Understanding the relationship between impact investment firms and market volatility: An analysis of European impact firms amidst COVID-19 and ongoing conflict

Mas e ’s Deg ee P og am in S a is ics and In o ma ion Managemen UNDERSTANDING THE RELATIONSHIP BETWEEN IMPACT INVESTMENT FIRMS AND MARKET VOLATILITY An analysis o Eu opean Impac Fi ms amids COVID-19 and Ongoing Con lic Fábio Alexand e Al es de Sousa Mas e Thesis p esen ed as pa ial equi emen o ob aining he Mas e ’s Deg ee in S a is ics and In o ma ion Managemen NOVA In o ma ion Managemen School Ins i u o Supe io de Es a ís ica e Ges ão de In o mação Uni e sidade No a de Lisboa NOVA In o ma ion Managemen School Ins i u o Supe io de Es a ís ica e Ges ão de In o mação Uni e sidade No a de Lisboa UNDERSTANDING THE RELATIONSHIP BETWEEN IMPACT INVESTMENT FIRMS AND MARKET VOLATILITY An analysis o Eu opean Impac Fi ms amids COVID-19 and Ongoing Con lic by Fábio Alexand e Al es de Sousa Mas e Thesis p esen ed as pa ial equi emen o ob aining he Mas e ’s deg ee in S a is ics and In o ma ion Managemen , wi h a specializa ion in Risk Analysis and Managemen Supe ised by A shin Asho eh, PhD, NOVA In o ma ion Managemen School July, 2024 i STATEMENT OF INTEGRITY 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 Rules o Conduc and Code o Hono om he NOVA In o ma ion Managemen School. [Lisbon, July 15, 2024] ii DEDICATION This hesis is dedica ed o he many people who suppo ed and encou aged me h oughou his jou ney. To my gi l iend, whose uncondi ional suppo p o ided me wi h he s eng h and mo i a ion I needed du ing he mos s ess ul imes. You unde s anding and pa ience we e in aluable. To my pa en s, o hei suppo and encou agemen , allowing me o pu sue my educa ion and ollow my d eams. You sac i ices and belie in me ha e made his achie emen possible. To my amily and iends, o hei companionship and o helping me ake b eaks and dis ac mysel om he s ess o hesis wo k. You suppo kep me g ounded and balanced. To my colleagues and he company, I wo ked wi h du ing his p ocess, o p o iding he en i onmen and condi ions ha enabled me o comple e his wo k. To my ellow mas e 's s uden s, o hei in aluable assis ance and sha ed knowledge. You help and iendship made his jou ney mo e manageable and enjoyable. Finally, o my supe iso , o hei guidance, suppo , and expe ise h oughou his p ocess. Thank you all o you signi ican con ibu ions o his achie emen . iii ABSTRACT The pas decade has been ma ked by signi ican social and acial inequali ies, global con lic s, clima e challenges, and he ongoing epe cussions o he COVID-19 pandemic. Amid hese issues, he g ow h and de elopmen o impac in es ing signal a subs an ial shi in he in es men landscape, sugges ing a p omising u u e whe e in es men s ha d i e posi i e social and en i onmen al change will be c ucial in add essing majo global p oblems. Despi e he apid e olu ion o impac in es ing, exis ing esea ch on he inancial pe o mance and isk associa ed wi h hese in es men s p esen s a a ie y o pe spec i es. This pape employs Dynamic Condi ional Co ela ion GARCH modelling o analyse he ime- a ying ola ili y and co ela ion dynamics be ween Eu opean Impac Fi ms and Eu opean ma ke e e ence indices, including he STOXX Eu ope 600 Index and ola ili y indices such as he EURO STOXX 50 Vola ili y and DAX-New Vola ili y Indices. The analysis is conduc ed ac oss h ee dis inc pe iods: p e-COVID-19, du ing COVID-19, and he pandemic-wa pe iod, encompassing he pandemic and he ongoing con lic . Ou indings sugges po en ial sho - e m di e si ica ion and isk-hedging bene i s o EU impac i ms, while long- e m ola ili y con agion highligh s he subs an ial and pe sis en impac o ma ke ola ili y on hese i ms o e ex ended pe iods. KEYWORDS Impac in es ing; Vola ili y; COVID-19; Di e si ica ion; DCC-GARCH model Sus ainable De elopmen Goals (SDG): i TABLE OF CONTENTS 1. In oduc ion .................................................................................................................. 1 2. Li e a u e e iew .......................................................................................................... 4 2.1. Can impac in es men s be jus i ied h ough a sa e-ha en, isk hedging and di e si ica ion ole? ........................................................................................................... 4 2.2. Do impac in es men s show esilience du ing c isis pe iods? ............................ 6 3. Da a and Me hodology ................................................................................................. 8 3.1. Da a Collec ion ...................................................................................................... 8 3.1.1. Sc eening Impac Fi ms .................................................................................. 8 3.1.2. Cons uc ing ou Impac In es ing Index ..................................................... 10 3.1.3. B oade Ma ke and Vola ili y Indices ......................................................... 11 3.2. Me hodology ....................................................................................................... 12 3.2.1. Exponen ially weigh ed mo ing a e age (EWMA) ....................................... 12 3.2.2. Dynamic Condi ional Co ela ion (DCC) GARCH........................................... 13 3.2.3. Copula Dynamic Condi ional Co ela ion GARCH ........................................ 15 4. Empi ical S udy ........................................................................................................... 17 5. Resul s and discussion ................................................................................................ 26 Sho - e m ola ili y con agion be ween Eu opean impac i ms and Eu opean ma ke ’s ola ili y (V2TX/V1XI) ...................................................................................................... 28 Long- e m ola ili y con agion be ween Eu opean impac i ms and Eu opean ma ke ’s ola ili y (V2TX/V1XI) ...................................................................................................... 29 Sho - e m ola ili y con agion be ween Eu opean impac i ms and Eu opean ma ke ’s benchma k (STOXX) ......................................................................................................... 31 Long- e m ola ili y con agion be ween Eu opean impac i ms and Eu opean ma ke ’s benchma k (STOXX) ......................................................................................................... 32 6. Conclusions and u u e wo ks .................................................................................... 37 Bibliog aphical Re e ences .............................................................................................. 39 Appendix A ...................................................................................................................... 44 Appendix B ...................................................................................................................... 50 Appendix C ...................................................................................................................... 52 Appendix D ...................................................................................................................... 53 Appendix E ....................................................................................................................... 55 LIST OF FIGURES Figu e 1 - P ices and alues o Impac In es ing Fi ms Index (IIFX) , STOXX Eu ope 600 Index (STOXX), and he wo ma ke ola ili y indices (V2TX and V1XI) om Janua y 4, 2016, and Ma ch 27, 2024. ............................................................................................................... 17 Figu e 2 - Log Re u ns o Impac In es ing Fi ms Index (IIFX) , STOXX Eu ope 600 Index (STOXX), and he wo ma ke ola ili y indices (V2TX and V1XI) om Janua y 5, 2016, and Ma ch 27, 2024. ........................................................................................................................... 18 Figu e 3 - Sca e plo o e u ns be ween Impac In es ing Fi ms Index (IIFX) and STOXX Eu ope 600 Index (STOXX), and he wo ma ke ola ili y indices (V2TX and V1XI) om Janua y 5, 2016, and Ma ch 27, 2024. .............................................................................................. 19 Figu e 4 - Rolling co a iances and co ela ions be ween he Impac In es ing Fi ms Index (IIFX) and STOXX Eu ope 600 Index (STOXX), and he wo ma ke ola ili y indices (V2TX and V1XI), using di e en ime windows. ............................................................................... 23 Figu e 5 - EWMA co a iances and co ela ions be ween he Impac In es ing Fi ms Index (IIFX) and STOXX Eu ope 600 Index (STOXX), and he wo ma ke ola ili y indices (V2TX and V1XI), using di e en ime windows. ............................................................................... 24 Figu e 6 – Dynamic Condi ional Co ela ion (DCC) be ween he Impac In es ing Fi ms Index (IIFX) and he h ee b oade ma ke (STOXX) and ola ili y indices (V2TX and V1XI) om Janua y 5, 2016, o Ma ch 27, 2024. ................................................................................ 34 Figu e B1 - Au oco ela ion unc ion (ACF) o he ou e u n se ies (𝑦𝑡): Impac In es ing Fi ms Index (IIFX), STOXX Eu ope 600 EUR P ice Index (STOXX), EURO STOXX 50 Vola ili y (VSTOXX) EUR P ice Index (V2TX), and DAX-New Vola ili y Index (V1XI) ........................ 50 Figu e B2 - Au oco ela ion unc ion (ACF) o he ou e u n se ies as powe o 2 (𝑦𝑡2): Impac In es ing Fi ms Index (IIFX), STOXX Eu ope 600 EUR P ice Index (STOXX), EURO STOXX 50 Vola ili y (VSTOXX) EUR P ice Index (V2TX), and DAX-New Vola ili y Index (V1XI) ......... 51 Figu e D1 - Dynamic Condi ional Co ela ion (DCC) be ween he Impac In es ing Fi ms Index (IIFX) and he h ee b oade ma ke (STOXX) and ola ili y indices (V2TX and V1XI) om Janua y 5, 2016, o Janua y 23, 2020 (P e-COVID-19 pe iod). ........................................ 53 i Figu e D2 - Dynamic Condi ional Co ela ion (DCC) be ween he Impac In es ing Fi ms Index (IIFX) and he h ee b oade ma ke (STOXX) and ola ili y indices (V2TX and V1XI) om Janua y 24, 2020, o May 05, 2023 (Du ing COVID-19 pe iod). ....................................... 54 Figu e D3 - Dynamic Condi ional Co ela ion (DCC) be ween he Impac In es ing Fi ms Index (IIFX) and he h ee b oade ma ke (STOXX) and ola ili y indices (V2TX and V1XI) om Janua y 24, 2020, o Ma ch 27, 2024 (Pandemic-Wa pe iod). ....................................... 54 LIST OF TABLES Table 1 - Impac in es ing i ms’ dis ibu ion o impac a eas ................................................ 10 Table 2 - Summa y s a is ics on indices’ daily e u n se ies, Janua y 5, 2016, o Ma ch 27, 2024. .......................................................................................................................................... 20 Table 3 - P elimina y e idence o uncondi ional co ela ion .................................................. 22 Table 4 - Es ima ion esul s o DCC-GARCH model .................................................................. 27 Table 5 - Es ima ion esul s o Copula-DCC-GARCH model ..................................................... 33 Table C1 - Maximum likelihood es ima ion o uni a ia e GARCH model ............................. 