The paradox of progress: Technological advancements in banking and the dual impact on SME bank borrowing

H yckiewicz, Ane a; Ko os ele a, Julia; Kozłowski, Łukasz; Rzepka, Malwina; Wang,
Ruomeng
Wo king Pape
The pa adox o p og ess: Technological ad ancemen s in
banking and he dual impac on SME bank bo owing
ADBI Wo king Pape , No. 1468
P o ided in Coope a ion wi h:
Asian De elopmen Bank Ins i u e (ADBI), Tokyo
Sugges ed Ci a ion: H yckiewicz, Ane a; Ko os ele a, Julia; Kozłowski, Łukasz; Rzepka, Malwina; Wang,
Ruomeng (2024) : The pa adox o p og ess: Technological ad ancemen s in banking and he dual
impac on SME bank bo owing, ADBI Wo king Pape , No. 1468, Asian De elopmen Bank Ins i u e
(ADBI), Tokyo,
h ps://doi.o g/10.56506/DOZI7138
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/305426
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by-nc-nd/3.0/igo/
ADBI Wo king Pape Se ies
THE PARADOX OF PROGRESS:
TECHNOLOGICAL ADVANCEMENTS
IN BANKING AND THE DUAL IMPACT
ON SME BANK BORROWING
Ane a H yckiewicz, Julia Ko os ele a,
Lukasz Kozlowski, Malwina Rzepka,
and Ruomeng Wang
No. 1468
July 2024
Asian De elopmen Bank Ins i u e
The Wo king Pape se ies is a con inua ion o he o me ly named Discussion Pape se ies; he
numbe ing o he pape s con inued wi hou in e up ion o change. ADBI’s wo king pape s e lec
ini ial ideas on a opic and a e pos ed online o discussion. Some wo king pape s may de elop in o
o he o ms o publica ion.
The a icle has been sponso ed by he Na ional Science Cen e (NCN) in Poland unde
2021/41/B/HS4/03586.
The Asian De elopmen Bank e e s o “China” as he People’s Republic o China.
Sugges ed ci a ion:
H yckiewicz, A., J. Ko os ele a, L. Kozlowski, M. Rzepka, and R. Wang. 2024. The Pa adox o
P og ess: Technological Ad ancemen s in Banking and he Dual Impac on SME Bank Bo owing.
ADBI Wo king Pape 1468. Tokyo: Asian De elopmen Bank Ins i u e. A ailable:
h ps://doi.o g/10.56506/DOZI7138
Please con ac he au ho s o in o ma ion abou his pape .
Email: ane a.h yckiewi[email p o ec ed], [email p o ec ed], j.ko os ele [email p o ec ed]c.uk
Ane a H yckiewicz is a Visi ing Schola and Associa e, Said Business School, Uni e si y o
Ox o d, and an Associa e P o esso a Kozminski Uni e si y, Wa saw, Poland. Julia Ko os ele a is
a p o esso a Uni e si y College London, Uni ed Kingdom (UK). Lukasz Kozlowski is an associa e
p o esso a Kozminski Uni e si y, Depa men o Banking, Insu ance, and Risk, Wa saw, Poland.
Malwina Rzepka is a PhD candida e, Economic Ins i u e o Empi ical Analysis, Wa saw, Poland.
Ruomeng Wang is a PhD candida e, Uni e si y College London, London, UK.
The iews exp essed in his pape a e he iews o he au ho and do no necessa ily e lec he
iews o policies o ADBI, ADB, i s Boa d o Di ec o s, o he go e nmen s hey ep esen . ADBI
does no gua an ee he accu acy o he da a included in his pape and accep s no esponsibili y
o any consequences o hei use. Te minology used may no necessa ily be consis en wi h ADB
o icial e ms.
Discussion pape s a e subjec o o mal e ision and co ec ion be o e hey a e inalized and
conside ed published.
The au ho s a e g a e ul o all commen s ecei ed om he pa icipan s o he esea ch semina
a he Saïd Business School, Uni e si y o Ox o d (2022); he IFABS con e ence in Naples (2021);
and he Wo kshop on SME Pe o mance unde Unce ain y (2024) in Tokyo. The au ho s do no
epo any con lic o in e es . All opinions a e hei own and should no be associa ed wi h any
ela ed ins i u ion.
Asian De elopmen Bank Ins i u e
Kasumigaseki Building, 8 h Floo
3-2-5 Kasumigaseki, Chiyoda-ku
Tokyo 100-6008, Japan
Tel: +81-3-3593-5500
Fax: +81-3-3593-5571
URL: www.adbi.o g
E-mail: [email p o ec ed]
© 2024 Asian De elopmen Bank Ins i u e
ADBI Wo king Pape 1468 A. H yckiewicz e al.
Abs ac
This s udy del es in o he impac o echnological bank inno a ions on small and
medium-sized en e p ise (SME) bo owing ac oss he Eu opean Union. By analyzing a
comp ehensi e da ase o 179,921 SME-bank lending ela ionships om 2009 o 2019, we
explo e he mechanisms h ough which echnological ad ancemen s in banking eshape
adi ional lending p ac ices. Ou empi ical analysis documen s ha banks’ echnological
inno a ions ha e a mo e subs an ial impac on SMEs’ long- e m c edi g ow h han on hei
sho - e m g ow h, indica ing he use ulness o hese echnologies in p o iding da a ha
no only educe in o ma ion asymme y bu also enhance long- e m decision channels.
Speci ically, blockchain and au oma ion play a c ucial ole in expanding bank c edi o SMEs.
Howe e , we also iden i y a pa adoxical dual e ec : while echnological ad ancemen s
acili a e c edi access, hey simul aneously inc ease he cos o bo owing o SMEs. This
inding highligh s a complex in e play whe e echnological p og ess in banking p esen s bo h
oppo uni ies and challenges, especially o mo e opaque i ms seeking inancing.
Ou s udy con ibu es o he unde s anding o he nuanced ole o inno a ion in banking,
o e ing insigh s in o he dualis ic na u e o he impac o echnology on SME inancing.
Keywo ds: SME inancing, bank echnology, inno a ion, colla e al, cos o in e media ion
JEL Classi ica ion: G21, O16, O33, G23, G32
ADBI Wo king Pape 1468 A. H yckiewicz e al.
Con en s
1. INTRODUCTION .......................................................................................................... 1
2. DATA AND METHODOLOGY ...................................................................................... 4
2.1 Sample Desc ip ion .......................................................................................... 4
2.2 Me hodology ..................................................................................................... 7
3. EMPIRICAL RESULTS .............................................................................................. 11
3.1 Banks’ Digi aliza ion and SME Bo owing ...................................................... 11
3.2 Banks’ Technology and SMEs’ Bo owing: A Di e ence-in-Di e ence
App oach ........................................................................................................ 13
3.3 The Role o Di e en Types o Bank Technology in SME Bo owing ............. 15
4. THE CHANNELS OF THE TECHNOLOGICAL EFFECTS ON SME BORROWING . 18
4.1 Reduc ion in In o ma ion F ic ions and Reliance on Colla e al ....................... 18
4.2 C edi Access and Cos o In e media ion ...................................................... 20
5. ROBUSTNESS CHECKS ........................................................................................... 21
5.1 Endogenei y Conce ns and Ins umen al Va iable Reg ession (IV) ............... 21
5.2 Iden i ica ion S a egy ..................................................................................... 25
6. CONCLUSIONS ......................................................................................................... 29
REFERENCES ...................................................................................................................... 30

ADBI Wo king Pape 1468 A. H yckiewicz e al.
1
1. INTRODUCTION
In he ace o apid echnological e olu ion, he banking indus y has been a he
o e on o emb acing change. F om he con enience o elec onic paymen s o he
insigh ul wo ld o big da a analy ics and he in elligence o AI solu ions and blockchain,
banks ha e ha nessed hese inno a i e echnologies o e olu ionize hei decision-
making p ocesses. These ecen ad ancemen s in echnology can in oduce new
lending pa adigms ha could ex end no el oppo uni ies o adi ionally unde se ed
cus ome s, like SMEs.
Small and medium-sized en e p ises (SMEs) play a i al ole in p omo ing sus ainable
economic g ow h wo ldwide ia os e ing inno a ion and compe i ion, and c ea ing
employmen . In Eu ope, SMEs accoun o 99% o he en e p ise popula ion and
o mo e han hal o i s GDP and employmen .1 Despi e hei well-acknowledged
impo ance o he economy, SMEs ecei e a disp opo iona ely small sha e o c edi
om inancial ins i u ions, and such a end pe sis s ac oss bo h de eloped and
de eloping coun ies (Beck and Demi güç-Kun 2006; ECB SAFE Su ey 2020; Wo ld
Bank 2018, 2023). These inancing cons ain s a e oo ed in he inhe en in o ma ion
opaci y o SMEs, which exace ba es he asymme y be ween lende s and bo owe s
and leads o c edi a ioning, as p o en by S igli z and Weiss (1981). Addi ionally,
he lack o aluable colla e al and unp opo ionally high cos s o bank inancing o
SMEs exclude he la e om bank inancing e en u he (Beck and Demi güç-Kun
2006; De Blick, Paeleman, and La e en 2023; Ha ison e al. 2022; Yaldiz Haneda ,
B occa do, and Bazzana 2014).
In his esea ch, we in es iga e whe he he mos ecen echnological inno a ions
adop ed by banks can esol e he exis ing challenges aced by SMEs when hey
seek bank inancing, and which echnological ad ancemen s can con ibu e o SMEs’
imp o emen in accessing bank c edi .
Using he Amadeus i m-le el panel da ase , ou sample includes 179,921 SMEs
ac oss he EU, om 2009 o 2019, pai ed wi h da a om 54 majo Eu opean banks.
We ocus on SMEs wi h exis ing bank ela ionships o disce n he addi i e alue o
echnology in p o iding da a o banks. We hen me ge bank da a wi h in o ma ion on
echnological solu ions implemen ed by each bank a ilia ed wi h an SME. Mo e
impo an ly, o each bank, we can iden i y a echnological inno a ion ha a bank has
implemen ed as well as he yea o i s implemen a ion. Thus, by agg ega ing all he
echnological inno a ions implemen ed by a bank be ween 2009 and 2019, we can
e alua e he le el o he bank’s echnological inno a ion and ack i s p og ession o e
ime. We e ie e echnological da a om he C unchbase and CBInsigh s da abases
and supplemen his in o ma ion by web-mining p ocesses allowing us o iden i y
banks’ announcemen s on bank echnological p oduc acquisi ion and/o de elopmen
as well as i s na u e.
Addi ionally, we also use al e na i e measu es o bank echnological de elopmen ,
such as: (i) he numbe o iled applica ions by a bank; (ii) he numbe o pa en s
g an ed by a bank; and (iii) he numbe o deals a bank has been in ol ed wi h as
a en u e capi alis (VC). We e ie e his in o ma ion a a bank-yea le el om
GlobalDa a. Howe e , a ce ain cau ion o hese measu es mus be g an ed as hey
a e a ailable only o 3,500 public companies, wi h he banking sec o being highly
unde ep esen ed. We use hese al e na i e measu es o bank echnological
inno a i eness o es he obus ness o ou esul s.
1 h ps://ec.eu opa.eu/g ow h/smes_en.
