes
Yes
OBSERVATIONS
5581
5581
R-SQUARED
0.302
0.751
28
6. Conclusions
In his Disse a ion, he aim was o examine he e ec s o ca bon emissions on he one-yea ahead c ash
isk o a sample o 1,279 US-lis ed i ms. The esea ch was based on he model de eloped by Jin and
Mye s (2006), which a gues ha s ock p ice c ashes esul om in o ma ion asymme y. When a ce ain
h eshold is eached, all he hidden in o ma ion comes simul aneously, leading o he collapse in s ock
p ices. Two measu es o c ash isk we e employed: he Nega i e Condi ional Skewness, 𝑁𝐶𝑆𝐾𝐸𝑊, which
cap u es he asymme y o he e u n dis ibu ion, and he Down- o-Up Vola ili y, 𝐷𝑈𝑉𝑂𝐿, which assesses
he asymme y in s ock e u n ola ili y be ween pe iods o posi i e and nega i e e u ns ela i e o a s ock’s
a e age e u n.
The indings indica e no s a is ically signi ican ela ionship be ween ca bon emissions, 𝐶𝐴𝑅𝐵𝑂𝑁𝑡−1, and
one-yea ahead c ash isk measu ed by bo h 𝑁𝐶𝑆𝐾𝐸𝑊𝑡 and 𝐷𝑈𝑉𝑂𝐿𝑡. Rega ding he a iables o con ol,
despi e some indings lacking s a is ical signi icance, he analysis iden i ied s a is ically signi ican
ela ionships be ween c ash isk and pas e u ns, 𝑅𝐸𝑇𝑡−1, ola ili y, 𝑆𝐼𝐺𝑀𝐴𝑡−1, i m size, 𝑆𝐼𝑍𝐸𝑡−1,
ma ke - o-book a io, 𝑀𝐵𝑡−1, and inancial le e age, 𝐿𝐸𝑉𝑡−1, in line wi h he li e a u e.
I is impo an o acknowledge ha his s udy encoun e ed se e al da a a ailabili y challenges, pa icula ly
ega ding ca bon emissions. Following he me hodology o Landis and Skou as (2021), signi ican ime and
e o we e equi ed o compile an accu a e sample. Due o hese cons ain s, di ec compa isons wi h
exis ing li e a u e may no be s aigh o wa d. Mos p io s udies used di e en da abases ha p o ided
easie access o mo e comp ehensi e da ase s, which we e no a ailable o his analysis. These limi a ions
unde sco e he need o imp o ed da a accessibili y and consis ency in u u e esea ch o acili a e mo e
obus compa isons and conclusions.
Fo u u e esea ch, i would be aluable o conduc u he in es iga ion in o he a ious measu es used o
compu e c ash isk and explo e how hey migh cap u e di e en in o ma ion du ing he same e en o ime
pe iod. Expanding he analysis o a la ge sample o i ms and including companies om di e se coun ies
could p o ide mo e meaning ul compa isons and a iche unde s anding o he ela ionship be ween c ash
isk and ca bon emissions in di e en con ex s. Addi ionally, explo ing whe he companies wi h highe
ca bon emissions expe ience lowe in o ma ion asymme y compa ed o hose wi h lowe emissions could
o e aluable insigh s in o how his cha ac e is ic in luence c ash isk dynamics. This u he in es iga ion
may e eal nuances in how ca bon emissions impac in o ma ion anspa ency and, consequen ly, s ock
p ice s abili y.
29
Re e ences
Adhika i, A., & Zhou, H. (2021). Volun a y disclosu e and in o ma ion asymme y: do in es o s in US capi al
ma ke s ca e abou ca bon emission?
Sus ainabili y Accoun ing, Managemen and Policy Jou nal
,
13(1)
, 195-220.
Aljughaiman, A. A., Cao, N. D., & Alba ak, M. S. (2024). The impac o g eenhouse gas emission on
co po a e’s ail isk.
Jou nal o Sus ainable Finance & In es men
,
14 (1)
, 68-85.
Assa , R., Gup a, D., & Kuma , R. (2023). The p ice o wa : E ec o he Russia-Uk aine wa on he global
inancial ma ke .