52 Table E1 - In o ma ion C i e ia, Log-likelihood, MSE and RMSE o mul i a ia e GARCH models wi h m -dis ibu ion o IIFX-STOXX se ies .................................................................... 55 Table E2 - In o ma ion C i e ia, Log-likelihood, MSE and RMSE o mul i a ia e GARCH models wi h m -dis ibu ion o IIFX-V2TX se ies ...................................................................... 56 Table E3 - In o ma ion C i e ia, Log-likelihood, MSE and RMSE o mul i a ia e GARCH models wi h m -dis ibu ion o IIFX-V1IX se ies ....................................................................... 57 LIST OF ABBREVIATIONS AND ACRONYMS IF Impac Fi ms IIFX Impac In es ing Fi ms Index STOXX STOXX Eu ope 600 EUR P ice Index V2TX EURO STOXX 50 Vola ili y (VSTOXX) EUR P ice Index V1XI DAX-New Vola ili y Index DCC Dynamic Condi ional Co ela ion GARCH Gene alized Au o eg essi e Condi ional He e oskedas ici y 1 1. INTRODUCTION Impac in es ing has e ol ed om a niche concep o a widely adop ed in es men app oach. In es o s a e inc easingly aligning hei inancial goals wi h posi i e social and en i onmen al ou comes (Höchs äd e & Scheck, 2015). Ea ly pionee s, such as he Global Impac In es ing Ne wo k (GIIN), we e among he ea ly isiona ies behind wha we now ecognize as impac in es ing. The GIIN de ines impac in es ing as “in es men s made wi h he in en ion o gene a e measu able posi i e social and en i onmen al impac alongside inancial e u ns, and speci ically use ha in es men capi al along wi h engagemen o in es men e ms o posi i ely in luence a ge ed impac esul s.” This in en ionali y di e en ia es impac in es ing om pu ely inancially-mo i a ed in es men s. The apid g ow h o he impac in es ing ma ke is e lec ed in he subs an ial inc ease in asse s unde managemen , which ha e isen om $95 billion in 2017 o $213 billion by he end o 2022. This su ge indica es a p omising ajec o y, sugges ing ha impac in es men s will play a c ucial ole in add essing some o he wo ld's mos p essing challenges in he yea s o come (Hand e al., 2023). Howe e , despi e he quick expansion o he impac in es ing landscape, he e is s ill a lack o su icien li e a u e on he opic (see, Chiappini e al., 2023 and Miglia acca e al., 2022). Ques ions emain a ound he de ini ion o impac in es ing and how i di e s om ela ed concep s like socially esponsible in es ing (SRI). Cla i ying hese bounda ies is essen ial o a be e unde s anding and applica ion o impac in es ing. Exis ing esea ch has con ibu ed o a clea e delinea ion o impac in es ing, and o a clea delinea ion o bounda ies o b oade concep s like socially esponsible in es ing (Höchs äd e & Scheck, 2015). I is commonly pe cei ed as ei he di e en om o o go beyond ela ed concep s like socially esponsible in es ing. The easons a e ha while impac in es o s seek o und p ojec s wi h posi i e and measu able socie al and en i onmen al goals, and a e open o accep ing ade-o s be ween hose social and also inancial goals (Ba be e al., 2021), socially esponsible in es o s o en ocus on En i onmen al, Social and Go e nance (ESG) c i e ia o imp o e co po a e p ac ices, he eby enhancing inancial pe o mance h ough epu a ion. Some companies may mis ep esen social and en i onmen al impac o po ay hemsel es as “pu pose-led b ands” (Mo duch & Ogden, 2019). Geczy e al. (2021) delinea ed he con ac ual and go e nance p ac ices inhe en in impac in es ing, emphasizing he delica e balance equi ed be ween inancial objec i es and social- bene i goals. In a complemen a y s udy, Bu on e al. (2021) del ed in o he cha ac e is ics o impac in es o s, de ining hem as indi iduals o en i ies wi h a dual manda e o gene a ing social good and inancial e u ns. These in es o s me iculously selec companies ha demons a e a capaci y o deli e signi ican social alue, o en main aining highe equi y s akes in such en e p ises (Chowdh y e al., 2019). 8 3. DATA AND METHODOLOGY 3.1. DATA COLLECTION 3.1.1. SCREENING IMPACT FIRMS To be eligible o impac in es men s, he i m's p oduc and mission should a ge speci ic social o en i onmen al ou comes, such as con ibu ing o educe po e y and hunge , imp o e he quali y o educa ion and p o ide clean wa e and sani a ion. Ano he common s a egy is o in es in companies wi h posi i e impac h ough hei ope a ional p ac ices. This in ol es selec ing companies ha demons a e esponsible and sus ainable business p ac ices in hei day- o-day ope a ions, such as educing hei ca bon oo p in , p omo ing ai labou p ac ices, o implemen ing en i onmen ally iendly policies (Hand e al., 2020). Accessing he igh da a and me ics o impac in es ing is a big challenge. E en hough e o s ha e been made o help make da a mo e a ailable, mos s udies on impac in es men s s ill ely on p i a e da a. P i a e impac unds, especially, end o keep hei in es men da a o hemsel es. And when hey do sha e, hey o en lea e ou any in es men s ha didn' do well, which makes hei o e all pe o mance look be e han i migh be (Wa s & Scales, 2020). Hence, o see i impac in es men s can be jus i ied h ough a sa e-ha en, isk-hedging, and di e si ica ion ole, and i hey do show esilience du ing c isis pe iods, we me iculously ga he ed a unique da ase om a selec ion o EU lis ed companies. These companies p ima ily ocus on add essing a ious social and en i onmen al challenges ou lined in he Uni ed Na ions Sus ainable De elopmen Goals (SDGs). The SDGs, comp ising 17 goals ha add ess a wide a ay o sus ainable de elopmen issues, se e as a globally ecognized amewo k o de ining impac a eas, e lec ing a consensus among s akeholde s on ambi ious de elopmen objec i es. Impac in es o s inc easingly a ou ed companies whose o e ings align wi h a leas one SDG, using hem as a benchma k o es ablish impac pe o mance goals, as highligh ed in GIIN's epo (Hand e al., 2020). Consequen ly, companies a e op ing o inco po a e SDGs in o hei non- inancial disclosu es (Pizzi e al., 2022). Following a simila selec ion p ocess o Giacomini e al. (2023) , we ini ia ed ou s udy by selec ing companies lis ed in he wikiposi i e.o g da abase, which ea u ed close o 800 companies globally by he end o Ap il 2024. F om his ex ensi e da abase, we speci ically ocused on EU-lis ed companies, na owing down he selec ion o jus o e 200 companies. Subsequen ly, we go ano he lis o companies by u ilizing Re ini i Sc eene and applying wo il e s: ESG nega i e sc eening AND alignmen wi h a leas one Sus ainable De elopmen Goal (SDG). I used web sc aping o e ie e a able o company names om he wikiposi i e.o g da abase, whe e each name included a hype link o addi ional de ails abou he company. By ollowing each link, I ga he ed speci ic in o ma ion, such as he ISIN code, allowing me o c oss- e e ence bo h lis s ISIN codes o iden i y which ones mee ou s udy c i e ia. Finally, by conduc ing sea ches on each company’s websi e, we sc u inize hei co e 9 business desc ip ions o pinpoin i ms explici ly commi ed o social and en i onmen al sus ainabili y h ough hei p oduc s o se ices. This comp ehensi e app oach enabled us o include companies dedica ed o posi i e social and en i onmen al impac , while also conside ing ESG c i e ia and SDG alignmen , ensu ing a comp ehensi e and obus analysis o impac in es men s in ou esea ch. This me iculous sc eening p ocess yields ou inal sample comp ising 55 EU impac i ms (he ea e , IF) spanning he pe iod om beginning 2016 o Ma ch 2024. The size o ou sample e lec s he niche na u e o impac in es ing compa ed o mo e adi ional ESG and socially esponsible in es men s. As p e iously men ioned, Be nal e al. (2021) examined he public equi y ma ke in Eu ope, analysing a sample o only 16 lis ed i ms. Ou da ase , which includes 55 Eu opean impac i ms, ma ks a no able ad ancemen compa ed o hei s udy. Appendix A p o ides a comp ehensi e lis o companies included in his s udy along wi h desc ip ions o hei p ima y business ac i i ies. These companies showcase a di e se ange o impac ocuses. While some p io i ize sus ainable echnology, in as uc u e, o cons uc ion, o he s may concen a e on heal hca e, educa ion, o en i onmen al conse a ion. Fu he mo e, ou analysis does no include companies in he banking o go e nmen sec o s which p e en s egula o y in luences om biasing ou esul s. To ca ego ize hese 55 impac i ms ac oss each impac goal, we ha e es ablished i e impac ca ego ies ollowing he ecommenda ions o MSCI ESG Resea ch, which o e s a amewo k encompassing all SDGs. Table 1 illus a es he dis ibu ion o he 55 companies ac oss hese impac ca ego ies. The Empowe men ca ego y ep esen s he smalles segmen , accoun ing o 6% o he i ms. In con as , Na u al Capi al eme ges as he mos p e alen ca ego y, comp ising 38.2%, ollowed by Basic Needs and Clima e Change, each ep esen ing 21.8%. 10 Table 1 - Impac in es ing i ms’ dis ibu ion o impac a eas Impac a eas SDGs Numbe Pe cen Basic needs No po e y, Ze o Hunge , Good Heal h and Well-Being, Clean Wa e and Sani a ion, Sus ainable Ci ies and Communi ies 12 21,8% Clima e Change A o dable and Clean Ene gy, Clima e Ac ion 12 21,8% Empowe men Quali y Educa ion, Gende Equali y, Decen Wo k and Economic G ow h, Reduced Inequali ies, Indus y, Inno a ion, and In as uc u e 10 18,2% Na u e Capi al Sus ainable Consump ion and P oduc ion, Li e on Land, Li e Below Wa e 21 38,2% Go e nance Peace, Jus ice and S ong Ins i u ions, Pa ne ship 0 0,0% To al 55 100,0% 3.1.2. CONSTRUCTING OUR IMPACT INVESTING INDEX In o de o cons uc ou Impac In es ing Fi ms Index (he ea e , IIFX), we ga he ed he daily closing p ices o each o he 55 impac i ms om Janua y 2016 o Ma ch 2024, along wi h hei ma ke capi aliza ion and ee- loa a e as o Ma ch 2024, using Thomson Reu e s Da as eam. Fo he index cons uc ion, we adop ed he me hodology employed by STOXX, u ilizing he STOXX Eu ope 600 as ou benchma k. This comp ehensi e index ep esen s he pe o mance o he 600 la ges cap companies ac oss 17 Eu opean coun ies, se ing as a eliable measu e o Eu opean equi y ma ke pe o mance. The STOXX Eu ope 600 employs a ee- loa ma ke capi aliza ion weigh ing me hodology o ensu e ha i accu a ely e lec s he po ion o he ma ke a ailable o in es o s o ading. Each cons i uen 's weigh ing wi hin he index is de e mined by di iding i s ee- loa a e ma ke capi aliza ion by he sum o all cons i uen s' ee- loa a e ma ke capi aliza ion. The esul ing weigh o each cons i uen is hen mul iplied by i s sha e p ice o de e mine i s impac on he o e all index pe o mance. 11 Regula qua e ly e iews a e conduc ed, and du ing he e iew p ocess, he mos ecen ee- loa da a a ailable o STOXX is u ilized o componen selec ion. This ensu es ha he index accu a ely e lec s he cu en s a e o he ma ke . Fo ou analysis, we u ilized he mos ecen ee- loa and ma ke capi aliza ion da a a ailable a he ime o ou analysis, Ma ch 2024. This app oach ensu es ha he index accu a ely e lec s he ma ke landscape a he end o ou analysis pe iod, aligning wi h indus y s anda ds and bes p ac ices o index cons uc ion. I is impo an o no e ha we ha e no inco po a ed adjus men s o co po a e ac ions as sugges ed by STOXX index me hodology. Co po a e ac ions such as s ock spli s, me ge s, acquisi ions, o spin-o s can impac he composi ion and weigh ing o he index, bu we ha e op ed o a simpli ied app oach ha ocuses solely on he ee loa ma ke capi aliza ion o each company. 