ADBI Wo king Pape 1468 A. H yckiewicz e al.
2
Ou me hodology in ol es wo-way ixed-e ec eg ession models, including in e ac ion
e ec s, di e ence-in-di e ence (DID) es ima ions, and wo-s age ins umen al a iable
(2SLS IV) echniques o add ess po en ial endogenei y be ween he adop ion o
bank echnology and SME bo owing. We show ha he Second Paymen Se ices
Di ec i e (PSD2) adop ed by he Eu opean Commission a he end o 2015 has
caused an exogenous shock, signi ican ly speeding up he digi iza ion o he en i e
inancial sec o in Eu ope a e wa ds. The occu ence o his shock has ex ended he
access o di e en echnologies o banks and p o ided a solid ounda ion o ou
DID es ima ions.
In con as o exis ing li e a u e ha examines FinTech and BigTech i ms’ ole in
ca e ing o unde se ed cus ome s, ou s udy o e s a unique pe spec i e on he ole o
echnological inno a ion a banks in accessing c edi o opaque bo owe s. While he e
a e some academic s udies documen ing how FinTech and BigTech companies ex end
inancial se ices o o e looked o unde se ed cus ome s ( o example, Balyuk (2022);
Beaumon e al. (2021); Be ge e al. (2021); Co nelli e al. (2023); Gambaco a e al.
(2019); Gopal and Schnabl (2022); Jag iani e al. (2021); Jag iani and Lemieux (2018,
2019);, Ouyang (2022); Palladino (2021)), li le is known abou how ecen banking
echnological inno a ions a e changing he lending amewo k owa d opaque
cus ome s. Though he ole o bank digi aliza ion in supplying supe io c edi has been
ecen ly e idenced, mainly du ing he pandemic c isis (see, o example, B anzoli e al.
(2021; Fe i e al. (2019); Kwan e al. (2021); o mo e gene ally D’And ea and Limodio
(2023)), hese s udies a e mainly silen conce ning he ole o echnology in supplying
c edi o unde se ed cus ome s. To he bes o ou knowledge, he only s udy ha
a emp s o look a his ques ion is ha o Seduno (2017), who uses US bank da a
o he pe iod be ween 2001 and 2008 – long be o e he eal FinTech de elopmen
a ea began. Simila ly, Sheng (2021) examines he ole o FinTech ins i u ions in
p o iding bank c edi o SMEs in he People’s Republic o China (PRC); howe e , he
au ho uses mac o da a a he han bank-le el da a. Wi h ou s udy, we close he
gaps in he exis ing li e a u e by u ilizing he mos ecen echnological solu ions
adop ed by indi idual banks o assess hei ole in imp o ing access o bank c edi o
unde se ed cus ome s.
Ou s udy e eals ha he ecen adop ion o echnological inno a ions by banks has
enabled SMEs o o e come ce ain ba ie s and ge be e access o bank c edi .
Howe e , we also ound ha a ce ain le el o echnological de elopmen a banks is
necessa y o e ec i ely add ess he challenges associa ed wi h he in o ma ion opaci y
o SMEs. Impo an ly, we obse e a mo e p onounced impac o bank echnology
on inc eased SMEs’ long- e m bo owing compa ed o sho - e m bo owing. This
sugges s ha echnological inno a ion no only mi iga es asymme ic in o ma ion
p oblems bu also enhances in o ma ion e iciency, a ec ing o he channels o bank
c edi decision-making. This unde sco es he mul i ace ed ole o di e en echnological
inno a ions in assis ing banks wi h c edi decisions o opaque companies.
Addi ionally, ou analysis en iches he li e a u e by examining he ole o speci ic ypes
o echnological inno a ions in banks’ c edi decision-making o SMEs. We explo e
he complexi y and inno a i eness o echnological de elopmen a banks, including
he adop ion o elec onic paymen s, online lending, pe sonal inance solu ions, da a
analy ics, egula o y echnology, blockchain, and au oma ion. Subsequen ly, we es
how hese echnologies add ess he unique challenges aced by opaque bo owe s.
The uniqueness and complexi y o ou da ase signi ican ly dis inguish us om p e ious
academic s udies ha examine he impac o he gene al le el o bank digi aliza ion,
mos ly measu ed by bank IT spending (B anzoli, Rainone, and Supino 2021; D’And ea
and Limodio 2023; Kwan e al., 2020; Ma inez Pe ia e al. 2022; Pie i and Timme
ADBI Wo king Pape 1468 A. H yckiewicz e al.
3
2022) o by access o he in e ne (D’And ea and Limodio 2023). The g adualness o
ou da a allows us o in es iga e he con ibu ion o each echnological solu ion o
add ess he p oblems o SMEs and hei access o bank c edi .
Ou esea ch unde sco es he pi o al ole o blockchain echnology in mi iga ing he
di icul ies encoun e ed by SMEs in ob aining bank c edi . In pa icula , ou eg ession
analysis e eals ha blockchain echnology signi ican ly b oadens he spec um o
c edi op ions a ailable o SMEs. Th ough i s abili y o collec and p ocess la ge se s o
da a, i allows banks o educe he asymme ic in o ma ion p oblem in an e icien way,
imp o ing access o bank c edi o opaque i ms. Mo eo e , we also ind ha da a
analy ics and au oma ion also appea o be highly s a is ically signi ican in imp o ing
SMEs’ access o c edi . Consequen ly, ou eg essions e eal ha access o da a and
e iciency in collec ing and p ocessing hem seem o be he mos impo an ac o s in
ex ending c edi o opaque cus ome s.
Mo eo e , ou s udy o e s no el and comp ehensi e insigh s in o a ious channels
h ough which echnological inno a ion in banks can lead o inc eased SME bo owing.
We ocus on how such echnology in luences he easing o colla e al equi emen s and
he cos o in e media ed bank c edi . While a ew s udies, such as hose by Buchak
e al. (2018), Jag iani and Lemieux (2018), and Beaumon e al. (2021), compa e he
cos o c edi o e ed by FinTech companies o ha o banks, he p ecise impac o
echnological ad ancemen on he cos o bank in e media ion emains unde explo ed.
These aspec s ha e been p ima ily discussed in FinTech and BigTech li e a u e, o en
wi hou de ini i e conclusions, bu a e less widely co e ed in banking li e a u e.
Simila ly, he cu en li e a u e has no add essed how bank echnology a ec s he ole
o colla e al in opaque companies accessing bank c edi . Al hough Holms om and
Ti ole (1997) and Gambaco a e al. (2023) a gue ha g ea e access o ha d da a
could ease bank equi emen s o colla e al, he exis ing li e a u e has no empi ically
e i ied he link be ween i m colla e al and bank c edi . Ou da a c ea e a g ea es ing
g ound o es his ela ionship.
Ou eg ession esul s documen ha banks wi h mo e inno a i e echnology equi e
less aluable colla e al om SMEs han less digi alized banks. In o he wo ds, SMEs
associa ed wi h echnologically ad anced banks can access c edi wi h lowe colla e al
equi emen s han hose associa ed wi h less digi alized banks. This inding seems o
sugges ha ansac ional da a may also imp o e he sc eening p ocesses a banks. A
he same ime, ou eg ession esul s challenge he p esump ion ha echnological
inno a ion leads o lowe inancing cos s. In u n, we ind ha mo e echnologically
ad anced banks end o cha ge SMEs highe in e es a es. Ou s udy may indica e
ha he echnology migh no pe ec ly eplace he ela ional so da a ha banks
accumula e o e ime, and he ease o he colla e al o ces banks o cha ge a highe
isk p emium.
In ou s udy, we igo ously add ess a ious po en ial biases and endogenei y conce ns
ac oss a ious speci ica ions, measu es o bank echnological de elopmen , and
he na u e o SME-bank ela ionships. Fi s ly, we employ OLS eg ession wi h an
in e ac ion e m as well as DID o es he po en ial issues ela ed o he assump ions
o DID. Secondly, we u ilize TWFE DID s agge ed wi h iming as well as a single
ea men pe iod o es he obus ness o ou DID esul s. Thi dly, o add ess po en ial
endogenei y ela ed o banks’ indi idual ea u es and hei echnological de elopmen ,
we employ wo-s age ins umen al a iable (2SLS IV) eg ession using coun y-le el
a ia ion ela ed o he adop ion o PSD2 as an exogenous ins umen o banks’
echnological de elopmen . Howe e , o enhance he echnological ad ancemen o
banks, we conduc addi ional eg essions o add ess he iden i ica ion p oblem. These
eg essions in ol e examining he ela ionship be ween indi idual SMEs and banks,
ADBI Wo king Pape 1468 A. H yckiewicz e al.
4
such as i m digi al in ensi y o indus y ype, which could po en ially in oduce bias in o
ou es ima ed esul s. All ou obus ness analyses highligh he c ucial ole o bank
echnological inno a ion in shaping SMEs’ access o c edi .
Ou pape is s uc u ed as ollows. The nex sec ion discusses he da a and
me hodology, while Sec ions 3 and 4 p esen he esul s, which a e u he es ed
o hei obus ness in Sec ion 5. Sec ion 6 o e s conclusions and policymaking
implica ions.
2. DATA AND METHODOLOGY
2.1 Sample Desc ip ion
To in es iga e ou esea ch ques ions, we assemble a a ie y o da a, such as SME
da a, bank-le el da a, including in o ma ion on each bank-implemen ed echnological
solu ion, and mac oeconomic coun y-le el da a.
Ou da a collec ion p ocess s a s wi h he cons uc ion o he SME panel sample.
Fo his, we use he Amadeus da abase p o ided by he Bu eau an Dijk, which is he
majo sou ce o EU-compa able inancial and accoun ing da a on i ms. The p ocess
is comp ised o wo s ages: (i) cons uc ion o he key inancial indica o s and i m-le el
con ols o hose i ms wi h unconsolida ed accoun s; and (ii) ga he ing o he
i ms’ bank a ilia ion in o ma ion. To iden i y SMEs, we ollow he Eu opean
Commission’s de ini ion o SMEs – which is also used by Eu os a – as ha ing ewe
han 250 pe sons employed and an annual u no e o up o EUR50 million o a o al
balance shee o no mo e han EUR43 million. The Amadeus da abase is also a
p ima y sou ce o in o ma ion allowing us o link SMEs wi h hei a ilia ed banks.
We hus es ic ou da abase o only hose i ms o which he in o ma ion on hei
bank a ilia ion was a ailable in he da abase. In o al, we can iden i y 179,921 i ms
om 15 coun ies, i.e., Aus ia, C oa ia, Denma k, Es onia, F ance, Ge many, G eece,
Hunga y, I eland, La ia, Poland, Po ugal, Slo enia, Spain, and he UK. Table 1 gi es
an o e iew o he sample s uc u e by he yea and by he numbe o banks a ilia ed
wi h a i m. I also p o ides he ea u es o SMEs used in ou analysis.
Table 1: Sample S uc u e
This able p esen s a sample s uc u e based on he obse a ions employed in eg essions om
Speci ica ion 1 in Table 4.
Panel A. Sample s uc u e by yea
Yea
Coun ies
Obse a ions
% o Obse a ions
2009
10
7,395
0.7
2010
12
30,061
3.0
2011
12
74,473
7.4
2012
12
68,994
6.9
2013
13
103,770
10.3
2014
13
116,038
11.6
2015
13
119,877
11.9
2016
14
129,938
12.9
2017
14
133,456
13.3
2018
14
134,228
13.4
2019
15
85,178
8.5
All yea s
15
1,003,408
100.0
con inued on nex page
ADBI Wo king Pape 1468 A. H yckiewicz e al.