The Jou nal o Economic Asymme ies
,
28
, e00328.
Bis line, J., Blan o d, G., B own, M., Bu aw, D., Domeshek, M., Fa bes, J.,…Jones, R. (2023). Emissions
and ene gy impac s o he In la ion Reduc ion Ac .
Science
,
380
(6652), 1324-1327.
Bol on, P., & Kacpe czyk, M. (2021). Do in es o s ca e abou ca bon isk?
Jou nal o inancial economics
,
142
(2), 517-549.
Bo ghei, Z., Leung, P., & Gu h ie, J. (2018). Does olun a y g eenhouse gas emissions disclosu e educe
in o ma ion asymme y? Aus alian e idence.
A o-Asian Jou nal o Finance and Accoun ing
,
8
(2),
123-147.
Campbell, J. Y., Hilsche , J., & Szilagyi, J. (2008). In sea ch o dis ess isk.
The Jou nal o inance
,
63
(6),
2899-2939.
Capasso, G., Gian a e, G., & Spinelli, M. (2020). Clima e change and c edi isk.
Jou nal o Cleane
P oduc ion
,
266
, 121634.
Ca agnano, A., Ma iani, M., Pizzu ilo, F., & Zi o, M. (2020). Is i wo h educing GHG emissions? Explo ing
he e ec on he cos o deb inancing.
Jou nal o En i onmen al Managemen
,
270
, 110860.
Chen, J., Hong, H., & S ein, J. C. (2001). Fo ecas ing c ashes: ading olume, pas e u ns, and condi ional
skewness in s ock p ices.
Jou nal o Financial Economics
,
61
(3), 345-345-381.
Ch is o e sen, P. (2012).
Elemen s o inancial isk managemen
(2nd ed.). Academic p ess.
da Sil a, P. P. (2022). C ash isk and ESG disclosu e.
Bo sa Is anbul Re iew
,
22(4)
, 794-811.
Dai, P.-F., Xiong, X., Liu, Z., Huynh, T. L. D., & Sun, J. (2021). P e en ing c ash in s ock ma ke : The ole
o economic policy unce ain y du ing COVID-19.
Financial Inno a ion
,
7
, 1-15.
30
Delbeke, J., Runge-Me zge , A., Slingenbe g, Y., & We ksman, J. (2019). The pa is ag eemen . In
Towa ds
a clima e-neu al Eu ope
(pp. 24-45). Rou ledge.
Dimson, E. (1979). Risk measu emen when sha es a e subjec o in equen ading.
Jou nal o inancial
economics
,
7
(2), 197-226.
Fama, E. F., & F ench, K. R. (2006). P o i abili y, in es men and a e age e u ns.
Jou nal o inancial
economics
,
82
(3), 491-518.
Fama, E. F., & F ench, K. R. (2015). A i e- ac o asse p icing model.
Jou nal o inancial economics
,
116
(1), 1-22.
Feng, J., Goodell, J. W., & Shen, D. (2022). ESG a ing and s ock p ice c ash isk: E idence om China.
Finance Resea ch Le e s
,
46
, 102476.
Habib, A., Hasan, M. M., & Jiang, H. (2018). S ock p ice c ash isk: e iew o he empi ical li e a u e.
Accoun ing & Finance
,
58
, 211-251.
Ha ey, C. R., & Siddique, A. (2000). Condi ional skewness in asse p icing es s.
The Jou nal o inance
,
55
(3), 1263-1295.
Hong, H., & S ein, J. C. (1999).
Di e ences o opinion, a ional a bi age and ma ke c ashes
. Na ional
bu eau o economic esea ch Camb idge, Mass., USA.
Hunj a, A. I., Mehmood, R., & Tayachi, T. (2020). How do co po a e social esponsibili y and co po a e
go e nance a ec s ock p ice c ash isk?
Jou nal o Risk and Financial Managemen
,
13
(2), 30.
Hu on, A. P., Ma cus, A. J., & Teh anian, H. (2009). Opaque inancial epo s, R2, and c ash isk.
Jou nal
o inancial Economics
,
94
(1), 67-86.