3.1.3. BROADER MARKET AND VOLATILITY INDICES In examining he ela ionship be ween Eu opean impac in es ing i ms and he Eu opean ma ke , i is impe a i e o es ablish a comp ehensi e unde s anding o he ma ke dynamics. This en ails a ho ough analysis o key ma ke e e ence indices ha se e as benchma ks o assessing he pe o mance and ola ili y o Eu opean equi ies. The STOXX Eu ope 600 index s ands as a p ominen benchma k, encompassing a di e se a ay o la ge, mid, and small capi aliza ion companies spanning 17 Eu opean coun ies. I s b oad co e age, ep esen ing app oxima ely 90% o he Eu opean s ock ma ke capi aliza ion, p o ides i an impo an ool o in es o s seeking insigh s in o he o e all pe o mance o Eu opean equi ies. Mo eo e , ola ili y indices such as he EURO STOXX 50 Vola ili y and he DAX-New Vola ili y Index o e c ucial pe spec i es on ma ke sen imen and isk pe cep ion. The EURO STOXX 50 Vola ili y Index acks he implied ola ili y o op ions on he EURO STOXX 50 Index, p o iding aluable insigh s in o in es o s' expec a ions ega ding u u e ola ili y wi hin he Eu ozone ma ke . Simila ly, he DAX-New Vola ili y Index moni o s he implied ola ili y o op ions on he DAX, shedding ligh on he an icipa ed ola ili y o Ge many's leading blue-chip s ocks. High le els o he EURO STOXX 50 Vola ili y and DAX-New Vola ili y Indices o en indica e ma ke s ess, while low le els sugges ma ke calmness. Implied ola ili y e lec s ma ke expec a ions o u u e p ice mo emen s, de i ed om op ion p ices. We elied on hese key ma ke e e ence indices o in es iga e he po en ial oles o impac i ms as sa e-ha en, isk-hedging, and di e si ica ion in he con ex o he Eu opean ma ke . 12 3.2. METHODOLOGY To assess he po en ial bene i s o di e si ica ion by including impac asse s in a adi ional inancial po olio, we analysed he co ela ion o inancial e u ns be ween Eu opean impac i ms and Eu opean ma ke benchma k indices. Impac in es ing has unde gone signi ican changes in ecen yea s, so we wan ed o de e mine i he beha iou o impac asse s has changed, pa icula ly du ing pe iods o ma ke s ess such as he COVID-19 pandemic and i s a e ma h, as well as he cu en con ex o geopoli ical ensions in Eas e n Eu ope and he Middle Eas . 3.2.1. EXPONENTIALLY WEIGHTED MOVING AVERAGE (EWMA) Be o e employing he DCC-GARCH (Dynamic Condi ional Co ela ion - Gene alized Au o eg essi e Condi ional He e oskedas ici y) model in ou analysis, we i s u ilized he Exponen ially Weigh ed Mo ing A e age (EWMA) me hod o in es iga e he dynamic ela ionships be ween he Impac In es ing Fi ms Index (IIFX) and he indices STOXX, V2TX, and V1XI. The a ionale o using EWMA lies in i s abili y o p o ide a p elimina y unde s anding o he ime- a ying co a iances and co ela ions wi h less compu a ional complexi y compa ed o DCC-GARCH. EWMA o e s a clea and in ui i e app oach o ack changes in he ela ionships o e ime, making i an excellen p ecu so o he mo e sophis ica ed DCC-GARCH analysis. The EWMA me hod assigns exponen ially dec easing weigh s o olde obse a ions, allowing ecen da a o ha e a mo e signi ican impac on he calcula ed a e ages. This weigh ing scheme is con olled by a smoo hing pa ame e , lambda (λ), se o 0.94. By using exponen ially dec easing weigh s, he EWMA me hod places mo e emphasis on he mos ecen obse a ions, which is c ucial in inancial ma ke s whe e ecen e en s o en ha e a g ea e impac on cu en ma ke condi ions han olde ones. This p ope y makes EWMA pa icula ly use ul o modelling inancial ime se ies da a, as i allows o a mo e esponsi e adjus men o changes in ma ke ola ili y and co ela ions. In ou analysis, we u ilized bo h a sho e window (lambda = 0.94) and a longe window (hal - li e = 125 days) o he EWMA calcula ions. The sho e window p o ided de ailed insigh s in o he sho - e m luc ua ions and immedia e impac s o ma ke e en s on he ela ionships. We also explo ed he use o a longe hal -li e pa ame e , which e ec i ely adjus s lambda o p o ide a di e en pe spec i e on he da a. This longe window smoo hs ou sho - e m luc ua ions and highligh s longe - e m ends, p o iding a complemen a y iew o he s anda d lambda se ing. By applying EWMA, we aimed o gain p elimina y insigh s in o he dynamic beha iou o he ela ionships be ween EU impac i ms and he indices be o e mo ing o he mo e complex and compu a ionally in ensi e DCC-GARCH model. The insigh s ob ained om EWMA analysis 13 se ed as a ounda ion o unde s anding he empo al e olu ion o co a iances and co ela ions, he eby in o ming and guiding he subsequen DCC-GARCH analysis. 3.2.2. DYNAMIC CONDITIONAL CORRELATION (DCC) GARCH In his s udy, Dynamic Condi ional Co ela ion GARCH modelling is used o model he ime- a ying ola ili y and co ela ion dynamics be ween Eu opean impac i ms and Eu opean ma ke e e ence indices. Unde s anding how hey in e ac is i al as impac in es men g ows. A e conduc ing se e al s a is ical es s, such as ARCH and ACF es s, we ound ha he DCC-GARCH model p o ided he bes i o he da a. This inding is u he suppo ed by p e ious s udies, including Yousa e al. (2022); Be nal e al. (2021); Haq e al. (2021); Saeed e al. (2020); and Sado sky (2014) , which ha e also demons a ed he e ec i eness o he DCC-GARCH model in cap u ing he dynamic ela ionships wi hin simila inancial da ase s. The Dynamic Condi ional Co ela ion GARCH model (DCC-GARCH; Engle, 2002; Engle & Sheppa d, 2001), in oduced by Engle (2002) , is a powe ul ool o cap u ing ime- a ying and dynamic ela ionships in mul i a ia e se ings. By combining he GARCH model o es ima e indi idual asse s' condi ional a iance and he dynamic condi ional co ela ion model o assess ime- a ying co ela ions, he DCC-GARCH model o e s a comp ehensi e unde s anding o ola ili y and co ela ion dynamics wi hin inancial ma ke s. When s udying he ola ili y o inancial ime se ies, esea che s equen ly u n o GARCH, a me hod ha uses an au o eg essi e amewo k o model a ime se ies' condi ional a iance, allowing ola ili y shocks o pe sis o e ime. Mul i a ia e GARCH models such as DCC-GARCH ha e been ex ensi ely used o explo e con agion e ec s be ween ma ke s, p o iding insigh s in o he in ica e in e connec ions and in e dependencies among a ious asse s (see, o example, Nguyen e al., 2022; Zhang e al., 2022; Celık, 2012; Ma ens & Poon, 2001). The DCC-GARCH model ollows a wo-s ep p ocess: 1. Es ima ion o he condi ional a iance (ℎ,) o each a iable, using GARCH(1,1) model, speci ied in Eqs. (1) and (2). 𝑟 = 𝜇+ 𝜖 ( 1 ) ℎ, = 𝜔+ 𝛼𝜀 + 𝛽ℎ, ( 2 ) whe e, 𝛼 and 𝛽 a e non-nega i e pa ame e s, whe e 𝛼 signi ies sho - e m ola ili y sensi i i y and 𝛽 deno es long- e m ola ili y pe sis ence. The sum o alpha and be a coe icien s se es as a measu e o he pe sis ence o ola ili y shocks. This sum is expec ed o be less han 1, indica ing ha he impac o shocks diminishes o e ime. 14 I he sum exceeds 1, i sugges s ha he shock has an explosi e e ec , pe sis ing inde ini ely and ampli ying ola ili y. 2. Es ima ion o DCC pa ame e s, o compu e he ime- a ying condi ional co a iance ma ix 𝐻, which is a unc ion o he condi ional co ela ion ma ix 𝑅 and diagonal ma ix o ime- a ying s anda d de ia ions 𝐷. 𝐻= 𝐷𝑅𝐷 ( 3 ) whe e 𝑅 is he condi ional co ela ion ma ix and 𝐷 ep esen s he diagonal ma ix o ime- a ying s anda d de ia ions, speci ied in Eqs. (4) and (5). 𝑅= 𝑄 ∗𝑄𝑄 ∗ ( 4 ) 𝐷= 𝑑𝑖𝑎𝑔ℎ∙   ⁄.....ℎ∙   ⁄ ( 5 ) 𝑄 ∗ is calcula ed as ollows: 𝑄 ∗ = (𝑑𝑖𝑎𝑔(𝑄)) . In his model, he diagonal elemen s o he condi ional co a iance ma ix a e modelled using uni a ia e GARCH p ocesses, while he o -diagonal elemen s o 𝐻 a e ep esen ed as nonlinea unc ions o he diagonal e ms and compu ed as: [𝐻] = ℎ   ⁄ℎ   ⁄𝜌,𝑖 ≠ 𝑗 ( 6 ) The e o e, he ime- a ying condi ional co ela ion is es ima ed as ollows: 𝜌∙ =𝒬∙ 𝒬∙𝒬∙ ,𝑖 ≠ 𝑗 ( 7 ) This wo-s ep app oach allows o he modelling o ime- a ying co ela ions in a mul i a ia e se ing while p ese ing he simplici y o uni a ia e GARCH models o indi idual asse s' ola ili y. Mo eo e , we elied on he Maximum Likelihood Es ima ion (MLE) me hod o e alua e model i , ensu ing ha he pa ame e s we e es ima ed in a way ha maximizes he likelihood. As a esul , we ound ha he inclusion o ARMA(1,1) componen s enhances he log-likelihood o ou model, signi ying a be e i o he da a by cap u ing linea dependencies in he mean 15 p ocess ha a plain GARCH model may o e look. While ou e u n se ies 𝑦 is app oxima ely unco ela ed bu dependen , mo e han 5 lags all ou side he con idence limi s, sugges ing he p esence o weak se ial co ela ions ha an ARMA model could cap u e. (see, Appendix B). 𝑟 = 𝜇+ 𝜙𝑟 + 𝜃𝜀 + 𝜀 ( 8 ) This inclusion o he au o eg essi e (AR) componen allows he mean equa ion o accoun o se ial co ela ion in he e u ns, p o iding a mo e accu a e ep esen a ion o he inancial da a. Addi ionally, he mo ing a e age (MA) componen helps o cap u e he impac o pas shocks on cu en e u ns. By inco po a ing bo h AR and MA componen s in he ARMA(1,1) model, we can e ec i ely model he dynamics o he e u ns, allowing o a mo e comp ehensi e and nuanced unde s anding o he unde lying p ocesses. This combined app oach helps o educe he esiduals' au oco ela ion, leading o be e pe o mance in subsequen ola ili y and co ela ion modelling s eps. Addi ionally, we es ed a ious dis ibu ions o he e o e ms, including he no mal Gaussian and S uden 's T dis ibu ions (see, Appendix C). Among hese, he S uden 's T dis ibu ion p o ed o be he mos e ec i e, hus being he one which esul s we decided o p esen in he s udy. The S uden 's T dis ibu ion was p e e ed because i accoun s o hea y ails and ex eme alues, which a e common in inancial da a. This dis ibu ion enhances he model's abili y o cap u e he isk and ola ili y associa ed wi h inancial e u ns, leading o mo e obus and ealis ic modelling. By inco po a ing hese elemen s, he model wi h a S uden 's T dis ibu ion o e s a mo e comp ehensi e unde s anding o he unde lying ma ke dynamics and imp o es he o e all p edic i e pe o mance o he model. 