11
Table 3: Desc ip i e S a is ics
This able p esen s desc ip i e s a is ics o he sample.
Va iable
Obse a ions
Fi ms
Mean
S d. De .
Min.
1s Qua .
2nd Qua .
3 d Qua .
Max.
A. Dependen a iables
DEBT.GR
1,003,408
179,921
–0.004
0.116
–0.547
–0.042
–0.016
0.006
0.856
INT.COST
629,578
128,922
0.086
0.274
–0.057
0.007
0.029
0.064
4.003
LT.DEBT.GR
1,001,487
179,830
–0.008
0.098
–0.547
–0.036
–0.016
0.003
0.763
ST.DEBT.GR
1,002,321
179,867
–0.008
0.069
–0.536
–0.024
–0.012
0.003
0.756
B. O he i m-le el a iables
PROFIT
1,003,408
179,921
0.026
0.159
–2.000
0.007
0.028
0.070
0.600
FIXED.ASSETS
1,003,408
179,921
0.299
0.259
0.000
0.075
0.231
0.473
1.000
LOW.COLLAT
977,667
176,325
0.499
0.500
0.000
0.000
0.000
1.000
1.000
EQUITY
1,003,408
179,921
0.466
0.264
0.000
0.246
0.450
0.679
1.000
ASSET.TURN
1,003,408
179,921
1.679
1.514
0.000
0.750
1.303
2.102
14.999
FIRM.SIZE
1,003,408
179,921
–0.268
1.688
–10.125
–1.392
–0.229
0.835
3.912
LN.FIRM.AGE
1,003,408
179,921
2.724
0.818
0.000
2.398
2.890
3.219
5.541
DIGITAL.FIRM
1,731,608
173,161
0.030
0.171
0.000
0.000
0.000
0.000
1.000
HIGH.CAPITAL
1,731,608
173,161
0.014
0.117
0.000
0.000
0.000
0.000
1.000
C. Coun y–le el a iables
PRI.CREDIT
1,003,408
179,921
1.081
0.366
0.324
0.937
1.112
1.306
1.921
GDP.GROWTH
1,003,408
179,921
0.017
0.022
–0.143
0.007
0.020
0.029
0.084
GDP.PC
1,003,408
179,921
35.385
5.202
21.024
31.305
35.969
38.906
86.550
UNEMPL
1,003,408
179,921
0.155
0.067
0.031
0.097
0.153
0.214
0.275
D. Bank undamen als
BANK.SIZE
1,003,408
179,921
12.073
1.502
8.252
10.920
12.301
13.288
14.625
BANK.LOANS
1,003,408
179,921
0.597
0.093
0.131
0.561
0.595
0.655
0.863
BANK.EQUITY
1,003,408
179,921
0.080
0.031
0.011
0.063
0.070
0.081
0.224
BANK.DEPO.GR
1,003,408
179,921
0.055
0.105
–0.424
–0.003
0.034
0.085
1.311
E. Financial inno a ions a a bank
AUT.SOFT
1,003,408
179,921
0.223
0.375
0.000
0.000
0.000
0.500
1.000
BLOCKCHAIN
1,003,408
179,921
0.153
0.338
0.000
0.000
0.000
0.000
1.000
ANALYTICS
1,003,408
179,921
0.134
0.301
0.000
0.000
0.000
0.000
1.000
LENDING
1,003,408
179,921
0.191
0.348
0.000
0.000
0.000
0.333
1.000
PAYMENTS
1,003,408
179,921
0.266
0.422
0.000
0.000
0.000
0.500
1.000
PERSON.FIN
1,003,408
179,921
0.089
0.260
0.000
0.000
0.000
0.000
1.000
REGULAT
1,003,408
179,921
0.238
0.387
0.000
0.000
0.000
0.500
1.000
BANK.INNOV
1,003,408
179,921
1.294
1.770
0.000
0.000
0.000
2.000
7.000
HIGH.DIGITAL
1,731,608
173,161
0.138
0.345
0.000
0.000
0.000
0.000
1.000
MAX.DIGITAL
1,643,512
164,351
2.750
2.096
0.000
1.000
3.000
4.000
7.000
3. EMPIRICAL RESULTS
3.1 Banks’ Digi aliza ion and SME Bo owing
We s a ou analysis by in es iga ing he gene al impac o bank echnological
ad ancemen on SME bo owing o e ime. Table 4 p esen s he eg ession esul s
on whe he , and i so how, bank echnological inno a ion (BANK.INNOV) is co ela ed
wi h di e en ypes o c edi g ow h a SMEs. Mo e speci ically, we examine he ole
o echnology in o e all c edi g ow h, as well as long- e m and sho - e m g ow h,
p esen ed in Speci ica ions (1)–(3), espec i ely.
As discussed in he p e ious sec ion, he PSD2 has os e ed inno a ion and
echnological ad ancemen s h ough he en y o new i ms o e ing inno a i e inancial
p oduc s and se ices, he eby inc easing he echnological de elopmen o many o he
inancial ins i u ions, including banks. To see whe he we could also no ice a change in
SME bo owing be o e and a e a po en ial shi in bank echnological inno a ion,

ADBI Wo king Pape 1468 A. H yckiewicz e al.
12
we in e ac he BANK.INNOV a iable wi h he yea dummies equal o one o he yea s
a e he adop ion o PSD2 (2016–2019) and ze o o he wise (YEAR2016_DUMMY).
The esul s a e p esen ed in Speci ica ions (4)–(6). All esul s om his sec ion a e
p esen ed in Table 4.
Table 4: The Role o Bank Technological Inno a ions in SMEs’ C edi G ow h
The able p esen s he eg ession esul s o i m- and yea - ixed-e ec panel models o bank
echnological inno a ion on SMEs’ c edi g ow h. DEBT.GR ep esen s he g ow h o SMEs’
combined sho - e m and long- e m bank deb a ime , di ided by he p e ious yea ’s o al
asse s (in la ion-adjus ed). LT.DEBT.GR indica es he g ow h o SMEs’ long- e m c edi a
ime , while ST.DEBT.GR deno es he g ow h o he SMEs’ sho - e m c edi . BANK.INNOV is a
measu e o a bank’s echnological inno a ion, de ined as he sum o all echnological solu ions
adop ed by bank i a ime . YEAR2016_DUMMY akes one o he pe iod a e PSD2 adop ion
(yea s 2016–2019). In he in e es o b e i y, we do no p esen coe icien s o he yea dummy
a iables. S anda d e o s, clus e ed a he i m le el, a e shown in pa en heses. *, **, and
*** e e o signi icance a he 10%, 5%, and 1% le els, espec i ely.
(1)
(2)
(3)
(4)
(5)
(6)
Va iables
DEBT.GR
LT.DEBT.GR
ST.DEBT.GR
DEBT.GR
LT.DEBT.GR
ST.DEBT.GR
L. BANK.INNOV
0.00109***
0.000984***
0.000240**
0.00119***
0.000921***
0.000824***
(0.000166)
(0.000137)
(0.000101)
(0.000281)
(0.000232)
(0.000163)
YEAR2016_DUMMY*
BANK.INNOV
0.000530**
0.000462**
–1.91e-06
(0.000221)
(0.000185)
(0.000122)
L.FIX_A
–0.0573***
–0.0634***
0.00627***
–0.0572***
–0.0633***
0.00631***
(0.00171)
(0.00145)
(0.000856)
(0.00171)
(0.00145)
(0.000856)
L.EBIT_S
–0.00573***
–0.00643***
0.000909
–0.00579***
–0.00648***
0.000870
(0.00120)
(0.00104)
(0.000639)
(0.00120)
(0.00104)
(0.000639)
L.EQUITY
0.135***
0.0909***
0.0393***
0.135***
0.0910***
0.0395***
(0.00144)
(0.00119)
(0.000747)
(0.00144)
(0.00119)
(0.000747)
L.TAT
0.0140***
0.00819***
0.00508***
0.0140***
0.00818***
0.00507***
(0.000283)
(0.000213)
(0.000151)
(0.000283)
(0.000213)
(0.000151)
L.LN_SALES
–0.0100***
–0.00628***
–0.00282***
–0.0100***
–0.00627***
–0.00281***
(0.000461)
(0.000375)
(0.000252)
(0.000461)
(0.000375)
(0.000252)
L.LN_FIRM_AGE
–0.0135***
–0.00906***
–0.00456***
–0.0132***
–0.00884***
–0.00437***
(0.000958)
(0.000757)
(0.000516)
(0.000960)
(0.000759)
(0.000517)
PRICREDIT
0.0165***
0.0234***
0.00394***
0.0187***
0.0252***
0.00522***
(0.00218)
(0.00179)
(0.00119)
(0.00219)
(0.00180)
(0.00121)
GDPGROWTH
0.331***
0.286***
0.215***
0.358***
0.307***
0.232***
(0.0150)
(0.0120)
(0.00939)
(0.0155)
(0.0125)
(0.00978)
GDPPCPPP
–0.000544*
–0.000439*
–0.00167***
–0.000470
–0.000414*
–0.00158***
(0.000324)
(0.000251)
(0.000189)
(0.000322)
(0.000250)
(0.000188)
UNEMPL
–0.223***
–0.194***
–0.177***
–0.226***
–0.198***
–0.177***
(0.0172)
(0.0137)
(0.00944)
(0.0172)
(0.0136)
(0.00949)
L.BANK_LN_ASSETS
–0.00122
–0.00176*
–0.00554***
–0.00235*
–0.00258***
–0.00635***
(0.00120)
(0.000990)
(0.000698)
(0.00121)
(0.00100)
(0.000708)
L.BANK_LOANS
–0.00919**
–0.00696*
–0.0140***
–0.0147***
–0.0109***
–0.0175***
(0.00436)
(0.00377)
(0.00246)
(0.00445)
(0.00386)
(0.00254)
L.BANK_EQUITY
–0.0299**
–0.000445
–0.0289***
–0.0269**
0.00141
–0.0273***
(0.0129)
(0.0107)
(0.00790)
(0.0130)
(0.0107)
(0.00796)
L.BANK_DEPO_GR
0.00253*
0.00277**
0.00285***
0.00286**
0.00283**
0.00344***
(0.00137)
(0.00112)
(0.000880)
(0.00135)
(0.00110)
(0.000866)
0.0335*
0.0384***
0.142***
0.0452**
0.0481***
0.149***
Cons an
(0.0175)
(0.0148)
(0.00953)
(0.0176)
(0.0149)
(0.00961)
Obse a ions
1,003,408
1,034,658
1,011,309
1,003,408
1,034,658
1,011,309
R-squa ed
0.043
0.036
0.041
0.043
0.036
0.041
Numbe o FIRM_ID
179,921
183,557
180,751
179,921
183,557
180,751
Time FE
Yes
Yes
Yes
Yes
Yes
Yes
Fi m FE
Yes
Yes
Yes
Yes
Yes
Yes
ADBI Wo king Pape 1468 A. H yckiewicz e al.