In, S. Y., & Schumache , K. (2021). Ca bonwashing: A New Type o Ca bon Da a-Rela ed ESG
G eenwashing.
A ailable a SSRN
.
Jin, L., & Mye s, S. C. (2006). R2 a ound he wo ld: New heo y and new es s.
Jou nal o Financial
Economics
,
79
(2), 257-257-292.
Kabi , M. N., Rahman, S., Rahman, M. A., & Anwa , M. (2021). Ca bon emissions and de aul isk:
In e na ional e idence om i m-le el da a.
Economic Modelling
,
103
, 105617.
37
WC03351
To al Liabili ies
All sho - and long- e m obliga ions
expec ed o be sa is ied by he
company.
Time Se ies
WC04101
De e ed Income Taxes &
In es men Tax C edi
The inc ease o dec ease in he
de e ed ax liabili y om one yea
o he nex esul ing om iming
di e ences in ecogni ion o
e enues and expenses o ax and
inancial epo ing pu poses.
Time Se ies
WC03451
P e e ed S ock
A claim p io o he common
sha eholde s on he ea nings o a
company and on he asse s in he
e en o liquida ion.
Time Se ies
WC03251
Long-Te m Deb
All in e es -bea ing inancial
obliga ions, excluding amoun s
due wi hin one yea . I is shown
ne o p emium o discoun .
Time Se ies
WC08326
Re u n On Asse s
(Ne Income – Bo om Line +
((In e es Expense on Deb -In e es
Capi alized) * (1-Tax Ra e))) /
A e age o Las Yea 's and Cu en
Yea ’s To al Asse s
Time Se ies

38
D. Indus y Classi ica ion o Sample Fi m
This Appendix p esen s a b eakdown o he 1,279 US-lis ed s ocks by indus y, classi ied acco ding o he
Fama-F ench 48 indus y ca ego ies desc ibed in sec ion 4. Acco dingly, inancials (SIC codes 6000-6999)
and u ili ies (SIC codes 4900-4949) a e excluded om his classi ica ion.
Indus y Classi ica ion
Numbe o Fi ms
% o he To al
34 Business Se ices
204
15,95%
36 Elec onic Equipmen
89
6,96%
42 Re ail
81
6,33%
30 Pe oleum and Na u al Gas
78
6,10%
40 T anspo a ion
60
4,69%
21 Machine y
60
4,69%
14 Chemicals
57
4,46%
13 Pha maceu ical P oduc s
48
3,75%
41 Wholesale
44
3,44%
23 Au omobiles and T ucks
42
3,28%
43 Res au an s, Ho els, Mo els
38
2,97%
12 Medical Equipmen
37
2,89%
35 Compu e s
34
2,66%
32 Communica ion
34
2,66%
37 Measu ing and Con ol Equipmen
32
2,50%
2 Food P oduc s
31
2,42%
17 Cons uc ion Ma e ials
27
2,11%
9 Consume Goods
25
1,95%
19 S eel Wo ks E c
25
1,95%
18 Cons uc ion
23
1,80%
38 Business Supplies
20
1,56%
22 Elec ical Equipmen
19
1,49%
10 Appa el
15
1,17%
28 Non-Me allic and Indus ial Me al Mining
14
1,09%
33 Pe sonal Se ices
14
1,09%
4 Bee & Liquo
13
1,02%
24 Ai c a
13
1,02%
15 Rubbe and Plas ic P oduc s
12
0,94%
6 Rec ea ion
10
0,78%
8 P in ing and Publishing
10
0,78%
39
39 Shipping Con aine s
10
0,78%
11 Heal hca e
9
0,70%
7 En e ainmen
7
0,55%
48 O he
7
0,55%
3 Candy & Soda
6
0,47%
29 Coal
6
0,47%
1 Ag icul u e
4
0,31%
5 Tobacco P oduc s
4
0,31%
20 Fab ica ed P oduc s
4
0,31%
27 P ecious Me als
4
0,31%
16 Tex iles
4
0,31%
25 Shipbuilding, Rail oad Equipmen
3
0,23%
26 De ense
2
0,16%
To al
1279
100,00%