3.2.3. COPULA DYNAMIC CONDITIONAL CORRELATION GARCH The Copula-DCC-GARCH me hodology is designed o cap u e he complexi ies o inancial da a, pa icula ly he skewed, lep oku ic, and asymme ic dependence pa e ns exhibi ed by asse e u ns. I achie es his by linking uni a ia e ma ginal dis ibu ions o hei mul i a ia e dis ibu ion h ough copulas, a concep in oduced by Skla (1959) . This app oach allows o a mo e lexible and accu a e ep esen a ion o he dependencies be ween asse e u ns, add essing he limi a ions o adi ional models. The Copula-DCC-GARCH me hodology in ol es a wo-s ep p ocess. Fi s , uni a ia e e u n dis ibu ions a e es ima ed using GARCH models o cap u e ime- a ying ola ili y. This s ep ensu es ha he ma ginal dis ibu ions o each asse a e accu a ely modelled. Then, he condi ional dependence s uc u e is modelled using copulas, inco po a ing dynamic condi ional co ela ions (DCC) o cap u e ime- a ying co ela ions be ween asse s and 16 ma ke indices (Pa on, 2012, 2006). This s ep e ec i ely cap u es he dynamic in e dependencies be ween asse s and ma ke condi ions. The pa ame e s o he Copula-DCC-GARCH model a e es ima ed ia maximum likelihood es ima ion, a s anda d app oach in inancial econome ics. This ensu es obus ness in cap u ing he e ol ing ma ke dynamics ac oss di e en pe iods o s abili y and c isis. 17 4. EMPIRICAL STUDY The da ase used o he en i e sample pe iod has been ca e ully cleaned o ensu e accu acy and consis ency. Since he companies in he s udy a e om di e en pa s o Eu ope wi h a ying holidays and da a a ailabili y, i was impo an o add ess any po en ial biases. To achie e his, any ows wi h missing da a we e excluded. This means ha only days whe e all companies had comple e da a we e included in he analysis, esul ing in 1852 daily obse a ions om Janua y 5, 2016, o Ma ch 27, 2024. This me iculous p ocess ensu es ha he analysis is based on a s anda dized da ase , allowing o ai compa isons and accu a e e alua ions ac oss all companies. I conduc ed he analysis using R, u ilizing he ollowing packages: “ se ies”, “Pe o manceAnaly ics”, “quan mod”, “ uga ch”, “ca ”, “FinTS”, “ mga ch”, “copula”, and “lm es ”. Speci ically, “FinTS” helped jus i y he use o DCC-GARCH, h ough ARCH e ec es s, “ uga ch” was used o i he uni a ia e GARCH model, and “ mga ch” allowed o i and e alua e mul i a ia e GARCH models. Figu e 1 e eals a simila i y in ola ili y be ween EU impac i ms and STOXX, al hough he la e exhibi s mo e p onounced luc ua ions in p ices, pa icula ly du ing he COVID-19 pandemic pe iod. Bo h indices show simila non-posi i e ela ion wi h he ma ke ola ili y indices, which had a signi ican peak in hei alues coinciding wi h he COVID-19 pandemic and e lec ing ma ke expec a ions o u u e p ice mo emen s du ing his ime o heal h c isis. Figu e 1 - P ices and alues o Impac In es ing Fi ms Index (IIFX) , STOXX Eu ope 600 Index (STOXX), and he wo ma ke ola ili y indices (V2TX and V1XI) om Janua y 4, 2016, and Ma ch 27, 2024. 24 and 100-day olling co ela ion, which exhibi g ea e sensi i i y and ola ili y du ing c ises. The 252-day olling me ics smoo h ou hese luc ua ions, especially du ing he p olonged pe iods o c isis such as he pandemic and he wa . The IIFX and V1XI analysis ollows a simila pa e n o he V2TX analysis, wi h a nega i e co ela ion. Fo IIFX and STOXX, he sho - e m window e eals a ela i ely s able co a iance wi h a signi ican spike a ound he onse o he COVID-19 pandemic, indica ing a heigh ened co ela ion du ing his pe iod. The longe window, howe e , smoo hs ou hese sho - e m luc ua ions, sugges ing p olonged pe iods o highe co a iance du ing he pandemic. The co ela ion be ween IIFX and STOXX shows luc ua ions a ound a s ong posi i e alue, Figu e 5 - EWMA co a iances and co ela ions be ween he Impac In es ing Fi ms Index (IIFX) and STOXX Eu ope 600 Index (STOXX), and he wo ma ke ola ili y indices (V2TX and V1XI), using di e en ime windows. 25 e ealing a g adual decline since he pandemic, sugges ing a weakening bu s ill signi ican ela ionship o e ime. In con as , he analysis o IIFX and V2TX highligh s a mo e ola ile ela ionship. The sho - e m window displays equen spikes in co a iance du ing ma ke s ess pe iods, such as he COVID-19 pandemic. The co ela ion in he sho - e m window luc ua es conside ably, e lec ing ma ke ola ili y. Howe e , he longe window demons a es a consis en upwa d end, especially since he pandemic, indica ing a weakening bu s ill e y signi ican nega i e ela ionship be ween IIFX and V2TX du ing ola ile and unce ain ma ke condi ions. The IIFX and V1XI analysis ollows a simila pa e n o he V2TX analysis, wi h a nega i e co ela ion. The esul s show a mo e p onounced ela ionship be ween EU impac i ms and he STOXX index. This indica es ha EU impac i ms mo e s ongly in andem wi h he o e all ma ke , sugges ing limi ed di e si ica ion bene i s wi hin hese asse s. Con e sely, he nega i e co ela ions wi h V2TX and V1XI indices a e weake , implying ha EU impac i ms a e no as s ongly a ec ed by hese indices. This migh sugges some po en ial di e si ica ion bene i s agains implied ma ke ola ili y. 26 5. RESULTS AND DISCUSSION To de e mine whe he impac in es men s can se e as a sa e-ha en, isk-hedging, and di e si ica ion s a egy, and whe he hey exhibi esilience du ing c isis pe iods, we analysed he inancial e u ns o 55 EU impac i ms om Janua y 2016 o Ma ch 2024. Using a DCC Mul i a ia e GARCH model, we assessed he co ela ion o inancial e u ns be ween hese EU impac i ms and Eu opean ma ke benchma k indices, speci ically STOXX: STOXX Eu ope 600 Index, V2TX: EURO STOXX 50 Vola ili y Index, and V1XI: DAX-New Vola ili y Index. Ou da ase consis s o 1852 daily obse a ions om Janua y 5, 2016, o Ma ch 27, 2024, and is di ided in o h ee sub-samples. The i s sub-sample co e s he p e-COVID-19 pe iod, om Janua y 5, 2016, o Janua y 23, 2020. The second sub-sample, which includes he COVID-19 pe iod, spans om Janua y 24, 2020, ma king he beginning o COVID-19 in Eu ope, o May 5, 2023, ma king he end o COVID-19 as a global heal h eme gency by he Wo ld Heal h O ganiza ion (WHO). The hi d sub-sample co e s he pandemic-wa pe iod, which includes bo h he COVID-19 pe iod, s a ing om Janua y 24, 2020, and he ongoing wa ha began wi h Russia’s in asion o Uk aine on Feb ua y 24, 2022, and con inues o he p esen . The e is an o e lap du ing he COVID-19 and he pandemic-wa pe iod sub-samples, bo h ep esen ing pe iods o unce ain y and c isis in ou analysis, while he p e-COVID-19 pe iod ep esen s a mo e s able pe iod. 27 Table 4 - Es ima ion esul s o DCC-GARCH model IIFX - STOXX V2TX V1XI In e al A: Full sample pe iod (Janua y 5, 2016, o Ma ch 27, 2024) mu 0,00091 *** 0,00087 *** - 0,00234 *** - 0,00257 *** a 1 0,59805 ** 0,69392 *** 0,90520 *** 0,90349 *** ma1 - 0,65505 ** - 0,74200 *** - 0,97919 *** - 0,97445 *** omega 0,00001 0,00000 0,00020 0,00021 alpha1 0,12154 *** 0,14872 *** 0,07253 0,09254 ** be a1 0,82997 *** 0,83374 *** 0,89916 *** 0,87378 *** shape 6,20753 *** 4,11845 *** 3,87625 *** 3,87668 *** dcca1 0,05724 *** 0,04994 *** 0,04629 *** dccb1 0,88929 *** 0,85095 *** 0,85290 *** mshape 5,25560 *** 5,05504 *** 5,07079 *** In e al B: P e-COVID-19 (Janua y 5, 2016, o Janua y 23, 2020) mu 0,00072*** 0,00069*** -0,00202*** -0,00207*** a 1 -0,98386*** 0,89013*** 0,89743*** 0,88094*** ma1 0,99404*** -0,91941*** -0,98258*** -0,97340*** omega 0,00001*** 0,00000 0,00165** 0,00131** alpha1 0,14919*** 0,15629** 0,13187** 0,13571*** be a1 0,73270*** 0,79554*** 0,58204*** 0,58696*** shape 6,23262*** 4,45051*** 4,13100*** 4,47376*** dcca1 0,02638** 0,03193* 0,02366 dccb1 0,89077*** 0,86290*** 0,77044*** mshape 5,99599*** 4,97099 5,27980*** In e al C: Du ing COVID-19 (Janua y 24, 2020, o May 5, 2023) mu 0,00115*** 0,00116*** -0,00550*** -0,00597*** a 1 0,43800** 0,35614* 0,56564*** 0,58181*** ma1 -0,53108*** -0,45583** -0,70558*** -0,70600*** omega 0,00001** 0,00001 0,00010 0,00009 alpha1 0,10927*** 0,17879*** 0,05428*** 0,06326*** be a1 0,84601*** 0,82020*** 0,93334*** 0,92752*** shape 5,77550*** 3,77321*** 4,09054*** 3,64227*** dcca1 0,07553*** 0,05139*** 0,05852*** dccb1 0,87301*** 0,86181*** 0,85152*** mshape 4,79235*** 5,21968*** 4,70195*** In e al D: Pandemic-Wa pe iod (Janua y 24, 2020, o Ma ch 27, 2024) mu 0,00119*** 0,00106*** -0,00235 -0,00304*** a 1 0,39749** 0,43635** 0,63472*** 0,91642*** ma1 -0,48883*** -0,51604*** -0,75888*** -0,97543*** omega 0,00001 0,00000 0,00004 0,00006 alpha1 0,10810 0,14620** 0,04981*** 0,06896*** be a1 0,85384*** 0,84787*** 0,94483*** 0,92499*** shape 6,35171*** 3,93751*** 4,36091*** 3,75488*** dcca1 0,07188*** 0,04203** 0,04609*** dccb1 0,88166*** 0,87768*** 0,87235*** mshape 5,05991*** 5,74517*** 5,09007*** 28 Coe icien es ima es de i ed om uni a ia e GARCH and Dynamical Condi ional Co ela ions (DCC). Uni a ia e GARCH p ocesses a e calcula ed o each index (wi h pa ame e s mu, omega, alpha1, be a1), be o e es ima ing he DCC pa ame e s dcca1 and dccb1. The indices conside ed include all h ee pai s o Impac In es ing Fi ms Index (IIFX) agains each o STOXX Eu ope 600 index (STOXX), EURO STOXX 50 Vola ili y index (V2TX) and DAX-New Vola ili y index (V1XI). The use o he mul i a ia e - dis ibu ion allowed he model o accoun o hea y ails and ex eme alues in he da a, which a e common in inancial da a, leading o mo e obus and ealis ic modelling o inancial e u ns. ***, **, * indica e s a is ical signi icance a he 1%, 5% and 10% le els, espec i ely. SHORT-TERM VOLATILITY CONTAGION BETWEEN EUROPEAN IMPACT FIRMS AND EUROPEAN MARKET’S VOLATILITY (V2TX/V1XI) Sho - e m ola ili y con agion desc ibes he immedia e and signi ican esponse o ola ili y in one ma ke —in his case, EU impac i ms— o changes in ola ili y in ano he ma ke , ep esen ed by Eu opean ma ke ola ili y indices V2TX and V1XI. The DCC-GARCH model cap u es his phenomenon using he alpha 1 coe icien s (impac o pas shocks on cu en ola ili y es ima e) and he dcca1 coe icien s (sho - e m dynamics in he co ela ion be ween he Eu opean Impac In es ing Fi ms Index (IIFX) and he V2TX/V1XI indices). Du ing he ull sample pe iod, he indings using he DCC-GARCH model, as shown in Table 4, sugges a mode a e esponse o ecen shocks in bo h V2TX and V1XI, indica ing ha 7.253% and 9.254% o he ecen shocks pe sis in he ola ili y dynamics o hese indices. The sho - e m con agion wi h EU impac i ms, as measu ed by he dcca1 alues o 0.04994 o V2TX and 0.04629 o V1XI, appea s o be sligh bu s a is ically signi ican . This means ha sudden changes in Eu opean ma ke ola ili y led o a sligh sensi i i y in he ola ili y o EU impac i ms. The ela i ely low dcca1 alues sugges ha while he e is some sho - e m con agion, EU impac i ms a e no highly eac i e o ex eme mo emen s in Eu opean ma ke ola ili y, indica ing po en ial di e si ica ion bene i s. In he p e-COVID-19 pe iod, he alpha1 coe icien s o 0,13187 o V2TX and 0,13571 o V1XI e eal a highe deg ee o sho - e m ola ili y pe sis ence. This means ha in s able imes, pas shocks ha e a s onge in luence on cu en ola ili y es ima es. The dcca1 alues o 0,03193 o V2TX and 0,02366 o V1XI sugges weake sho - e m co ela ions wi h EU impac i ms. This indica es ha EU impac i ms we e no highly sensi i e o changes in Eu opean ma ke ola ili y. This weake sensi i i y sugges s po en ial o di e si ica ion du ing s able pe iods. Howe e , he s a is ical signi icance o hese alues is no s ong o all measu es, implying ha he obse ed ela ionship migh be due o andom chance a he han a obus co ela ion. Du ing he COVID-19 pe iod, he e is a educed esponse o ecen shocks, e ealed by he alpha1 coe icien s o 0,05428 o V2TX and 0,06326 o V1XI, indica ing ha immedia e pas shocks had a lesse impac on cu en ola ili y. This could e lec he unique ma ke condi ions du ing he pandemic, whe e ola ili y was d i en mo e by ongoing unce ain y han by ecen shocks. Addi ionally, he dcca1 alues o 0,05139 o V2TX and 0,05852 o V1XI 29 sugges sligh ly highe sho - e m con agion compa ed o he p e-COVID-19 pe iod. This indica es ha EU impac i ms became somewha mo e esponsi e o ecen changes in Eu opean ma ke ola ili y du ing he pandemic. The inc eased con agion e lec s a mo e in e connec ed and esponsi e ma ke en i onmen , ye he ela i ely low alues s ill indica e po en ial di e si ica ion bene i s as EU impac i ms we e no excessi ely sensi i e o ma ke ola ili y. In he pandemic-wa pe iod, which encompasses bo h he COVID-19 pandemic and he ongoing wa , he alpha1 alues 0,04981 o V2TX and 0,06896 o V1XI sugges a mino impac o immedia e pas shocks on cu en ola ili y. This con inued mino in luence e lec s he p olonged pe iod o unce ain y whe e ola ili y was in luenced mo e by ongoing c ises a he han ecen shocks. The dcca1 alues o 0,04203 o V2TX and 0,04609 o V1XI indica e ha he sho - e m sensi i i y