13
Ou indings o e in e es ing insigh s in o SME bo owing and bank echnological
ad ancemen . P ima ily, we see ha bank echnological ad ancemen is posi i ely
co ela ed wi h all o ms o SME c edi g ow h. Mo e speci ically, we ind ha SMEs
a ilia ed wi h mo e echnologically ad anced banks expe ience highe c edi g ow h.
This gi es us a i s insigh in o he ole o bank echnological inno a ion in ex ending
access o SMEs o bank c edi . In e es ingly, we ind ha he impac o bank
echnological inno a ion appea s o be mo e p onounced on long- e m han on sho -
e m c edi . The SMEs a ilia ed wi h a bank ha has one addi ional solu ion expe ience
a 0.09% highe c edi g ow h in long- e m c edi , whe eas he e ec on sho - e m
c edi g ow h is only 0.024%. This esul is economically alid as he mean o he
c edi g ow h a SMEs o he whole sample is nega i e. The esul means ha SMEs
could o se he nega i e ma ke end associa ed wi h mo e echnologically ad anced
banks. These indings a e pa icula ly p omising as hey sugges ha echnological
solu ions no only educe he asymme ic in o ma ion p oblems o acili a e SME
sho - e m c edi bu p obably also imp o e o he channels a ec ing banks’ long- e m
lending decisions.
The eg ession esul s on in e ac ion p esen ed in Speci ica ions (4)–(6) o e addi ional
insigh in o ou analysis. They documen ha he e ec o bank echnological
inno a ion on SME bo owing did no occu homogeneously o e ime. The eg ession
esul s seem o sugges ha when bank echnological inno a ion has sped up, we
can see highe c edi g ow h a SMEs. This conclusion is suppo ed by he posi i e
and s a is ically signi ican coe icien s o in e ac ion a iables be ween BANK.INNOV
and YEAR2016_DUMMY. A he same ime, we no ice ha BANK.INNOV a iables
emain highly s a is ically signi ican wi h a posi i e coe icien ac oss all speci ica ions.
This inding is in line wi h o he academic indings on he ole o paymen da a in
banks’ lending decisions (Ghosh, Vallee, and Zeng 2021; Ouyang 2022). Ou
esul s documen ha he pe iods o hese echnological de elopmen s coincided
wi h highe SME c edi g ow h, po en ially sugges ing ha hese solu ions could suppo
banks in hei lending decisions by inc easing he e iciency in collec ing and
p ocessing in o ma ion.
3.2 Banks’ Technology and SMEs’ Bo owing:
A Di e ence-in-Di e ence App oach
In his sec ion we compa e SME bo owing be o e and a e he in oduc ion o he
PSD2. Subsequen ly, we also g oup banks a ilia ed wi h SMEs based on he le el o
hei echnological inno a ion depending on he numbe o adop ed solu ions o be able
o compa e he SME bo owing a ilia ed wi h highly digi ized and less digi alized banks
as desc ibed in he Me hodology sec ion. Table 5 p esen s he eg ession esul s o
DID es ima ions.
Ou esul s p esen in e es ing insigh s. Fi s , we no e ha he esul s om he DID
ende he same conclusions as om he linea eg ession. This could sugges ha
endogenei y ela ed o he bank-SME ela ionship migh no be p esen . Mo e
speci ically, we see ha SMEs associa ed wi h mo e echnologically ad anced banks
ha e expe ienced much highe c edi g ow h, bo h sho - e m and long- e m, han
hose associa ed wi h less inno a i e banks. This inding has signi ican implica ions,
sugges ing ha banks ha ha e ollowed he echnological e olu ion since he
PSD2 ha e p obably imp o ed hei access o di e en da a and hei p ocessing. This
inding is consis en wi h Angelini, Tollo, and Roli (2008), Baza bash (2019), Fus e
e al. (2019), Jag iani and Lemieux (2019), Khandani e al. (2010)documen ing he
ADBI Wo king Pape 1468 A. H yckiewicz e al.
14
impo ance o big da a, da a sha ing, AI, and o he au oma ed p ocedu es in imp o ing
banks’ c edi sco ing p ocesses.
Table 5: The Role o Bank Technological Inno a ions in SMEs’ C edi G ow h
The able p esen s he DID es ima ions o he i m- and yea - ixed-e ec panel models
examining he impac o bank echnological inno a ion on SMEs’ c edi g ow h. The ea men
pe iod commenced in 2016 and con inued onwa ds. T ea ed banks a e hose ha adop ed mo e
han ou echnological solu ions in a gi en yea (HIGH.DIGITAL) o banks ha ha e adop ed
any echnological solu ion since 2016, while he maximum numbe is aken be ween 2016 and
2019 as a measu e o bank echnological de elopmen (MAX.DIGITAL). DEBT.GR ep esen s
he g ow h o an SME’s combined sho - e m and long- e m bank deb a ime , di ided by he
p e ious yea ’s o al asse s (in la ion-adjus ed). LT.DEBT.GR indica es he g ow h o an SME’s
long- e m c edi a ime , while ST.DEBT.GR deno es he g ow h o he SME’s sho - e m c edi .
Fo he sake o b e i y, we do no p esen coe icien s o he cons an e m and yea dummy
a iables. S anda d e o s, clus e ed a he i m le el, a e shown in pa en heses. *, **, and
*** indica e signi icance a he 10%, 5%, and 1% le els, espec i ely.
(1)
(2)
(3)
(4)
(5)
(6)
Va iables
DEBT.GR
LT.DEBT.GR
ST.DEBT.GR
DEBT.GR
LT.DEBT.GR
ST.DEBT.GR
YEAR2016_DUMMY*
HIGH.DIGITAL
0.00341***
0.00273***
0.00118***
(0.000578)
(0.000455)
(0.000364)
YEAR2016_DUMMY*
MAX.DIGITAL
0.00108***
0.000918***
0.000468***
(0.000137)
(0.000115)
(8.06e-05)
L.FIX_A
–0.0573***
–0.0634***
0.00626***
–0.0572***
–0.0633***
0.00629***
(0.00171)
(0.00145)
(0.000856)
(0.00171)
(0.00145)
(0.000856)
L.EBIT_S
–0.00570***
–0.00641***
0.000914
–0.00573***
–0.00644***
0.000900
(0.00120)
(0.00104)
(0.000639)
(0.00120)
(0.00104)
(0.000639)
L.EQUITY
0.135***
0.0908***
0.0393***
0.135***
0.0909***
0.0394***
(0.00144)
(0.00118)
(0.000747)
(0.00144)
(0.00119)
(0.000747)
L.TAT
0.0140***
0.00819***
0.00508***
0.0140***
0.00818***
0.00508***
(0.000283)
(0.000213)
(0.000151)
(0.000283)
(0.000213)
(0.000151)
L.LN_SALES
–0.0100***
–0.00629***
–0.00282***
–0.0101***
–0.00631***
–0.00283***
(0.000461)
(0.000375)
(0.000252)
(0.000461)
(0.000375)
(0.000252)
L.LN_FIRM_AGE
–0.0135***
–0.00909***
–0.00455***
–0.0133***
–0.00893***
–0.00444***
(0.000957)
(0.000757)
(0.000516)
(0.000959)
(0.000758)
(0.000517)
PRICREDIT
0.0153***
0.0224***
0.00358***
0.0180***
0.0247***
0.00470***
(0.00219)
(0.00180)
(0.00119)
(0.00220)
(0.00180)
(0.00121)
GDPGROWTH
0.325***
0.280***
0.215***
0.339***
0.293***
0.222***
(0.0149)
(0.0120)
(0.00927)
(0.0151)
(0.0122)
(0.00940)
GDPPCPPP
–0.000716**
–0.000600**
–0.00169***
–0.000763**
–0.000645***
–0.00170***
(0.000320)
(0.000247)
(0.000187)
(0.000319)
(0.000246)
(0.000186)
UNEMPL
–0.235***
–0.205***
–0.179***
–0.227***
–0.199***
–0.175***
(0.0170)
(0.0135)
(0.00931)
(0.0170)
(0.0135)
(0.00941)
L.BANK_LN_ASSETS
–0.000302
–0.000963
–0.00532***
–0.00180
–0.00220**
–0.00597***
(0.00119)
(0.000991)
(0.000692)
(0.00120)
(0.000995)
(0.000707)
L.BANK_LOANS
–0.00649
–0.00450
–0.0136***
–0.0143***
–0.0110***
–0.0171***
(0.00431)
(0.00373)
(0.00243)
(0.00449)
(0.00390)
(0.00257)
L.BANK_EQUITY
–0.0317**
–0.00205
–0.0286***
–0.0298**
–0.000674
–0.0273***
(0.0129)
(0.0107)
(0.00790)
(0.0129)
(0.0107)
(0.00789)
L.BANK_DEPO_GR
0.00140
0.00167
0.00272***
0.00153
0.00179
0.00286***
(0.00135)
(0.00110)
(0.000864)
(0.00134)
(0.00110)
(0.000860)
Cons an
0.0305*
0.0361**
0.141***
0.0499***
0.0522***
0.149***
(0.0175)
(0.0149)
(0.00954)
(0.0176)
(0.0148)
(0.00965)
Obse a ions
1,003,408
1,034,658
1,011,309
1,003,408
1,034,658
1,011,309
R-squa ed
0.043
0.036
0.041
0.043
0.036
0.041
Numbe o FIRM_ID
179,921
183,557
180,751
179,921
183,557
180,751
Time FE
Yes
Yes
Yes
Yes
Yes
Yes
Fi m FE
Yes
Yes
Yes
Yes
Yes
Yes
ADBI Wo king Pape 1468 A. H yckiewicz e al.
15
Simila ly, as in he p e ious eg essions, we also ind ha bank echnological
de elopmen ende s a di e en e ec on sho - e m e sus long- e m c edi g ow h a
SMEs. We no ice ha echnology impac s long- e m SME c edi mo e signi ican ly han
sho - e m SME c edi . Simila ly, as in he linea eg ession, his could sugges ha
echnology is e icien in educing he in o ma ion asymme y by imp o ing da a
collec ion and p ocessing. This could be highly bene icial, especially o he sho - e m
na u e o c edi . Howe e , o long- e m c edi ou inding migh sugges ha bank
echnologies also imp o e o he decision channels (such as, o example, c edi
sco ing models o sc eening p ocedu es). These ad an ages a e pa icula ly signi ican
o banks when making long- e m a he han sho - e m c edi decisions.
In e es ingly, while he s a is ical e ec s emain cons an ac oss di e en de ini ions o
bank ea men g oup (HIGH.DIGITAL and MAX.DIGITAL), we obse e some
he e ogenei y in e ms o economic e ec s. As an icipa ed, he economic in luence o
bank echnological de elopmen on SME c edi g ow h is less p ominen when we
ca ego ize banks in o a ea men g oup, de ined as hose wi h any adop ed solu ion
a e 2015, while he echnological p og ess o hese banks is gauged as he maximum
numbe o solu ions adop ed be ween 2016 and 2019 (Speci ica ions (4)–(6)). This
de ini ion in oduces addi ional he e ogenei y ac oss banks, as banks wi h one solu ion
as well as hose wi h se en solu ions en e he ea men g oup, ende ing di e en
e ec s on SMEs’ c edi g ow h. A he same ime, HIGH.DIGITAL a iable includes
highly echnologically ad anced banks. In e es ingly, he eg ession esul s indica e
ha ou e ec s apply o bo h sho - e m and long- e m SME c edi , hough he e ec on
long- e m SME bo owing is again mo e p onounced. These esul s unambiguously
end o sugges ha SMEs a ilia ed wi h mo e echnologically ad anced banks
expe ience highe c edi g ow h han hose a ilia ed wi h less inno a i e banks. A he
same ime, his highligh s he ans o ma i e po en ial o bank echnological solu ions in
eshaping he landscape o SME bo owing.