o EU impac i ms o ma ke ola ili y emained sligh bu s a is ically signi ican . Despi e he ex ended c isis pe iod, impac i ms exhibi ed ela i ely s able beha iou in he ace o Eu opean ma ke ola ili y, indica ing hei esilience. The sligh ly highe dcca1 alues compa ed o p e-COVID-19 sugges a main ained, ye s ill mode a e, sensi i i y o ma ke ola ili y. O e all, Eu opean impac i ms show a ying deg ees o sensi i i y o ma ke ola ili y ac oss di e en pe iods. Du ing s able, p e-COVID-19 pe iods, hei ola ili y was less co ela ed wi h b oade ma ke ola ili y, p o iding po en ial di e si ica ion bene i s. Howe e , du ing imes o c isis, such as he COVID-19 pandemic and he ongoing geopoli ical con lic , hei sho - e m sensi i i y inc eased, e lec ing a mo e in e connec ed ma ke en i onmen . Despi e he inc eased sho - e m sensi i i y du ing c ises, he ela i ely low dcca1 alues indica e ha Eu opean impac i ms s ill o e some di e si ica ion bene i s. This analysis suppo s he s a egic inclusion o Eu opean impac i ms in in es men po olios o mi iga e isk and enhance s abili y in he sho - e m, especially du ing s able pe iods. Howe e , i also highligh s he necessi y o ongoing esea ch o con i m hei di e si ica ion and isk-hedging bene i s du ing di e en ma ke condi ions. LONG-TERM VOLATILITY CONTAGION BETWEEN EUROPEAN IMPACT FIRMS AND EUROPEAN MARKET’S VOLATILITY (V2TX/V1XI) Long- e m ola ili y con agion desc ibes he endu ing and sus ained esponse o ola ili y in EU impac i ms o changes in he Eu opean ma ke ola ili y indices. This can be assessed h ough he be a1 and dccb1 coe icien s. Be a1 cap u es he pe sis ence o shocks o e ime, while dccb1 ep esen s he long- e m dynamics in he co ela ion be ween Eu opean impac i ms and ma ke ola ili y indices (V2TX/V1XI). Du ing he ull sample pe iod, he be a1 coe icien s o 0,89916 o V2TX and 0,87378 o V1XI sugges a high pe sis ence o ola ili y in hese indices o e ime. This indica es ha ola ili y shocks ha e a long-las ing impac , e lec ing sus ained ma ke unce ain y and in es o 30 sen imen . Co espondingly, he dccb1 alues o 0,85095 o V2TX and 0,85290 o V1XI imply a s ong long- e m con agion be ween EU impac i ms and he b oade ma ke ola ili y indices. This sugges s ha he long- e m ola ili y dynamics o EU impac i ms a e signi ican ly in luenced by he b oade Eu opean ma ke ola ili y. This high pe sis ence and s ong long- e m con agion imply ha in he long un, EU impac i ms a e mo e likely o ollow he ends o expec ed ma ke ola ili y. In he p e-COVID-19 pe iod, he be a1 alues o 0,58204 o V2TX and 0,58696 o V1XI a e signi ican ly lowe han in he ull sample pe iod. This indica es ha du ing s able imes, he long- e m pe sis ence o ola ili y in hese indices was educed, e lec ing a mo e s able ma ke en i onmen . The dccb1 alue o 0,86290 o V2TX and 0.77044 o V1XI sugges a s ong long- e m co ela ion wi h EU impac i ms, al hough sligh ly weake han in he ull sample pe iod. This indica es ha e en du ing s able pe iods, EU impac i ms we e in luenced by he ma ke 's long- e m ola ili y, bu o a lesse ex en han du ing c isis pe iods. The high dccb1 alues imply ha while EU impac i ms migh p o ide some di e si ica ion in he sho - e m, hei long- e m ola ili y was e y closely aligned wi h ma ke ola ili y ends. Du ing he COVID-19 pe iod, he be a1 coe icien s inc eased signi ican ly o 0,93334 o V2TX and 0,92752 o V1XI, e lec ing e y high long- e m pe sis ence o ola ili y. This e lec s he sus ained unce ain y and heigh ened isk en i onmen du ing he pandemic. The dccb1 alues o 0,86181 o V2TX and 0,85152 o V1XI e eal a s ong long- e m con agion wi h EU impac i ms, simila o he ull sample pe iod. This high long- e m co ela ion indica es ha du ing he pandemic, EU impac i ms' ola ili y was closely ied o he Eu opean ma ke 's long- e m ola ili y, educing hei di e si ica ion bene i s in imes o p olonged c isis. In he pandemic-wa pe iod, he be a1 alues emained high a 0.94483 o V2TX and 0.92499 o V1XI, indica ing a pe sis ence o ecen shocks simila o ha obse ed du ing COVID-19. Co espondingly, he dccb1 alues o 0.87768 o V2TX and 0.87235 o V1XI indica e a e y s ong long- e m co ela ion dynamic, compa able o he COVID-19 pe iod. This high long- e m co ela ion du ing he p olonged pe iod o unce ain y unde sco es he con inued in luence o ma ke ola ili y on EU impac i ms. The long- e m alignmen o hei ola ili y wi h ma ke ends sugges s limi ed di e si ica ion bene i s du ing ex ended pe iods o c isis. O e all, EU impac i ms exhibi s ong long- e m ola ili y con agion wi h Eu opean ma ke ola ili y indices ac oss all pe iods. Du ing s able, p e-COVID-19 pe iods, hei long- e m ola ili y pe sis ence was lowe bu s ill signi ican . Du ing imes o c isis, such as he COVID- 19 pandemic and he ongoing geopoli ical con lic , hei long- e m ola ili y closely ollowed he ma ke ola ili y ends, educing hei e ec i eness as a di e si ica ion ool. In summa y, he esul s o he DCC-GARCH model indica e ha while EU impac i ms o e some di e si ica ion bene i s in he sho - e m, hei s ong long- e m con agion wi h ma ke ola ili y sugges s limi ed isk-hedging bene i s du ing ex ended pe iods o c isis. Thei ola ili y emains closely aligned wi h ma ke ola ili y sen imen /expec ancy, highligh ing he 31 need o a di e si ied po olio s a egy ha includes o he asse s o e ec i ely manage long- e m isk. SHORT-TERM VOLATILITY CONTAGION BETWEEN EUROPEAN IMPACT FIRMS AND EUROPEAN MARKET’S BENCHMARK (STOXX) Du ing he ull sample pe iod, he indings sugges a high esponse o ecen shocks, wi h an alpha1 alue o 0,14872, indica ing ha app oxima ely 14.872% o ecen shocks pe sis in he ola ili y dynamics o he STOXX Index. This pe sis ence is highe han ha obse ed o V2TX and V1XI, indica ing a s onge endency o STOXX ola ili y o ca y o e pas shocks. The sho - e m con agion wi h EU impac i ms, as measu ed by he dcca1 alue o 0.05724, appea s o be sligh bu s a is ically signi ican . This implies a modes sensi i i y o EU impac i ms o sudden changes in he b oade Eu opean ma ke , highligh ing hei po en ial o p o iding di e si ica ion bene i s. Al hough con agion was compa able o V2TX and V1XI, he highe dcca1 alue alida ed he s onge ela ionship be ween EU impac i ms and STOXX. In he p e-COVID-19 pe iod, he alpha1 coe icien o 0.15629 e eals a sligh ly highe deg ee o sho - e m ola ili y pe sis ence compa ed o he ull sample pe iod. The dcca1 alue o 0.02638 indica es a s a is ically signi ican bu modes sho - e m co ela ion be ween EU impac i ms and STOXX. This alue, simila o he con agion le els wi h V2TX and V1XI, sugges s ha EU impac i ms we e no highly sensi i e o changes in he b oade Eu opean ma ke du ing his ime. The e o e, EU impac i ms o e ed no able di e si ica ion bene i s, mo ing independen ly om he b oade ma ke . Du ing he COVID-19 pe iod, he esponse o ecen shocks inc eased, as e idenced by he alpha1 coe icien o 0.17879. This high pe sis ence o ola ili y is g ea e han he alpha1 alues o V2TX and V1XI, indica ing ha STOXX expe ienced mo e endu ing ola ili y du ing he pandemic. The dcca1 alue o 0.07553 sugges s ha EU impac i ms we e mo e esponsi e o changes in he b oade Eu opean ma ke compa ed o V2TX and V1XI. This inc eased sho - e m con agion implies ha he di e si ica ion bene i s o impac in es men s we e lowe du ing he pandemic. Howe e , hei ela i e s abili y s ill p o ided some le el o esilience, unde sco ing he in e connec ed na u e o ma ke s du ing c ises. In he pandemic-wa pe iod, he alpha1 alue o 0.14620 indica es a sligh ly educed pe sis ence o ola ili y shocks compa ed o he p e-COVID-19 pe iod. Fu he mo e, he dcca1 alue o 0.07188 e eals ha al hough EU impac i ms exhibi ed highe sho - e m sensi i i y o changes in STOXX ola ili y compa ed o V2TX and V1XI, hei ola ili y emained ela i ely s able. This s abili y du ing a p olonged pe iod o unce ain y sugges s ha EU impac i ms main ained po en ial di e si ica ion bene i s, exhibi ing esilience despi e he heigh ened in e connec edness wi h he b oade ma ke . In summa y, while EU impac i ms p o ide some di e si ica ion bene i s du ing s able pe iods, hei inc eased sho - e m sensi i i y o STOXX ola ili y du ing c ises sugges s limi ed isk- 32 hedging bene i s in he sho - e m. In es o s seeking o use EU impac i ms o di e si ica ion should conside he heigh ened co ela ion wi h b oade Eu opean ma ke du ing imes o c isis. LONG-TERM VOLATILITY CONTAGION BETWEEN EUROPEAN IMPACT FIRMS AND EUROPEAN MARKET’S BENCHMARK (STOXX) Du ing he ull sample pe iod, he be a1 coe icien o STOXX is 0.83374, lowe han he es ima es o V2TX and V1XI. This indica es ha while he pe sis ence o ma ke shocks in STOXX is subs an ial, i is somewha less p onounced han in he o he wo indices. Addi ionally, he dccb1 alue o 0.88929 sugges s a s ong long- e m co ela ion be ween EU impac i ms and STOXX, implying signi ican in luence by long- e m ola ili y ends in he b oade Eu opean ma ke and a high le el o in e connec edness. In he p e-COVID-19 pe iod, he be a1 alue is 0.79554, indica ing sligh ly lowe long- e m ola ili y pe sis ence compa ed o he ull sample pe iod, al hough highe han he es ima es o V2TX and V1XI. This sugges s ha he pe sis ence o ma ke shocks was mo e endu ing in STOXX compa ed o he o he indices. The dccb1 alue o 0.89077 e eals a s ong long- e m co ela ion wi h EU impac i ms, simila o he ull sample pe iod. This indica es ha e en du ing s able pe iods, EU impac i ms we e s ill signi ican ly in luenced by he long- e m ola ili y ends in he b oade Eu opean ma ke . Du ing he COVID-19 pe iod, he be a1 coe icien inc eased o 0.82020, which is lowe han he alues o V2TX and V1XI, bu i s ill e lec s a high le el o pe sis ence in ola ili y es ima es o e he long e m. The dccb1 alue o 0.87301 indica es a s ong long- e m co ela ion wi h EU impac i ms, hough sligh ly lowe han he p e-COVID-19 pe iod. This sugges s ha while EU impac i ms we e in luenced by long- e m ends in STOXX ola ili y du ing he pandemic, hei co ela ion was sligh ly educed and he e ec s less p onounced. In he ull pandemic-wa pe iod, he be a1 coe icien is 0.84787, indica ing e y high long- e m ola ili y pe sis ence. The dccb1 alue o 0.88166, compa able o V2TX and V1XI, e eals a e y s ong long- e m co ela ion dynamic. This sugges s ha du ing pe iods o p olonged unce ain y, EU impac i ms emain highly in e connec ed wi h he b oade ma ke , educing hei po en ial o p o iding di e si ica ion bene i s. O e all, he analysis o long- e m ola ili y con agion be ween EU impac i ms and STOXX e eals a pa e n simila o ha obse ed wi h V2TX and V1XI. The be a1 and dccb1 coe icien s indica e ha EU impac i ms a e hea ily in luenced by long- e m ola ili y ends in he b oade ma ke . This educes hei po en ial o p o iding di e si ica ion bene i s o e he long e m, as hei pe o mance is closely ied o he b oade ma ke 's long- e m ends. 33 Table 5 - Es ima ion esul s o Copula-DCC-GARCH model STOXX V2TX V1XI In e al A: Full sample pe iod (Janua y 5, 2016, o Ma ch 27, 2024) dcca1 0,05768*** 0,04982*** 0,04799*** dccb1 0,88277*** 0,84471*** 0,84092*** mshape 5,20945*** 7,66998*** 7,07260*** In e al B: P e-COVID-19 (Janua y 5, 2016, o Janua y 23, 2020) dcca1 0,02796** 0,03164* 0,02405 dccb1 0,88473*** 0,85604*** 0,77232*** mshape 6,37226*** 5,37302*** 5,48550*** In e al C: Du ing COVID-19 (Janua y 24, 2020, o May 5, 2023) dcca1 0,08002*** 0,05316*** 0,05911*** dccb1 0,86273*** 0,85805*** 0,85198*** mshape 5,52369*** 12,25248** 8,92788*** In e al D: Pandemic-Wa pe iod (Janua y 24, 2020, o Ma ch 27, 2024) dcca1 0,07006*** 0,04550*** 0,04919*** dccb1 0,88069*** 0,86967*** 0,86887*** mshape 5,22330*** 13,64738** 10,43416** The esul s o he Copula-DCC-GARCH and DCC-GARCH models indica e ha while bo h models yield simila dynamic condi ional co ela ions (DCC) be ween he asse e u ns, he Copula-DCC-GARCH model demons a es a be e o e all i . This is e iden om be e alues in me ics such as In o ma ion C i e ia (AIC, BIC, Shiba a, and HQC), Log-likelihood, Mean Squa ed E o (MSE), and Roo Mean Squa ed E o (RMSE). The es ima ed pa ame e s (dcca1 and dccb1) o bo h models show close co espondence ac oss di e en pe iods, indica ing simila co ela ion dynamics. Howe e , he shape pa ame e s (mshape) and o he i indices consis en ly a ou he Copula-DCC-GARCH model, sugges ing i cap u es he da a cha ac e is ics mo e accu a ely. The imp o emen in model i o he Copula-DCC-GARCH model is pa icula ly no iceable when examining he s a is ical signi icance and he magni udes o he es ima ed coe icien s. Fo ins ance, du ing he COVID-19 and Pandemic-Wa pe iods, he mshape alues a e ma kedly highe in he Copula-DCC-GARCH model, indica ing a be e accommoda ion o he da a's hea y- ailed na u e and ex eme e en s. Consequen ly, i indica es ha he Copula- DCC-GARCH model p o ides mo e p ecise and eliable o ecas s o ola ili y and co ela ions compa ed o he adi ional DCC-GARCH model (see Appendix E). 