As ega ds o he con ol a iables, we ind ha mos coe icien s a e s ongly
s a is ically signi ican and exhibi he expec ed signs. Fo example, unsu p isingly we
obse e ha highe bank deb g ow h is epo ed by younge (LN.FIRM.AGE) and
smalle (FIRM.SIZE) companies wi h he capaci y o inc ease he ole o deb in hei
inancing s uc u e. Howe e , i ms wi h a high asse u no e (ASSET.TURN) o
limi ed sha e o ixed asse s in o al asse s (FIXED.ASSETS) a e mo e likely o be on
he poin o eaching hei p oduc ion capaci y limi s and, as a esul , may be mo e
inclined o aise hem h ough in es men s inanced wi h addi ional deb . In e es ingly,
we also obse e ha mo e p o i able i ms (PROFIT) a e less likely o incu mo e deb ,
which is in line wi h pecking o de heo y: Fi ms i s inance hei in es men ou o
e ained ea nings, which is he cheapes and mos eadily a ailable al e na i e, hen
ou o deb , and las ly by issuing equi y, seen as he mos expensi e op ion o i ms.
3.3 The Role o Di e en Types o Bank Technology
in SME Bo owing
The impac o bank echnological inno a i eness on i m bo owing seems o be a
mul i ace ed issue. So a , ou esul s ha e documen ed ha he le el o bank
echnological ad ancemen migh imp o e an SME’s si ua ion wi h ega d o bank
c edi due o imp o ed access o echnologies suppo ing banks wi h he da a collec ion
and p ocessing, he eby a ec ing he c edi sco ing models and sc eening p ocedu es.
Ye , he exac ole played by di e en indi idual echnological inno a ions in shaping a
bank’s c edi decisions on SMEs’ c edi g ow h emains unclea . To in es iga e his, we
nex e alua e he impac o banks’ echnological solu ions on SMEs’ sho - e m and
ADBI Wo king Pape 1468 A. H yckiewicz e al.
16
long- e m bo owing. This allows us o assess he alue o each solu ion o SMEs’
bo owing. Table 6 p esen s he esul s o ou eg ession analyses. Panel A conside s
he e ec s on long- e m bo owing while Panel B o Table 7 ocuses on sho - e m
bo owing. In addi ion o a bank echnological solu ion ype, we include he gene al
le el o a bank’s echnological inno a ion as a sepa a e con ol a iable in all
speci ica ions. We also epo he esul s by including all solu ions in he same
eg ession model (Speci ica ion (8) o Panels A and B).
Table 6 (PANEL A): The Role o Indi idual Technological Inno a ions
in SMEs’ C edi G ow h
The able p esen s he eg ession esul s o i m- and yea - ixed-e ec panel models examining
he impac o bank echnological inno a ion on SMEs’ c edi g ow h. LT.DEBT.GR indica es he
g ow h o SMEs’ long- e m c edi a ime . BANK.INNOV is a measu e o a bank’s echnological
inno a ion, de ined as he o al numbe o all echnological solu ions adop ed by bank i a
ime . Fo b e i y easons, we do no p esen coe icien s o i m- (PROFIT, FIXED.ASSETS,
EQUITY, ASSET.TURN, LN.FIRM.AGE, and FIRM.SIZE), coun y- (PRI.CREDIT,
GDP.GROWTH, GDP.PC, and UNEMPL), o bank-le el con ol a iables (BANK.SIZE,
BANK.LOANS, BANK.EQUITY, and BANK.DEPO.GR), he cons an e m, and yea dummy
a iables. S anda d e o s, clus e ed a he i m le el, a e shown in pa en heses. *, **, and
*** indica e signi icance a he 10%, 5%, and 1% le els, espec i ely.
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
VARIABLES
LT.DEBT.GR
LT.DEBT.GR
LT.DEBT.GR
LT.DEBT.GR
LT.DEBT.GR
LT.DEBT.GR
LT.DEBT.GR
LT.DEBT.GR
L. ELECTRONIC.
PAYMENTS
–0.00219***
–0.000
(0.000691)
(0.000677)
L. ONLINE.LENDING
–0.00309***
–0.00179**
(0.000735)
(0.000897)
L. PERSONAL_FIN
0.000702
–0.000635
(0.000690)
(0.000858)
L. ANALYTICS
–0.000165
0.00175**
(0.000728)
(0.000738)
L. REG_TECH
–0.000231
0.00108
(0.000746)
(0.000813)
L. BLOCKCHAIN
0.00309***
0.00457***
(0.000629)
(0.000648)
L. AUTOMATIZATION
0.000738
0.00213***
(0.000641)
(0.000663)
L.BANK.INNOV
0.00144***
0.00146***
0.000919***
0.00100***
0.00102***
0.000491***
0.000901***
(0.000195)
(0.000180)
(0.000151)
(0.000159)
(0.000191)
(0.000166)
(0.000156)
Cons an
0.0416***
0.0349**
0.0379**
0.0383***
0.0383***
0.0321**
0.0402***
0.0339**
(0.0149)
(0.0149)
(0.0149)
(0.0149)
(0.0148)
(0.0150)
(0.0150)
(0.0152)
Obse a ions
1,034,658
1,034,658
1,034,658
1,034,658
1,034,658
1,034,658
1,034,658
1,034,658
R–squa ed
0.036
0.036
0.036
0.036
0.036
0.036
0.036
183,557
Numbe o FIRM_ID
183,557
183,557
183,557
183,557
183,557
183,557
183,557
0.036
Time FE
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Fi m FE
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

ADBI Wo king Pape 1468 A. H yckiewicz e al.
17
Table 6 (PANEL B): The Role o Indi idual Technological Inno a ions
in SMEs’ C edi G ow h
The able p esen s he eg ession esul s o i m– and yea - ixed-e ec panel models examining
he impac o bank echnological inno a ion on SMEs’ sho - e m c edi g ow h. ST.DEBT.GR
deno es he g ow h o he SMEs’ sho - e m c edi . BANK.INNOV is a measu e o a bank’s
echnological inno a ion, de ined as he o al numbe o all echnological solu ions adop ed by
bank i a ime . In he in e es o b e i y, we do no p esen coe icien s o i m- (PROFIT,
FIXED.ASSETS, EQUITY, ASSET.TURN, LN.FIRM.AGE, and FIRM.SIZE), coun y-
(PRI.CREDIT, GDP.GROWTH, GDP.PC, and UNEMPL), o bank-le el con ol a iables
(BANK.SIZE, BANK.LOANS, BANK.EQUITY, and BANK.DEPO.GR), he cons an e m, and
yea dummy a iables. S anda d e o s, clus e ed a he i m le el, a e shown in pa en heses. *,
**, and *** indica e signi icance a he 10%, 5%, and 1% le els, espec i ely.
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Va iables
ST.DEBT.GR
ST.DEBT.GR
ST.DEBT.GR
ST.DEBT.GR
ST.DEBT.GR
ST.DEBT.GR
ST.DEBT.GR
ST.DEBT.GR
L. ELECTRONIC.
PAYMENTS
–0.00493***
–0.00400***
(0.000503)
(0.000495)
L. ONLINE.LENDING
–0.00139**
0.00107
(0.000555)
(0.000728)
L. PERSONAL_FIN
0.000702
0.000712
(0.000690)
(0.000629)
L. ANALYTICS
0.000928*
0.00253***
(0.000542)
(0.000558)
L. REG_TECH
0.000732
–0.000577
(0.000711)
(0.000818)
L. BLOCKCHAIN
0.00209***
0.00334***
(0.000456)
(0.000497)
L. AUTOMATIZATION
0.00167***
0.00131**
(0.000474)
(0.000520)
L.BANK.INNOV
0.00126***
0.000456***
0.000919***
0.000145
0.000115
–9.26e-05
5.16e-05
(0.000141)
(0.000143)
(0.000151)
(0.000123)
(0.000165)
(0.000119)
(0.000110)
Cons an
0.150***
0.141***
0.0379**
0.143***
0.142***
0.138***
0.146***
0.146***
(0.00960)
(0.00955)
(0.0149)
(0.00953)
(0.00953)
(0.00959)
(0.00971)
(0.00978)
Obse a ions
1,011,309
1,011,309
1,034,658
1,011,309
1,011,309
1,011,309
1,011,309
1,011,309
R-squa ed
0.041
0.041
0.036
0.041
0.041
0.041
0.041
180,751
Numbe o FIRM_ID
180,751
180,751
183,557
180,751
180,751
180,751
180,751
0.041
Time FE
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Fi m FE
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
The eg ession esul s p o ide compelling e idence. We obse e ha blockchain
echnology s ands ou as a p edominan o ce, demons a ing he mos signi ican
economic impac among he echnologies e alua ed. I s capaci y o p o ide a wide
ange o eal- ime da a seems o be especially ad an ageous in educing asymme ic
in o ma ion and imp o ing c edi sco ing models, as suppo ed by ecen li e a u e
(Yang, Abedin, and Hajek 2023; Zheng e al. 2022). We also ind ha au oma ion
solu ions play an impo an ole in SMEs accessing long- e m inancing. This migh
sugges ha he e iciency o in o ma ion collec ion and p ocessing is ex emely
impo an (Ga g e al. 2021). No su p isingly, ou eg ession esul s documen ha
blockchain has he g ea es e ec on he imp o ed access o SMEs o long- e m c edi .
In e es ingly, ou indings e eal ha online lending solu ions a e nega i ely associa ed
wi h long- e m c edi g ow h, sugges ing ha banks may s ill p io i ize a mix o so and
ha d in o ma ion o e pu ely da a-d i en insigh s o long- e m loan decisions.
Howe e , a he same ime, we no e ha he BANK.INNOV a iable measu ing he
gene al le el o bank echnological inno a ion seems o be highly s a is ically and
economically signi ican , sugges ing ha he gene al le el o bank echnological
de elopmen , i.e., a mix o di e en echnologies adop ed by banks, is impo an while
ADBI Wo king Pape 1468 A. H yckiewicz e al.
18
conside ing he long- e m na u e o SME bo owing. These esul s con i m he complex
na u e o bank long- e m decisions equi ing di e en echnologies suppo ing banks’
c edi decisions.
In con as , ou indings ela ed o sho - e m SME bo owing (Panel B) illumina e he
signi icance o speci ic echnological applica ions such as au oma ion, da a analy ics,
and blockchain echnology. In e es ingly, he agg ega e le el o a bank’s echnological
ad ancemen seems o be less impo an . This can be a ibu ed o he na u e o sho -
e m loans, which a e cha ac e ized by smalle sums and sho e du a ions, equi ing
less exhaus i e da a and a simpli ied isk e alua ion p ocess. Fo such inancial
p oduc s, banks le e age au oma ed echnologies o educe in o ma ion asymme y
and expedi e c edi issuance e icien ly. In e es ingly, ad anced paymen solu ions
nega i ely impac sho - e m unding a SMEs, po en ially due o a c owding-ou e ec .