40 Bu on, M. D., Cole, S. A., De , A., Ja ymowycz, C., Jeng, L., Le ne , J., Mashwama, F., Xu, C., & Zochowski, R. (2021). The P ojec on Impac In es men s’ Impac In es men Da abase (SSRN Schola ly Pape 3930171). h ps://doi.o g/10.2139/ss n.3930171 Celık, S. (2012). The mo e con agion e ec on eme ging ma ke s: The e idence o DCC-GARCH model. Economic Modelling, 29(5), 1946–1959. h ps://doi.o g/10.1016/j.econmod.2012.06.011 Chiappini, H., Ma inelli, N., Jalal, R. N.-U.-D., & Bi indelli, G. (2023). Pas , p esen and u u e o impac in es ing and closely ela ed inancial ehicles: A li e a u e e iew. Sus ainabili y Accoun ing, Managemen and Policy Jou nal, 14(7), 232–257. h ps://doi.o g/10.1108/SAMPJ-09-2022-0471 Chowdh y, B., Da ies, S. W., & Wa e s, B. (2019). In es ing o Impac . The Re iew o Financial S udies, 32(3), 864–904. h ps://doi.o g/10.1093/ s/hhy068 Cole, S., Melecky, M., Mölde s, F., & Reed, T. (2020). Long- un Re u ns o Impac In es ing in Eme ging Ma ke s and De eloping Economies (Wo king Pape 27870). Na ional Bu eau o Economic Resea ch. h ps://doi.o g/10.3386/w27870 Engle, R. (2002). Dynamic condi ional co ela ion: A simple class o mul i a ia e gene alized au o eg essi e condi ional he e oskedas ici y models. Jou nal o Business and Economic S a is ics, 20(3), 339–350. h ps://doi.o g/10.1198/073500102288618487 Engle, R., & Sheppa d, K. (2001). Theo e ical and Empi ical p ope ies o Dynamic Condi ional Co ela ion Mul i a ia e GARCH (w8554; p. w8554). Na ional Bu eau o Economic Resea ch. h ps://doi.o g/10.3386/w8554 Geczy, C., Je e s, J. S., Mus o, D. K., & Tucke , A. M. (2021). Con ac s wi h (Social) bene i s: The implemen a ion o impac in es ing. Jou nal o Financial Economics, 142(2), 697–718. h ps://doi.o g/10.1016/j.j ineco.2021.01.006 Giacomini, E., Ma inelli, N., & Ricce i, L. (2023). The impac a s ake: Risk and e u n in publicly lis ed social impac i ms. Business E hics, he En i onmen & Responsibili y, 32(2), 713–741. h ps://doi.o g/10.1111/bee .12511 God ey, P. C., Me ill, C. B., & Hansen, J. M. (2009). The ela ionship be ween co po a e social esponsibili y and sha eholde alue: An empi ical es o he isk managemen hypo hesis. S a egic Managemen Jou nal, 30(4), 425–445. h ps://doi.o g/10.1002/smj.750 Hand, D., & Gilbe , S. (2023). Holis ic po olio cons uc ion wi h an impac lens. The Global Impac In es ing Ne wo k (GIIN). h ps:// hegiin.o g/asse s/Holis ic%20po olio%20cons uc ion%20wi h%20an%20impac % 20lens.pd 41 Hand, D., Sunde ji, S., Di h ich, H., & No a, N. (2020). GIIN Annual Impac In es o Su ey 2020. The Global Impac In es ing Ne wo k (GIIN). h ps:// hegiin.o g/asse s/GIIN%20Annual%20Impac %20In es o %20Su ey%202020.pd Hand, D., Sunde ji, S., & Pa do, N. M. (2023). 2023 GIINsigh – Impac In es ing Alloca ions, Ac i i y & Pe o mance. The Global Impac In es ing Ne wo k (GIIN). h ps:// hegiin.o g/asse s/documen s/pub/2023- GIINsigh /2023%20GIINsigh %20%E2%80%93%20Impac %20In es ing%20Alloca ions,%20Ac i i y%20&%20Pe o mance.pd Haq, I. U., Chup adi , S., & Huo, C. (2021). Do G een Bonds Ac as a Hedge o a Sa e Ha en agains Economic Policy Unce ain y? E idence om he USA and China. In e na ional Jou nal o Financial S udies, 9(3), A igo 3. h ps://doi.o g/10.3390/ij s9030040 Höchs äd e , A. K., & Scheck, B. (2015). Wha ’s in a Name: An Analysis o Impac In es ing Unde s andings by Academics and P ac i ione s. Jou nal o Business E hics, 132(2), 449–475. h ps://doi.o g/10.1007/s10551-014-2327-0 Hua Fan, J., & Michalski, L. (2020). Sus ainable ac o in es ing: Whe e doing well mee s doing good. In e na ional Re iew o Economics & Finance, 70, 230–256. h ps://doi.o g/10.1016/j.i e .2020.07.013 Je e s, J., Lyu, T., & Posenau, K. (2023). The Risk and Re u n o Impac In es ing Funds (SSRN Schola ly Pape 3949530). h ps://doi.o g/10.2139/ss n.3949530 Kim, Y., Li, H., & Li, S. (2014). Co po a e social esponsibili y and s ock p ice c ash isk. Jou nal o Banking & Finance, 43, 1–13. h ps://doi.o g/10.1016/j.jbank in.2014.02.013 Lee, M., Adbi, A., & Singh, J. (2020). Ca ego ical cogni ion and ou come e iciency in impac in es ing decisions. S a egic Managemen Jou nal, 41(1), 86–107. h ps://doi.o g/10.1002/smj.3096 Lenzi, D. (2021). Co po a e Social Bonds: A Legal Analysis. Eu opean Company and Financial Law Re iew, 18(2), 291–320. h ps://doi.o g/10.1515/ec -2021-0014 Liu, B., Ju, T., & Gao, S. S. S. (2021). The combined e ec s o inno a ion and co po a e social esponsibili y on i m inancial isk. Jou nal o In e na ional Financial Managemen & Accoun ing, 32(3), 283–310. h ps://doi.o g/10.1111/ji m.12135 Ma ens, M., & Poon, S.-H. (2001). Re u ns synch oniza ion and daily co ela ion dynamics be ween in e na ional s ock ma ke s. Jou nal o Banking & Finance, 25(10), 1805–1827. h ps://doi.o g/10.1016/S0378-4266(00)00159-X Miglia acca, M., Pa el, R., Pal inie i, A., & Goodell, J. W. (2022). Mapping impac in es ing: A bibliome ic analysis. Jou nal o In e na ional Financial Ma ke s, Ins i u ions and Money, 81, 101679. h ps://doi.o g/10.1016/j.in in.2022.101679 42 Mo duch, J., & Ogden, T. (2019). The challenges o social in es men h ough he lens o mic o inance. Em M. Hudon, M. Labie, & A. Sza a z (Eds.), A Resea ch Agenda o Financial Inclusion and Mic o inance. Edwa d Elga Publishing. h ps://doi.o g/10.4337/9781788114226.00009 Mudalia , A., & Bass, R. (2017). E idence on he Financial Pe o mance o Impac In es men s. h ps:// hegiin.o g/asse s/2017_GIIN_FinancialPe o manceImpac In es men s_Web.pd Nakai, M., Yamaguchi, K., & Takeuchi, K. (2016). Can SRI unds be e esis global inancial c isis? E idence om Japan. In e na ional Re iew o Financial Analysis, 48, 12–20. h ps://doi.o g/10.1016/j.i a.2016.09.002 Nguyen, T. N., Phan, T. K. H., & Nguyen, T. L. (2022). Financial Con agion Du ing Global Financial C isis and Co id–19 Pandemic: The E idence om Dcc–Ga ch Model. Cogen Economics & Finance, 10(1), 2051824. h ps://doi.o g/10.1080/23322039.2022.2051824 No singe , J. R., Sulaeman, J., & Va ma, A. (2019). Ins i u ional in es o s and co po a e social esponsibili y. Jou nal o Co po a e Finance, 58, 700–725. h ps://doi.o g/10.1016/j.jco p in.2019.07.012 No singe , J., & Va ma, A. (2014). Socially esponsible unds and ma ke c ises. Jou nal o Banking & Finance, 48, 180–193. h ps://doi.o g/10.1016/j.jbank in.2013.12.016 Pás o , Ľ., & Vo sa z, M. B. (2020). Mu ual Fund Pe o mance and Flows du ing he COVID-19 C isis. The Re iew o Asse P icing S udies, 10(4), 791–833. h ps://doi.o g/10.1093/ aps u/ aaa015 Pa on, A. J. (2006). Modelling Asymme ic Exchange Ra e Dependence*. In e na ional Economic Re iew, 47(2), 527–556. h ps://doi.o g/10.1111/j.1468-2354.2006.00387.x Pa on, A. J. (2012). A e iew o copula models o economic ime se ies. Jou nal o Mul i a ia e Analysis, 110, 4–18. h ps://doi.o g/10.1016/j.jm a.2012.02.021 Pizzi, S., Del Baldo, M., Capu o, F., & Ven u elli, A. (2022). Volun a y disclosu e o Sus ainable De elopmen Goals in manda o y non- inancial epo s: The mode a ing ole o cul u al dimension. Jou nal o In e na ional Financial Managemen & Accoun ing, 33(1), 83–106. h ps://doi.o g/10.1111/ji m.12139 Riedl, A., & Smee s, P. (2017). Why Do In es o s Hold Socially Responsible Mu ual Funds? The Jou nal o Finance, 72(6), 2505–2550. h ps://doi.o g/10.1111/jo i.12547 Sado sky, P. (2014). Modeling ola ili y and condi ional co ela ions be ween socially esponsible in es men s, gold and oil. Economic Modelling, 38, 609–618. h ps://doi.o g/10.1016/j.econmod.2014.02.013 43 Saeed, T., Bou i, E., & T an, D. K. (2020). Hedging S a egies o G een Asse s agains Di y Ene gy Asse s. Ene gies, 13(12), A igo 12. h ps://doi.o g/10.3390/en13123141 Skla , M. (1959). Fonc ions de épa i ion à N dimensions e leu s ma ges. Annales de l’ISUP, VIII(3), 229–231. To e, M. L., Chiappini, H., & Mango, F. (2017). Do Impac In es men s Con ibu e o Po olio Pe o mance? A P elimina y In es iga ion. Jou nal o Business and Economics, 8(10), 799–808. h ps://doi.o g/10.15341/jbe(2155-7950) Wa s, N., & Scales, I. R. (2020). Social impac in es ing, ag icul u e, and he inancialisa ion o de elopmen : Insigh s om sub-Saha an A ica. Wo ld De elopmen , 130, 104918. h ps://doi.o g/10.1016/j.wo ldde .2020.104918 Yousa , I., Suleman, M. T., & Demi e , R. (2022). G een in es men s: A luxu y good o a inancial necessi y? Ene gy Economics, 105, 105745. h ps://doi.o g/10.1016/j.eneco.2021.105745 Zhang, W., He, X., & Hamo i, S. (2022). Vola ili y spillo e and in es men s a egies among sus ainabili y- ela ed inancial indexes: E idence om he DCC-GARCH-based dynamic connec edness and DCC-GARCH -copula app oach. In e na ional Re iew o Financial Analysis, 83, 102223. h ps://doi.o g/10.1016/j.i a.2022.102223 44 APPENDIX A Table A1 Lis and summa y cha ac e is ics o impac i ms Company name Impac In es ing a eas Business ac i i y (b ie desc ip ion) Abb L d Na u al Capi al P o ides sus ainable and esou ce-e icien elec i ica ion, mo ion and obo ics, and au oma ion echnologies and solu ions o op imizing, elec i ying, and deca bonizing ope a ions, in o de o enhance ene gy e iciency, and a ain a low-ca bon socie y and ne -ze o u u e. I p o ides digi ally connec ed and enabled indus ial equipmen and sys ems o cus ome s in u ili ies, indus y, anspo , and in as uc u e sec o s. ACS Ac i idades de Cons uccion y Se icios SA Na u al Capi al Is a e e ence in he cons uc ion business: de eloping highways, ailways, ma i ime and ai po s p ojec s, also esiden ial buildings, social in as uc u e, and acili ies p ojec s; anspo concessions indus y: de eloping “g een ield” in as uc u es; and in as uc u es and en i onmen al se ices: such as s ee cleaning, was e collec ion and anspo a ion, was e ecycling, and o he s. Als om SA Na u al Capi al O e s p o en al e na i es o diesel on non-elec i ied ne wo ks by being he i s