Meanwhile, well-de eloped elec onic paymen sys ems a banks seem o imp o e
clien s’ liquidi y managemen , educing hei sho - e m bo owing needs, in line wi h
indings om Ca bó-Val e de, Cuad os-Solas, and Rod íguez-Fe nández (2020).
4. THE CHANNELS OF THE TECHNOLOGICAL
EFFECTS ON SME BORROWING
4.1 Reduc ion in In o ma ion F ic ions and Reliance
on Colla e al
So a , ou esul s sugges ha bank echnological ad ancemen s p o ide banks wi h a
wide spec um o di e en da a, which signi ican ly seems o mi iga e he asymme ic
in o ma ion p oblems, educing c edi isk o banks. Fu he mo e, access o a la ge se
o eal- ime da a could allow banks o swi ch o ad anced c edi sco ing models, hus
educing banks’ demand o colla e al.3 The e o e, echnological inno a ion could lead
o elaxed colla e al equi emen s o SMEs, which emains one o he majo ba ie s o
loan access iden i ied in he li e a u e (Beck and Demi güç-Kun 2006; Chan and
Thako 1987; Yaldiz Haneda , B occa do, and Bazzana 2014; Niinimäki 2018).
In his sec ion, we examine he impac o bank echnology on colla e al equi emen s
by in e ac ing a bank’s le el o echnological inno a ion (BANK.INNOV) and he alue
o an SME’s colla e al (COLLATERAL). A s a is ically signi ican coe icien o he
in e ac ion e m (COLLATERAL*BANK.INNOV) would indica e a mode a ing e ec o
bank echnology on he eliance o colla e al o SME bo owing (Speci ica ion (1)).
Fu he mo e, we conside SMEs wi h lowe - alue colla e al (LOW.COLLATERAL)
– hose whose ixed asse o o al asse a io is below he median – and in oduce i
in o he model as ano he in e ac ion e m (LOW.COLLATERAL*BANK.INNOV). The
eg ession esul s a e p esen ed in Table 7.
3 This is in line wi h he BASEL III equi emen s. Acco ding o hese, banks ha use ad anced c edi
sco ing models may ease he equi emen s o colla e al wi hou imposing addi ional capi al (BIS 2017).
ADBI Wo king Pape 1468 A. H yckiewicz e al.
19
Table 7: The Role o Colla e al in Accessing Bank C edi o SMEs
The able p esen s he eg ession esul s o i m- and yea - ixed-e ec panel models examining
he impac o bank echnological inno a ion on SMEs’ c edi g ow h. DEBT.GR ep esen s he
g ow h o an SME’s combined sho - e m and long- e m bank deb a ime , di ided by he
p e ious yea ’s o al asse s (in la ion-adjus ed). LT.DEBT.GR indica es he g ow h o an
SME’s long- e m c edi a ime , while ST.DEBT.GR deno es he g ow h o he SME’s
sho - e m c edi . BANK.INNOV is a measu e o a bank’s echnological inno a ion, de ined as
he o al numbe o echnological solu ions adop ed by bank i a ime . COLLATERAL e e s o
he alue o a i m’s ixed asse s as a p opo ion o i s o al asse s a ime . LOW.COLLATERAL
is a bina y a iable ha akes a alue o one i he i m’s ixed asse alue is below he sample
median, and ze o i i is abo e he median. The in e ac ion e m (COLLATERAL*BANK.INNOV
o LOW.COLLATERAL*BANK.INNOV) includes a one-pe iod lagged colla e al and a
measu e o bank echnological inno a ion (BANK.INNOV) a ime . Fo he sake o b e i y,
we do no p esen coe icien s o i m- (PROFIT, FIXED.ASSETS, EQUITY, ASSET.TURN,
LN.FIRM.AGE, and FIRM.SIZE), coun y- (PRI.CREDIT, GDP.GROWTH, GDP.PC, and
UNEMPL), o bank-le el con ol a iables (BANK.SIZE, BANK.LOANS, BANK.EQUITY, and
BANK.DEPO.GR), he cons an e m, and yea dummy a iables. S anda d e o s, clus e ed a
he i m le el, a e shown in pa en heses. *, **, and *** indica e signi icance a he 10%, 5%, and
1% le els, espec i ely.
(1)
(2)
(3)
(4)
Va iables
DEBT.GR
DEBT.GR
LT.DEBT.GR
ST.DEBT.GR
COLLATERAL*BANK.INNOV
0.00606***
(0.000406)
LOW.COLLATERAL*BANK.INNOV
0.00177***
0.00198***
–3.80e-06
(0.000171)
(0.000136)
(0.000105)
L.LOW.COLLATERAL
–0.0263***
–0.0267***
0.000304
(0.000635)
(0.000524)
(0.000345)
BANK.INNOV
–5.04e-05
0.000868***
0.000432***
0.000823***
(0.000205)
(0.000197)
(0.000166)
(0.000117)
L.COLLATERAL
–0.0615***
–0.0772***
–0.0833***
0.00656***
(0.00177)
(0.00192)
(0.00163)
(0.000949)
Obse a ions
1,003,165
1,003,408
1,034,658
1,011,309
R-squa ed
0.043
0.046
0.041
0.041
Numbe o FIRM_ID
179,904
179,921
183,557
180,751
Time FE
Yes
Yes
Yes
Yes
Fi m FE
Yes
Yes
Yes
Yes
The indings e eal ha he impo ance o a bank’s echnological inno a i eness is
neu alized when a i m has adequa e colla e al (Speci ica ion 1). The in e ac ion e m
(COLLATERAL*BANK.INNOV) eme ges as posi i e and signi ican , sugges ing ha
echnology complemen s i m colla e al in acili a ing unding access o SMEs.
Howe e , he nega i e coe icien o colla e al alone hin s a a po en ial educ ion in he
impac o colla e al on bo owing when bank echnology is conside ed, pa icula ly in
he con ex o mo e echnologically ad anced banks.
In e es ingly, we also obse e a posi i e coe icien o he in e ac ion o low- alue
colla e al and bank echnological inno a ion (LOW.COLLATERAL*BANK.INNOV)
(Speci ica ion 2), sugges ing ha i ms wi h less colla e al bene i om highe c edi
g ow h when aligned wi h echnologically adep banks. The bank’s echnological
inno a i eness a iable i sel is posi i ely co ela ed wi h i m bo owing, while a low
colla e al alue emains a signi ican nega i e ac o . This implies ha echnological
inno a ion may help i ms wi h limi ed colla e al ob ain ex e nal unding by
coun e balancing he nega i e e ec s o low colla e al alue.
ADBI Wo king Pape 1468 A. H yckiewicz e al.
20
Dis inc di e ences a e no ed when seg ega ing he e ec s o sho - e m and long- e m
SME bo owing. The posi i e impac s o bank echnological ad ancemen on colla e al
equi emen s a e pa icula ly p onounced o long- e m c edi , esona ing wi h
academic indings ha emphasize he c i ical ole o colla e al in secu ing long- e m
inancing (Schmalz, S ae , and Thesma 2017). These esul s unde sco e he po en ial
o bank echnology o aid opaque i ms in ob aining long- e m unding, an essen ial
componen o sus ainable business g ow h.
4.2 C edi Access and Cos o In e media ion
The po en ial o echnology o enhance in o ma ion collec ion and p ocessing e iciency
is signi ican , which in u n migh in luence he cos o in e media ion – a no able ba ie
o SMEs seeking ex e nal unding in Eu ope. Thus, unde s anding he e ec o a
bank’s echnological ad ancemen on c edi cos s could p o ide i al insigh s. This
sec ion p esen s eg ession esul s e lec ing he impac o bank echnological
inno a ion on in e media ed bank c edi o SMEs. Table 8 p esen s he eg ession
esul s. Speci ica ions (1)–(2) u ilize s anda d linea models wi h in e ac ion a iables,
while Speci ica ions (3)–(4) apply a DID app oach.
Table 8: The Impac o Bank Technological Inno a ions on he Cos
o C edi o SMEs
The able p esen s he eg ession esul s o i m- and yea - ixed-e ec panel models examining
he impac o bank echnological inno a ion on SMEs’ cos o c edi . INT.COST ep esen s he
sum o all in e es paymen s made by an SME a ime on i s a e age alue o sho - e m and
long- e m bank deb , adjus ed o in la ion. Speci ica ions (1) and (2) de ail he eg essions on
he in e ac ion be ween bank echnological inno a i eness (BANK.INNOV) and yea dummies
o pe iods a e 2015. Speci ica ions (3) and (4) p esen he esul s o di e ence-in-di e ence
eg essions, whe e he ea men e ec began in 2016 and con inues onwa ds. He e, ea ed
banks a e iden i ied as hose ha adop ed mo e han ou echnological solu ions in a gi en yea
(HIGH.DIGITAL) a e 2015. Al e na i ely, digi alized banks a e de ined as hose adop ing any
echnological solu ion a e 2015, wi h he le el o digi aliza ion measu ed by he maximum
numbe o solu ions adop ed h oughou he o al sample pe iod (MAX.DIGITAL). In he in e es
o b e i y, we do no p esen coe icien s o i m- (PROFIT, FIXED.ASSETS, EQUITY,
ASSET.TURN, LN.FIRM.AGE, and FIRM.SIZE), coun y- (PRI.CREDIT, GDP.GROWTH,
GDP.PC, and UNEMPL), o bank-le el con ol a iables (BANK.SIZE, BANK.LOANS,
BANK.EQUITY, and BANK.DEPO.GR), he cons an e m, and yea dummy a iables. S anda d
e o s, clus e ed a he i m le el, a e shown in pa en heses. *, **, and *** indica e signi icance
a he 10%, 5%, and 1% le els, espec i ely.
(1)
(2)
(3)
(4)
Va iables
INT.COST
INT.COST
INT.COST
INT.COST
YEAR2016_DUMMY*BANK.INNOV
0.00150**
(0.000657)
BANK.INNOV
0.000441
–0.00185**
(0.000468)
(0.000851)
YEAR2016_DUMMY*HIGH.DIGITAL
0.00342**
(0.00156)
YEAR2016_DUMMY*MAX.DIGITAL
0.000834**
(0.000406)
Obse a ions
634,770
634,770
634,770
634,770
R-squa ed
0.012
0.012
0.012
0.012
Numbe o FIRM_ID
129,387
129,387
129,387
129,387
Time FE
Yes
Yes
Yes
Yes
Fi m FE
Yes
Yes
Yes
Yes
ADBI Wo king Pape 1468 A. H yckiewicz e al.
27
in e ac ion, would indica e ha banks’ mo e owa ds digi aliza ion has an inhe en
alue in acili a ing SME c edi g ow h ha is sepa a e om any p eexis ing
ela ionships. This dis inc ion is c i ical, as ou HIGH.DIGITAL de ini ion speci ically
includes banks ha ha e unde gone digi aliza ion since 2015. The me hodology,
he e o e, allows us o accoun o his o ical SME bo owing pa e ns be o e 2016, and
obse e he e olu ion o c edi ela ionships ollowing he digi aliza ion shi . The esul s
p esen ed in Table 12 cap u e hese dynamics and p o ide insigh s in o he in luence o
bank echnological ad ancemen on SME inancing.
Table 12: Robus ness: The Role o SMEs’ Digi aliza ion
in he Access o Bank C edi
The able p esen s he eg ession esul s o i m- and yea - ixed-e ec panel models examining
he impac o bank echnological inno a ion on SMEs’ c edi g ow h. DEBT.GR ep esen s
he g ow h o an SME’s combined sho - e m and long- e m bank deb a ime , di ided by
he p e ious yea ’s o al asse s (in la ion-adjus ed). LT.DEBT.GR indica es he g ow h o an
SME’s long- e m c edi a ime , while ST.DEBT.GR deno es he g ow h o he SME’s
sho - e m c edi . BANK.INNOV is de ined as he numbe o echnological solu ions adop ed
by bank i a ime . DIGITAL.FIRM e e s o all i ms ope a ing in he digi al sec o , classi ied
acco ding o NACE codes. This ca ego y includes Compu e p og amming ac i i ies (NACE:
6201); Compu e consul ancy ac i i ies (NACE: 6202); Compu e acili ies managemen
ac i i ies (NACE: 6203); O he in o ma ion echnology and compu e se ice ac i i ies (NACE:
6209); Da a p ocessing, hos ing, and ela ed ac i i ies (NACE: 6311); Web po als (NACE:
6312); Publishing o compu e games (NACE: 5821); O he so wa e publishing (NACE: 5829);
and Re ail sale ia mail o de houses o ia he in e ne (NACE: 4791). HIGH.DIGITAL deno es
highly digi alized banks, de ined as hose exceeding ou echnological solu ions adop ed. Fo
he sake o b e i y, we do no p esen coe icien s o i m- (PROFIT, FIXED.ASSETS, EQUITY,
ASSET.TURN, LN.FIRM.AGE, and FIRM.SIZE), coun y- (PRI.CREDIT, GDP.GROWTH,
GDP.PC, and UNEMPL), o bank-le el con ol a iables (BANK.SIZE, BANK.LOANS,
BANK.EQUITY, and BANK.DEPO.GR), he cons an e m, and yea dummy a iables. S anda d
e o s, clus e ed a he i m le el, a e shown in pa en heses. *, **, and *** indica e signi icance
a he 10%, 5%, and 1% le els, espec i ely.
(1)
(2)
(3)
Va iables
DEBT.GR
LT.DEBT.GR
ST.DEBT.GR
YEAR2016_DUMMY*HIGH.DIGITAL
0.00341***
0.00284***
0.00110***
(0.000582)
(0.000459)
(0.000365)
YEAR2016_DUMMY*HIGH.DIGITAL*DIGITAL.FIRM
0.000217
–0.00400
0.00309
(0.00360)
(0.00249)
(0.00260)
Obse a ions
1,003,408
1,034,658
1,011,309
R-squa ed
0.043
0.036
0.041
Numbe o FIRM_ID
179,921
183,557
180,751
Time FE
Yes
Yes
Yes
Fi m FE
Yes
Yes
Yes
The econome ic analysis yields no e idence o suppo he hypo hesis ha i ms wi h a
p onounced digi al p esence secu ed mo e c edi in he a e ma h o 2015, coinciding
wi h an e a o in ensi ied bank digi aliza ion. Addi ionally, he empi ical da a do no
a i m he p edilec ion o digi ally ad anced i ms o es ablish c edi ela ionships wi h
simila ly digi alized banking ins i u ions. Con a ily, he a iable HIGH.DIGITAL, which
is indica i e o a bank’s echnological ad ancemen , exhibi s a s a is ically signi ican
and posi i e associa ion wi h he g ow h in SME c edi . This obus co ela ion
unde sco es ha he digi aliza ion o banks se es as a mo e c i ical de e minan in
augmen ing SMEs’ access o ex e nal inancing han he digi al a ibu es o he
SMEs hemsel es.

ADBI Wo king Pape 1468 A. H yckiewicz e al.
28
Las ly, we employ a e e se app oach using a andomized sample me hod, whe e
we c ea e a sample comp ising companies om he leas digi alized indus ies. We
hypo hesize ha hese companies a e likely o be om capi al-in ensi e indus ies.
We classi y hem based on hei NACE codes,6 assigning a dummy a iable o one
o companies assigned o hese codes (HIGH.CAPITAL). I a bank’s echnological
de elopmen a ou s mo e digi alized companies han companies ope a ing in less
digi al en i onmen , we could see a s a is ically signi ican e ec o he in e ac ion
e m be ween dummies iden i ying mo e capi al-in ensi e companies (HIGH.CAPITAL)
and mo e digi alized banks (HIGH.DIGITAL) wi h a nega i e sign a e 2016. Table 13
p esen s he eg ession esul s on in e ac ion and HIGH.DIGITAL as a sepa a e
con ol a iable.
Table 13: Robus ness: Randomized Con ol Sample
Using Capi al-In ensi e Fi ms
The able p esen s he eg ession esul s o i m- and yea - ixed-e ec panel models examining
he impac o bank echnological inno a ion on SMEs’ c edi g ow h. DEBT.GR ep esen s he
g ow h o an SME’s combined sho - e m and long- e m bank deb a ime , di ided by
he p e ious yea ’s o al asse s (in la ion-adjus ed). LT.DEBT.GR indica es he g ow h o an
SME’s long- e m c edi a ime , while ST.DEBT.GR deno es he g ow h o he SME’s
sho - e m c edi . BANK.INNOV is de ined as he numbe o echnological solu ions adop ed by
bank i a ime . HIGH.CAPITAL deno es all i ms ope a ing in he nondigi al sec o , classi ied
acco ding o NACE codes. This ca ego y includes codes such as: Cons uc ion o esiden ial
and non esiden ial buildings (NACE: 4120); F eigh anspo by oad (NACE: 4941);
Res au an s and mobile ood se ice ac i i ies (5610). HIGH.DIGITAL deno es highly digi alized
banks, de ined as hose exceeding ou echnological solu ions adop ed. Fo easons o b e i y,
we do no p esen coe icien s o i m- (PROFIT, FIXED.ASSETS, EQUITY, ASSET.TURN,
LN.FIRM.AGE, and FIRM.SIZE), coun y- (PRI.CREDIT, GDP.GROWTH, GDP.PC, and
UNEMPL), o bank-le el con ol a iables (BANK.SIZE, BANK.LOANS, BANK.EQUITY, and
BANK.DEPO.GR), he cons an e m, and yea dummy a iables. S anda d e o s, clus e ed
a he i m le el, a e shown in pa en heses. *, **, and *** indica e signi icance a he 10%, 5%,
and 1% le els, espec i ely.
(1)
(2)
(3)
Va iables
DEBT.GR
LT.DEBT.GR
ST.DEBT.GR
YEAR2016_DUMMY*HIGH.DIGITAL*HIGH.CAPITAL
0.00295
0.00231
0.000641
(0.00194)
(0.00159)
(0.00102)
HIGH.DIGITAL
0.00317***
0.00254***
0.00113***
(0.000594)
(0.000464)
(0.000380)
Obse a ions
1,003,408
1,034,658
1,011,309
R-squa ed
0.043
0.036
0.041
Numbe o FIRM_ID
179,921
183,557
180,751
Time FE
Yes
Yes
Yes
Fi m FE
Yes
Yes
Yes
The eg ession esul s p esen in e es ing indings. We no ice ha he in e ac ion e m
o HIGH.CAPITAL and HIGH.DIGITAL is s a is ically insigni ican . This inding sugges s
ha less digi alized i ms do no seem o expe ience lowe c edi g ow h han any
o he i ms when hey a e associa ed wi h mo e digi ized banks. A he same ime, we
can see ha he HIGH.DIGITAL a iable is highly s a is ically signi ican , indica ing
a posi i e sign. The eg ession esul s suppo ou p e ious inding ha his is a bank
echnological de elopmen ha ex ends access o unding o SMEs and no any
6 These indus ies and espec i e NACE codes a e: 4120 – cons uc ion o esiden ial and non esiden ial
buildings; 4941 – eigh anspo by oad; 5610 – es au an s and mobile ood se ice ac i i ies.
ADBI Wo king Pape 1468 A. H yckiewicz e al.
29
indi idual ea u es o he companies o banks. Mo e speci ically, we also no e ha bank
echnological de elopmen suppo s long- e m c edi g ow h in pa icula , and o a
lesse ex en sho - e m SME bo owing.
6. CONCLUSIONS
The ecen digi aliza ion o inancial se ices has aised a lo o public and academic
deba e on he ole o bank echnology in add essing inancing cons ain s aced by
SMEs. Fo yea s his speci ic g oup o companies has been unde unded by adi ional
banks due o hei in o ma ion opaqueness. The ecen adop ion o echnological
inno a ions by banks has b ough some hope o inc easing da a access and i s e icien
p ocessing, which should lead o be e c edi a ailabili y o hese companies.
This s udy in es iga es he impac o bank echnological inno a ions on bo owing by
SMEs and he way hey ans o m adi ional lending amewo ks o banks. The
esea ch u ilizes a comp ehensi e da ase encompassing 179,921 SME-bank lending
ela ionships ac oss he Eu opean Union om 2009 o 2019.
Ou esul s emphasize ha bank echnological inno a ions ha e a mo e subs an ial
impac on long- e m bo owing by SMEs han on sho - e m bo owing, indica ing
he use ulness o hese echnologies in p o iding da a ha no only educe in o ma ion
asymme y bu also enhance long- e m decision channels, such as banks’ c edi
sco ing p ocesses. Mo eo e , ou eg ession esul s highligh he pi o al ole o
blockchain echnology and au oma ion as echnologies ha enhance he e iciency o
da a collec ion and p ocessing, he eby mi iga ing he di icul ies encoun e ed by SMEs
in ob aining bank c edi .
The s udy also highligh s a dual e ec o echnology on SME bo owing. While i
eases colla e al cons ain s, i pa adoxically aises he cos o c edi o SMEs. This
is especially obse able since he in oduc ion o he Eu opean PSD2, which caused
a signi ican upsu ge in bank digi iza ion. The esul s indica e ha elaxing he
equi emen s o SMEs o access bank c edi may challenge banks o inc ease
he isk p emium due o a po en ially loose ela ionship o a elaxed a i ude owa d
he colla e al.
Ou in es iga ion in o he impac o bank echnological inno a ion on SME bo owing
sugges s a ge ed policy in e en ions. To imp o e he si ua ion o SMEs in hei
bank c edi access, egula o s should incen i ize he adop ion o speci ic echnological
solu ions. Mo eo e , hey mus also ensu e ha such inno a ions do no
disp opo iona ely aise he cos o c edi o SMEs. Suppo ing compe i i e p icing,
especially o he mos opaque i ms, and main aining he bene i s o ela ionship
banking amids digi aliza ion seem o be key. O e sigh o banks’ p icing s a egies is
essen ial o p e en he undue ans e o echnology in es men cos s o SME
bo owe s. Ul ima ely, policies should aim o balance echnological ad ancemen wi h
equi able c edi access, ein o cing he ounda ion o sus ainable economic g ow h
d i en by obus SME inancing.
ADBI Wo king Pape 1468 A. H yckiewicz e al.