playe wi h hyd ogen-powe ed lee s in comme cial se ice, and by de eloping and p omo ing inno a i e and sus ainable (hyd ogen and ba e y-elec ic) mobili y and in as uc u e solu ions ha help educe ene gy consump ion. Anheuse - Busch Inbe SA Na u al Capi al P oduces, dis ibu es, and ma ke s bee , alcoholic be e ages, and so d inks made om simple ing edien s, g own locally, and b ewed wi h he aim o p ese e na u al esou ces. They pa ne wi h local a me s and in es in esea ch and de elopmen o sus ainable b ewing me hods. Also, ha e a business- o-business e-comme ce pla o m o empowe small and medium-sized e ail pa ne s and help g ow hei businesses. A hens Wa e Supply and Sewe age Company SA Basic Needs Responsible o managing he wa e supply and sewe age sys ems in he egion o A ica, G eece, p o iding clean and sa e d inking wa e o esiden s, businesses, and o he cus ome s wi hin i s se ice a ea, and collec ing and ea ing was ewa e o p o ec public heal h and he en i onmen . BAE Sys ems PLC Empowe men Is a de ense, ae ospace, and secu i y company ha de elops, enginee s, manu ac u es, and suppo s p oduc s and sys ems o deli e mili a y capabili y, p o ec na ional secu i y and people, and keep c i ical in o ma ion and in as uc u e secu e. I also d i es he de elopmen o sus ainable elec ic solu ions in he de ence and comme cial sec o . Ba a De elopmen s P L C Na u al Capi al Is a comme cial p ope y de elope and esiden ial cons uc ion con ac o whose p ima y business in ol es he cons uc ion, sale, and de elopmen o esiden ial p ope ies, inco po a ing ene gy-e icien ea u es, sus ainable building ma e ials, and en i onmen ally iendly p ac ices in o i s p ojec s o minimize en i onmen al impac and mee sus ainabili y goals. Beie sdo AG Basic Needs De elops and ma ke s pe sonal ca e p oduc s and sel -adhesi e p oduc solu ions, such as ace ca e, body ca e, sun p o ec ion, and wound ca e, conside ing he deca bonisa ion o hei ope a ions and anspo by emb acing enewable ene gy. I commi s o sou cing aw ma e ials om a ious o igins, including palm oil, shea bu e and pape -based ma e ials, and p omo es was ewa e ecycling ac oss hei p oduc ion. Ma ke s i s p oduc s unde NIVEA, Euce in, and o he b ands. Boi on SA Basic Needs Engages in he p o ision o homeopa hic medicines p o iding sa e and quali y medicines and d ugs ha a e used as al e na i e o 45 complemen a y ea men s o a ious heal h condi ions. Commi s o sou cing aw ma e ials om a ious o igins, including plan s, mine als/chemicals, and animals (e.g. enom). Gi es p e e ence o sho channels and all cul i a ed plan s a e GM- ee. BT G oup PLC Empowe men Is a elecommunica ions and ne wo k p o ide which p o ides ixed and mobile elecommunica ions and ela ed secu e digi al p oduc s, solu ions and se ices powe ed by enewable elec ici y in hei jou ney o ne ze o emissions and a ci cula wo ld. Has empowe ed people and has accele a ed digi al ans o ma ion h ough hei esponsible, inclusi e, and sus ainable echnology. Cen ica PLC Clima e Change Ope a es as an in eg a ed ene gy company ha ca ies ou elec ici y gene a ion h ough combined cycle gas u bine (CCGT) powe s a ions, and nuclea powe plan s, and i deli e s in eg a ed ene gy solu ions, helping o ganiza ions balance he demands o p o i abili y and sus ainabili y by p o iding ailo ed ze o ca bon and enewable ene gy supply op ions. Coloplas A/S Basic Needs De elops, manu ac u es, and ma ke s medical de ices and se ices ha make li e easie o people wi h e y pe sonal and p i a e medical condi ions, like os omy ca e, u ology, con inence, and wound ca e. I uses aw ma e ials such as enewables and p omo es p oduc and packaging ecyclabili y. Co icei a Amo im SGPS SA Na u al Capi al Leading specialis in co k p ocessing, ha ing a po olio o co k p oduc s like co k loo ing, insula ion, and o he co k-based solu ions wi h high added alue. Focuses on sus ainable co k p oduc ion and inno a i e applica ions by o e ing eco- iendly solu ions ac oss a ious sec o s such as a chi ec u e and design, cons uc ion and in as uc u e, mobili y, ene gy, ae ospace, and o he s. Danone SA Basic Needs Is a manu ac u e and p oduce o essen ial dai y and plan -based p oduc s, ea ly li e nu i ion, medical nu i ion and wa e s ha p io i ize people's heal h and well-being. I dis ibu es i s p oduc s h ough e ail chains, adi ional ma ke ou le s and specialized dis ibu ion channels, including hospi als, clinics, and pha macies. Dassaul Sys emes SE Empowe men Ope a es as a so wa e company, which p o ides a 3Dexpe ience pla o m o c ea e inno a i e new p oduc s and se ices using i ual expe iences, and in his way model he social and en i onmen al impac s o e e y inno a ion and e alua e how i imp o es he li es o o he people and hei communi ies. Deu sche Telekom AG Empowe men Is a leading global elecommunica ions company ha p o ides mobile ixed-line, in e ne , and IT se ices o consume s and businesses and o e s sma p oduc s and se ices om enewable ene gy sou ces ha help educe ene gy consump ion and conse e esou ces. Also, empowe s p ojec s and ini ia i es wi h a posi i e con ibu ion o social and en i onmen al challenges in he digi al wo ld. EDP Reno a eis SA Clima e Change Is a enewable ene gy company ha concen a es on he p oduc ion o ene gy om enewable esou ces, and i s ac i i ies comp ise he de elopmen , ope a ion, and main enance o such elec ic powe s a ions as hyd oelec ic, wind, sola , idal, biomass and was e plan s, among o he s, educing ca bon emissions. EnBW Ene gie Baden Wue embe g AG Clima e Change T ansmi s, dis ibu es, and ades elec ic ene gy and na u al gas, and p o ides ene gy- ela ed se ices. I gene a es elec ici y om con en ional uels such as nuclea powe , as well as enewable sou ces such as wind, sola , and hyd oelec ic o educe i s ca bon oo p in and inc ease i s sha e o clean ene gy p oduc ion. Endesa SA Clima e Change Gene a es, dis ibu es, ansmi s, and sells ele ici y h ough he pa icipa ion in hyd o, he mal, nuclea , combined cycle, and enewable and cogene a ion plan s. Ope a es also in he na u al gas sec o and 46 p o ides o he alue-added se ices o wind a ms, mini hyd aulic, biomass, and was e ea men plan s. Enel SpA Clima e Change Gene a es, anspo s, and dis ibu es elec ici y and sells elec ic ene gy and na u al gas, g een ce i ica es, and CO2 emission igh s, wi h a s ong emphasis on enewable ene gy sou ces like wind, sola , geo he mal and hyd opowe , p o iding he solu ions o powe elec ic ehicles and sma ci ies as i pa icipa es in enewables ma ke s. Fage hul G oup AB Clima e Change Designs, manu ac u es, and sells a wide ange o ligh ing ix u es and sys ems ha ca e s o di e en ligh ing equi emen s and ene gy e iciency goals, minimizing ene gy consump ion and en i onmen al impac , o a ious p o essional applica ions, including o ices, schools, hospi als, e ail spaces, and indus ial acili ies, p o iding he e o e sus ainable solu ions o u ban a eas. Fo um Oyj Clima e Change P o ides elec ici y, gas, hea ing, and cooling se ices. The gene a ion o elec ici y comp ises nuclea , hyd o, wind, and he mal powe gene a ion, as well as powe po olio op imisa ion, ading, indus ial in elligence, and global nuclea se ices. I 's esponsible o sus ainable solu ions o u ban a eas. F esenius SE & Co KGaA Basic Needs De elops and manu ac u es a wide a ie y o heal h ca e p oduc s, which includes dialysis and non-dialysis p oduc s. P o ides dialysis ca e and ela ed se ices o pe sons su e ing om end-s age enal disease (ESRD), as well as o he heal hca e se ices o hospi als and medical ca e o pa ien s a home. Gebe i AG Na u al Capi al P o ides sani a y p oduc s and sys ems including auce s and lushing sys ems, and was e i ings and aps, ce amics and pipes and i ings o he esiden ial, comme cial, new cons uc ion and eno a ion ma ke s conside ing ecological and social aspec s. Con ibu es o sus ainable cons uc ion o buildings by p o iding wa e e iciency solu ions and sa ing ma e ials. Ge esheime AG Na u al Capi al De elops and manu ac u es special y and inno a i e p oduc s o d ug deli e y, such as insulin pens, inhale s, and p e illable sy inges, and i p oduces d ug packaging o glass o medicines and cosme ics, such as pha maceu icals ja s, ampoules, and o he s o he pha maceu ical, bio ech, and heal hca e indus ies. G i ols SA Basic Needs De elops, manu ac u es, and dis ibu es a b oad ange o biological medicines based on plasma de i ed p o eins ha enable pa ien s wi h a e and ch onic diseases o lead heal hie li es. I helps ensu e he sa e y o he wo ld’s blood supply, p o ides biological ma e ials, including plasma, o li e-science esea ch and u he manu ac u ing, and has a comp ehensi e po olio o ools, in o ma ion and se ices ha enable hospi als o e icien ly deli e medical ca e. Ind a Sis emas SA Empowe men P o ides consul ing and echnology solu ions such as sys ems in eg a ion and ou sou cing se ices, ins alls ailway managemen and icke ing sys ems, ins alls oad a ic oll and signalling sys ems, cons uc s sa elli e communica ions sys ems, manu ac u es ada ai de ence sys ems, and ins alls insu ance managemen sys ems o ae ospace, de ence & secu i y, and mobili y sec o s. Kingspan G oup PLC Na u al Capi al Is a global leade in high-pe o mance insula ion and building en elope solu ions enabling low-ene gy and low-ca bon buildings. I also specializes in he manu ac u e and dis ibu ion o sus ainable building ma e ials, including insula ion boa ds, insula ed panels, enewable ene gy sys ems, and wa e managemen solu ions. Leg and SA Na u al Capi al Is an elec ical and digi al building in as uc u e company o esiden ial, comme cial, and indus ial buildings ha o e s a wide ange o p oduc s and solu ions o elec ical ins alla ions, da a ne wo ks, building 47 au oma ion, and con ol sys ems ha a e simple, inno a i e, and sus ainable. Na u gy Ene gy G oup SA Clima e Change Is an in eg a ed gas and elec ic u ili y company ha ca ies ou he explo a ion, de elopmen , lique ac ion, egasi ica ion, anspo a ion, s o age, dis ibu ion, and comme cializa ion o na u al gas. Also, he gene a ion, dis ibu ion, and comme cializa ion o elec ici y h ough combined cycle, he mal, nuclea , hyd o, co-gene a ion and wind a m plan s, and o he special egime echnologies. NOS SGPS SA Empowe men Engages in he elecommunica ions, media and en e ainmen businesses o e ing s a e-o - he-a solu ions o ixed and mobile elephony, ele ision, In e ne , oice, and da a o all ma ke sec o s. Ope a es in he a ea o pay ele ision, nex -gene a ion b oadband se ices, and ilm dis ibu ion, and has been in es ing in ene gy-e icien echnologies, ansi ioning o enewable ene gy sou ces, and implemen ing ini ia i es o educe was e gene a ion and p omo e ecycling. No a is AG Basic Needs De elops, manu ac u es, and ma ke s pha maceu ical and consume heal hca e p oduc s such as b anded and gene ic p esc ip ion d ugs, ac i e pha maceu ical ing edien s (APIs), biosimila s and oph halmic p oduc s. I uses inno a i e science and digi al echnologies o de elop ans o ma i e ea men s and pa en -p o ec ed p esc ip ion medicines in a eas o g ea medical need. No o No disk A/S Basic Needs De elops, p oduces, and ma ke s pha maceu ical p oduc s ocused on diabe es ca e and o e s insulin deli e y sys ems and o he diabe es p oduc s. I also wo ks in he he apy a eas o haemophilia ca e, g ow h ho mone he apy