30
REFERENCES
Angelini, E., G. di Tollo, and A. Roli. 2008. A Neu al Ne wo k App oach o C edi Risk
E alua ion. Qua e ly Re iew o Economics and Finance 48: 733–755.
Babina, T., S. Bahaj, G. Buchak, F. De Ma co, A. Foulis, W. Go nall, F. Mazzola, T. Yu.
2024. Cus ome Da a Access and Fin ech En y: Ea ly E idence om Open
Banking, NCBR Wo king Pape 32089.
Bake , A. C., D. F. La cke , and C. Y. Wang. 2022. How Much Should We T us
S agge ed Di e ence-in-Di e ences Es ima es? Jou nal o Financial Economics
144: 370–395.
Balyuk, T. 2022. FinTech Lending and Bank C edi Access o Consume s.
Managemen Science 69: 555–575.
Baza bash, M. 2019. FinTech in Financial Inclusion: Machine Lea ning Applica ions in
Assessing C edi Risk, IMF Wo king Pape s 2019/109.
Beaumon , P., H. Tang, and E. Vans eenbe ghe. 2021. The Role o FinTech in Small
Business Lending. Wo king Pape . h ps://paulhbeaumon .gi hub.io/pd s/b .pd
(accessed 20 Decembe 2022).
Beck, T., and A. Demi güç-Kun . 2006. Small and Medium-Size En e p ises: Access
o Finance as a G ow h Cons ain . Jou nal o Banking and Finance 30:
2931–2943.
Beck, T., A. Demi güç-Kun , L. Lae en, and V. Maksimo ic, 2006. The De e minan s o
Financing Obs acles. Jou nal o In e na ional Money and Finance 25: 932–952.
Be ge , T., V. Bu g, A. Gombo ić, and M. Pu i. 2021. On he Rise o Fin echs: C edi
Sco ing Using Digi al Foo p in s. Re iew o Financial S udies 33: 2845–2897.
B anzoli, N., E. Rainone, and I. Supino. 2021. The Role o Banks’ Technology Adop ion
in C edi Ma ke s du ing he Pandemic. SSRN Elec onic Jou nal. A ailable a
SSRN: h ps://ss n.com/abs ac =3878254.
Buchak, G., G. Ma os, T. Pisko ski, and A. Se u. 2016. Fin ech, egula o y a bi age,
and he ise o shadow banks. Jou nal o Financial Economics 130: 453–483.
Ca bó-Val e de, S., P. J. Cuad os-Solas, and F. Rod íguez-Fe nández, 2020. The
E ec o Banks’ IT In es men s on he Digi aliza ion o hei Cus ome s. Global
Policy 11: 9–17.
Chan, Y.-S., and A. V. Thako . 1987. Colla e al and Compe i i e Equilib ia wi h Mo al
Haza d and P i a e In o ma ion. Jou nal o Finance 42: 345–363.
Chiu, J., and T. V. Koeppl. 2019. Blockchain-Based Se lemen o Asse T ading.
Re iew o Financial S udies 32: 1716–1753.
Coombs, C., D. Hislop, S. K. Tane a, and S. Ba na d. 2020. The S a egic Impac s o
In elligen Au oma ion o Knowledge and Se ice Wo k: An In e disciplina y
Re iew. Jou nal o S a egic In o ma ion Sys ems 29: 101600.
Co naggia, J., Y. Mao, X. Tian, and B. Wol e. 2015. Does Banking Compe i ion A ec
Inno a ion? Jou nal o Financial Economics 115: 189–209.
Co nelli, G., J. F os , L. Gambaco a, P. R. Rau, R. Wa d op, and T. Ziegle . 2023.
Fin ech and Big Tech C edi : D i e s o he g ow h o digi al lending. Jou nal o
Banking and Finance 148: 106742.
D’And ea, A., and N. Limodio. 2023. High-Speed In e ne , Financial Technology, and
Banking. Managemen Science 70: 671–1342.
ADBI Wo king Pape 1468 A. H yckiewicz e al.
31
De Blick, T., I. Paeleman, and E. La e en. 2023. Financing Cons ain s and SME
G ow h: The Supp ession E ec o Cos -Sa ing Managemen Inno a ions.
Small Business Economics 62: 961–986.
ECB SAFE Su ey. 2020. Su ey on he Access o Finance o En e p ises in he Eu o
A ea: Ap il o Sep embe 2020. Eu opean Commission, No embe (Sep embe ).
Fe i, G., P. Mu o, V. Pe uzzi, and Z. Ro ondi. 2019. Bank Lending Technologies and
C edi A ailabili y in Eu ope: Wha Can We Lea n om he C isis? Jou nal o
In e na ional Money and Finance 95: 128–148.
Fus e , A., M. Plosse , P. Schnabl, and J. Vicke y. 2019. The Role o Technology in
Mo gage Lending. The Re iew o Financial S udies 32: 1854–1899.
Gambaco a, L., Y. Huang, Z. Li, H Qiu, and S. Chen. 2023. Da a Ve sus Colla e al.
Re iew o Finance 27: 369–398.
Gambaco a, L., Y. Huang, H. Qiu, and J. Wang. 2019. How Do Machine Lea ning and
Non adi ional Da a A ec C edi Sco ing? New E idence om a Chinese
Fin ech Fi m. BIS Wo king Pape s No. 834.
Ga g, P., B. Gup a, A. K. Chauhan, U. Si a ajah, S. Gup a, and S. Modgil. 2021.
Measu ing he Pe cei ed Bene i s o Implemen ing Blockchain Technology in
he Banking Sec o . Technological Fo ecas ing and Social Change 163: 120407.
Ghosh, P., B. Vallee, and Y. Zeng. 2021. FinTech Lending and Cashless Paymen s.
SSRN: h ps://pape s.ss n.com/sol3/pape s.c m?abs ac _id=3766250.
Gopal, M., and P. Schnabl. 2022. The Rise o Finance Companies and FinTech
Lende s in Small Business Lending. The Re iew o Financial S udies 35:
4859–4901.
Ha ison, R., Y. Li, S. A. Vigne, and Y. Wu. 2022. Why Do Small Businesses Ha e
Di icul y in Accessing Bank Financing? In e na ional Re iew o Financial
Analysis 84: 102352.
Holms om, B., and J. Ti ole. 1997. Financial In e media ion, Loanable Funds, and he
Real Sec o . The Qua e ly Jou nal o Economics112: 663–691.
Jag iani, J., and C. Lemieux. 2018. Do Fin ech Lende s Pene a e A eas ha A e
Unde se ed by T adi ional Banks? Jou nal o Economics and Business
100: 43–54.
———. 2019. The Roles o Al e na i e Da a and Machine Lea ning in Fin ech Lending:
E idence om he Lending Club Consume Pla o m. Financial Managemen
48: 1009–1029.
Khandani, A., E., A. J., Kim, and A. W. Lo. 2010. Consume C edi - isk Models ia
Machine-Lea ning Algo i hms. Jou nal o Banking & Finance 34: 2767–2787.
Kwan, A., C. Lin, V. Pu siainen, and M. Tai. 2021. S ess Tes ing Banks’
Digi al Capabili ies: E idence om he COVID-19 Pandemic. SSRN:
h ps://pape s.ss n.com/sol3/pape s.c m?abs ac _id=3694288.
Le ne , J., A. Se u, N. Sho , and Y. Sun. 2023. Financial Inno a ion in he 21s
Cen u y: E idence om U.S. Pa en s. Jou nal o Poli ical Economy o hcoming.
Ma inez Pe ia, M., Y. Timme , N. Pie i, and K. Modi. 2022. The Ana omy o Banks’ IT
In es men s: D i e s and Implica ions. IMF Wo king Pape s 2022(244).
Ouyang, S. (2022). Cashless Paymen and Financial Inclusion. SSRN:
h ps://pape s.ss n.com/sol3/pape s.c m?abs ac _id=3948925.
ADBI Wo king Pape 1468 A. H yckiewicz e al.
32
Palladino, L. M. 2021. The Impac s o Fin ech on Small Business Bo owing. Jou nal o
Small Business and En ep eneu ship 33: 639–661.
Pie i, N., and Y. Timme . 2022. The impo ance o echnology in banking du ing a
c isis. Jou nal o Mone a y Economics 128: 88–104.
Polasik, M., A. Hu e ska, R. I ikha , and Š. Mikula. 2020. The Impac o Paymen
Se ices Di ec i e 2 on he PayTech Sec o De elopmen in Eu ope. Jou nal o
Economic Beha io and O ganiza ion 178: 385–401.
P eziuso, M., F. Koe e , and M. Eh enha d. 2023. Open Banking and Inclusi e Finance
in he Eu opean Union: Pe spec i es om he Du ch s akeholde Ecosys em.
Financial Inno a ion 9: 1–27.
Ro h, J., C., P. H., A. Bilinski, and J. Poe. 2022. Wha ’s T ending in Di e ence-in-
Di e ences? A Syn hesis o he Recen Econome ics Li e a u e. Jou nal o
Econome ics il. 235: 2218–2244.
Schmalz, M. C., D. A. S ae , and D. Thesma . 2017. Housing Colla e al and
En ep eneu ship. The Jou nal o Finance 72: 99–132.
Seduno , J. 2017. Does Bank Technology A ec Small Business Lending Decisions?
Jou nal o Financial Resea ch 40: 5–32.
Sheng, T. 2021. The E ec o Fin ech on Banks’ C edi P o ision o SMEs: E idence
om China. Finance Resea ch Le e s 39: 101558.
Sp ing, M., J. Faulconb idge, and A. Sa wa . 2022. How In o ma ion Technology
Au oma es and Augmen s P ocesses: Insigh s om A i icial-In elligence-Based
Sys ems in P o essional Se ice Ope a ions. Jou nal o Ope a ions
Managemen 68: 592–618.
Wood idge, M. J. 2010. Econome ic Analysis o C oss Sec ion and Panel Da a.
Second Edi ion, MIT.
Wo ld Bank. 2018. Imp o ing Access o Finance o SMEs: Oppo uni ies h ough
C edi Repo ing, Secu ed Lending and Insol ency P ac ices. Wo ld Bank
G oup, May 2018.
———. 2023. Small and Medium En e p ises (SMEs) Finance: Imp o ing SMEs’
Access o Finance and Finding Inno a i e Solu ions o Unlock Sou ces o
Capi al. Wo ld Bank G oup Wo king Pape , May 2023.
Yaldiz Haneda , E., E. B occa do, and F. Bazzana. 2014. Colla e al Requi emen s o
SMEs: The E idence om Less-De eloped Coun ies. Jou nal o Banking and
Finance 38: 106–121.
Yang, F., M. Z. Abedin, and P. Hajek. 2023. An Explainable Fede a ed Lea ning and
Blockchain-Based Secu e C edi Modeling Me hod. Eu opean Jou nal o
Ope a ional Resea ch, o hcoming.
Zheng, K., L. J. Zheng, J. Gau hie , L. Zhou, Y. Xu, A. Behl, J. Z. Zhang. 2022.
Blockchain Technology o En e p ise C edi In o ma ion Sha ing in Supply
Chain Finance. Jou nal o Inno a ion and Knowledge, 7: 100256.
Websi es:
h ps://ec.eu opa.eu/g ow h/smes_en.
BIS (2017): h ps://www.bis.o g/bcbs/publ/d424.pd .