and ho mone eplacemen he apy. No ozymes A/S Na u al Capi al Is a bio echnology company engaged in he p oduc ion and sale o indus ial enzymes, mic oo ganisms, and biopha maceu ical ing edien s o use in ag icul u e, bioene gy, ood & be e ages, household ca e, lea he , pha maceu icals, o es p oduc s, ex ile, and was ewa e ea men . Pea son PLC Empowe men Ope a es in he educa ion, assessmen , and ce i ica ions ma ke s, and p o ides digi al con en , lea ning expe iences, assessmen s, quali ica ions, and da a in he lea ning ma ke . I publishes ac oss he cu iculum and p o ides a ange o educa ion se ices, including eache de elopmen , educa ional so wa e, and sys em-wide solu ions. Also owns and ope a es schools and colleges. P oximus NV Empowe men Engaged in he elecommunica ion, in o ma ion, communica ion, and echnology (ICT) indus ies. I ans o ms a ange o echnologies, such as In e ne o Things (IoT), Big Da a, Cloud and Secu i y, in o new digi al ecosys ems and inno a i e solu ions, and p o ides eliable and secu e communica ions shaping a ai , us wo hy, and inclusi e digi al wo ld and accele a ing he ansi ion o a g een socie y. Rockwool A/S Na u al Capi al Is a manu ac u e and supplie o s one wool-based insula ion p oduc s such as loo insula ion, ceiling insula ion, in e nal wall insula ion ex e nal wall insula ion, HVAC insula ion, passi e i e p o ec ion, acous ic insula ion, and o he mine al wool p oduc s, o use in applica ions, including ic ion and wa e managemen while lowe ing emissions, imp o ing in as uc u e, o enabling mode n ho icul u e using ewe esou ces. Schneide Elec ic SE Clima e Change P o ides ene gy- ela ed solu ions ha combine ene gy managemen , au oma ion, and so wa e. Some o i s p oduc s include au oma ion and con ol p oduc s, low ol age p oduc s and sys ems, sola and ene gy s o age, medium ol age dis ibu ion and g id au oma ion, and c i ical powe , cooling, and acks. I also o e s ene gy and sus ainabili y se ices, ield and au oma ion, cloud, consul ing, and aining se ices. 48 Se e n T en PLC Basic Needs P o ides clean wa e and was ewa e se ices, and i 's also in ol ed in a eas ela ed o clima e change, p o ec ing wa e supplies, imp o emen o biodi e si y, and de eloping enewable ene gy solu ions, gene a ing enewable powe , h ough wind u bines, sola panels, and anae obic diges ion o specially g own c ops. SGL Ca bon SE Na u al Capi al Is a manu ac u e o p oduc s and solu ions based on ca bon ib es and special y g aphi e se ing se e al indus ies, such as s eel, aluminium, au omo i e, ae ospace, chemical, wind ene gy, and en i onmen al p o ec ion among o he s. I con ibu es o he de elopmen s o sus ainable mobili y, new ene gies, and c oss-indus y digi aliza ion. Siemens AG Na u al Capi al Ope a es in he elec i ica ion, au oma ion, and digi aliza ion ields, as i designs, de elops, and manu ac u es p oduc s, and ins alls complex sys ems and p ojec s, ocusing on powe gene a ion and dis ibu ion, and in elligen in as uc u e o buildings and dis ibu ed ene gy sys ems. I also p o ides sma mobili y solu ions o ail and oad anspo , medical echnology, and digi al heal hca e se ices. SKF AB Na u al Capi al Is a supplie o p oduc s, solu ions and se ices o bea ings, seals, mecha onics, and lub ica ion sys ems ha o e s se ices such as echnical suppo , condi ion moni o ing, main enance se ices, asse e iciency op imiza ion, enginee ing consul ancy, including ma e ial analysis and es ing, and aining. I p o ides clean echnology solu ions o he indus ial ans o ma ion, and has been de eloping p oduc s o become ligh e , mo e e icien , longe las ing, epai able, and, ul ima ely, ecyclable. Smi h & Nephew PLC Basic Needs Engages in he de elopmen , manu ac u e, ma ke ing, and sale o medical de ices and se ices ocused on he epai , egene a ion, and eplacemen o so and ha d issue, and o use in o hopaedic econs uc ion and auma, spo s medicine, ENT, and ad anced wound managemen . Some o i s p oduc s include knee and hip implan s o eplace diseased, damaged, o wo n join s; obo ics-assis ed echnologies; and auma- ela ed p oduc s o s abilize ac u es and de o mi y co ec ion p ocedu es. Smi hs G oup PLC Na u al Capi al I p o ides c i ical solu ions o he ene gy and p ocess, medical de ices and consumables, de ices o he de ec ion and iden i ica ion o secu i y h ea s and con aband, solu ions o high-speed connec i i y, and componen s o hea and mo e luids and gases. SSAB AB Na u al Capi al Is engaged in he de elopmen , manu ac u e and ma ke ing o high s eng h ossil- ee s eel p oduc s and solu ions in o de o c ea e an o e ing comp ised o solu ions anging om ligh ness and du abili y o s eng h, e iciency, sus ainabili y, and sa e y, o cus ome s in he hea y anspo , cons uc ion and in as uc u e, indus ial applica ions, cons uc ion machine y, ene gy, and ma e ial handling ma ke s. S o a Enso Oyj Na u al Capi al Is a p o ide o enewable solu ions in packaging, bioma e ials, wooden cons uc ion, and pape indus ies ha de elops and p oduces solu ions based on wood and biomass o a ange o indus ies and applica ions such as building, e ail, ood and be e ages, manu ac u ing, publishing, pha maceu ical, and o he s. Swisscom AG Empowe men Is a p o ide o elecommunica ion p oduc s and se ices such as mobile, ixed and In e ne p o ocol (IP) based oice and da a communica ions, and is cha ac e ized by i s sus ainabili y p ac ices by becoming clima e neu al and p omo ing ci cula economy. I o e s digi al access se ices, mobile ne wo k, and b oadband se ices o e exis ing subsc ibe lines. Tele onica SA Empowe men Ope a es as a elecommunica ions company ha o e s ixed and mobile elephone, b oadband, and subsc ip ion ele ision se ices, as well as p o ides cybe secu i y, IOT, big da a, and cloud solu ions. P omo es 49 deca bonisa ion and ci cula i y, economic and social p og ess h ough digi aliza ion. Tom a Sys ems ASA Na u al Capi al P o ides echnology-led solu ions ha enable he ci cula economy wi h ad anced collec ion and so ing sys ems ha op imize esou ce eco e y and minimize was e in he ood, ecycling, and mining indus ies. Ucb SA Basic Needs Is a biopha maceu ical company ha is engaged in he disco e y and de elopmen o no el medicines and solu ions o he ea men o a ious se e e diseases s i ing o de elop p oduc s o he ea men o neu ology and immunology ela ed condi ions. Umico e SA Na u al Capi al De elops echnologies and p oduces ma e ials o sola cells used in sa elli es, echa geable ba e ies, LED applica ions, ca alys s, e c. I ecycles and ex ac s p ecious me als om lap ops and mobile phones, among o he hings. I also educes ha m ul emissions, powe he ehicles and echnologies o he u u e, and gi es new li e o used me als. Veolia En i onnemen SA Clima e Change I designs and p o ides wa e , was e, and ene gy managemen solu ions: Wa e solu ions co e po able wa e plan s and supply sys ems, was ewa e sys ems and ea men plan s, among o he s; Was e solu ions include was e collec ion se ices on behal o local au ho i ies, managemen o specialis so ing cen es and pla o ms, among o he s; Ene gy solu ions o e powe gene a ion and ela ed se ices, such as consul ing and plan design. Ve bund AG Clima e Change P o ides in eg a ed elec ic gene a ions, ansmission, and dis ibu ion se ices. T ades powe and ene gy de i a i es o e he coun e and a he majo powe exchanges ac oss Eu ope. I gene a es elec ici y ia i s la ge ne wo k o hyd oelec ic powe plan s, which accoun s o app oxima ely 90% o he powe gene a ed, supplemen ed by wind, sola and he mal powe plan s, powe ed by ossil uels. Ves as Wind Sys ems A/S Clima e Change Is a enewable ene gy company, ac i e wi hin he wind powe indus y, ha designs, manu ac u es, ins alls, and p o ides se ices o onsho e and o sho e powe con e e s, wind u bines, blades, and owe s. I o e s se ices including da a-d i en consul ancy, lee op imiza ion, blade main enance and inspec ion, powe gene a o epai s, and gea box exchange. Vossloh AG Na u al Capi al Is a p o ide o ail in as uc u e se ices and anspo echnology o ail in as uc u e and echnology ma ke s ha i o e s a unique p oduc ange o ail in as uc u e, such as ail as ening sys ems, conc e e sleepe s, swi ch sys ems and c ossings as well as se ices ela ed o he alue main enance o he ails, he eby being an impo an con ibu o o he mobili y o people and he anspo o commodi ies. 56 Table E2 - In o ma ion C i e ia, Log-likelihood, MSE and RMSE o mul i a ia e GARCH models wi h m -dis ibu ion o IIFX-V2TX se ies Model In e al AIC BIC Shiba a HQC Log-Lik MSE RMSE DCC-GARCH Full Pe iod -9,729 -9,675 -9,729 -9,709 9027,033 0,003 0,055 P e-COVID-19 -10,043 -9,944 -10,044 -10,005 4376,766 0,003 0,054 Du ing COVID-19 -9,199 -9,091 -9,200 -9,158 3559,759 0,004 0,059 Pandemic-Wa Pe iod -9,462 -9,372 -9,462 -9,428 4673,116 0,003 0,055 ADCC-GARCH Full Pe iod -9,728 -9,671 -9,728 -9,707 9027,033 0,003 0,055 P e-COVID-19 -10,041 -9,937 -10,042 -10,001 4376,766 0,003 0,054 Du ing COVID-19 -9,197 -9,082 -9,198 -9,153 3559,824 0,004 0,059 Pandemic-Wa Pe iod -9,460 -9,365 -9,461 -9,424 4673,187 0,003 0,055 Copula-DCC-GARCH Full Pe iod -9,750 -9,699 -9,750 -9,731 9045,444 0,003 0,055 P e-COVID-19 -10,057 -9,964 -10,058 -10,021 4381,797 0,003 0,054 Du ing COVID-19 -9,222 -9,119 -9,223 -9,183 3567,452 0,004 0,059 Pandemic-Wa Pe iod -9,482 -9,398 -9,483 -9,450 4682,141 0,003 0,055 57 Table E3 - In o ma ion C i e ia, Log-likelihood, MSE and RMSE o mul i a ia e GARCH models wi h m -dis ibu ion o IIFX-V1IX se ies Model In e al AIC BIC Shiba a HQC Log-Lik MSE RMSE DCC-GARCH Full Pe iod -9,927 -9,873 -9,927 -9,907 9210,045 0,002 0,050 P e-COVID-19 -10,248 -10,149 -10,248 -10,210 4465,457 0,002 0,049 Du ing COVID-19 -9,408 -9,299 -9,409 -9,366 3639,993 0,003 0,054 Pandemic-Wa Pe iod -9,655 -9,566 -9,656 -9,621 4768,384 0,003 0,051 ADCC-GARCH Full Pe iod -9,926 -9,869 -9,926 -9,905 9210,045 0,002 0,050 P e-COVID-19 -10,245 -10,141 -10,246 -10,205 4465,457 0,002 0,049 Du ing COVID-19 -9,405 -9,291 -9,407 -9,361 3640,049 0,003 0,054 Pandemic-Wa Pe iod -9,654 -9,559 -9,655 -9,618 4768,636 0,003 0,051 Copula-DCC-GARCH Full Pe iod -9,947 -9,896 -9,947 -9,928 9227,438 0,002 0,050 P e-COVID-19 -10,256 -10,162 -10,257 -10,220 4468,012 0,002 0,049 Du ing COVID-19 -9,438 -9,336 -9,439 -9,399 3650,761 0,003 0,054 Pandemic-Wa Pe iod -9,687 -9,603 -9,688 -9,655 4783,094 0,003 0,051 58 This is o ce i y ha P ojec No.: FIN2024-6-149921 P ojec Ti le: Unde s anding he ela ionship be ween impac in es men i ms and ma ke ola ili y P incipal Resea che : Fábio Sousa acco ding o he egula ions o he E hics Commi ee o NOVA IMS and MagIC Resea ch Cen e his p ojec was conside ed o mee he equi emen s o he NOVA IMS In e nal Re iew Boa d, being conside ed APPROVED on 6/14/2024. I is he P incipal Resea che ’s esponsibili y o ensu e ha all esea che s and s akeholde s associa ed wi h his p ojec a e awa e o he condi ions o app o al and which documen s ha e been app o ed. The P incipal Resea che is equi ed o no i y he E hics Commi ee, ia amendmen o p og ess epo , o - Any signi ican change o he p ojec and he eason o ha change; - Any un o eseen e en s o unexpec ed de elopmen s ha me i no i ica ion; - The inabili y o he P incipal Resea che o con inue in ha ole o any o he change in esea ch pe sonnel in ol ed in he p ojec . Lisbon, 6/14/2024 NOVA IMS E hics Commi ee e hi[email p o ec ed] 59