ToniAhnert,SebastianDoerr,NicolaPierri, YannickTimmer WorkingPaperSeries DoesIThelp?Informationtechnologyinbankingandentrepreneurship No2956 Disclaimer:ThispapershouldnotbereportedasrepresentingtheviewsoftheEuropeanCentralBank(ECB).TheviewsexpressedarethoseoftheauthorsanddonotnecessarilyreflectthoseoftheECB. Abstract Westudytheimportanceofinformationtechnology(IT)inbankingforentrepreneurship.Guidedbyaparsimoniousmodel,weestablishthatjobcreationbyyoungfirmsisstrongerinUScountiesmoreexposedtobankswithgreaterITadoption.Wepresentevidenceconsistentwithbanks’ITadoptionspurringentrepreneurshipthroughacollateralchan-nel:entrepreneurshipincreasesbymoreinIT-exposedcountieswhenhousepricesrise.FurtheranalysissuggeststhatITimprovesbanks’abilitytodeterminecollateralvalues,inparticularwhencollateralappraisalismorecomplex.ITalsoreducesthetimeandcostofdisbursingcollateralizedloans. JELclassification:D82,G21,L26. Keywords:technologyinbanking,entrepreneurship,informationtechnology,collateral,screening. Non-technicalsummary Focus Informationtechnology(IT)hasdramaticallychangedhowinformationisusedinthefinancialsector.Thismayaffectthesupplyofcreditfrombanks,asakeyfunctionofbanksistoscreenandmonitorborrowers.Lendingtoopaqueborrowers,suchasyoungfirmsandstart-ups,islikelytobeespeciallysensitivetosuchchangesinIT.Thereasonisthatyoungfirmshavenotyetproducedsufficientquantitativeinformation,suchasbalancesheetdata.Instead,lendersrelyonsoftinformation.Asstart-upscontributedisproportionatelytojobcreationandproductivity,butareoftenfinanciallyconstrained,understandinghowtheITrevolutioninbankinghasaffectedstart-ups'accesstofinanceisofparamountimportance.Yet,directevidencefortheimpactoflenders'ITcapabilitiesonentrepreneurshipisscarce. Contribution First,ourpaperrelatestotheliteratureinvestigatingtheeffectsofITinthefinancialsectoroncreditprovisionandsmallbusinesses.Second,wespeaktopapersthatanalysetheimportanceofcollateralforentrepreneurialactivity.Weprovidefirstevidencethatbanks'ITadoptionincreasestheimportanceofcollateralinbanks'financingofyoungfirms.Third,wecontributetotherecentliteraturethatinvestigateshowtheriseoffintechaffectscredit-scoringandcreditsupply.AnadvantageoffocusingonthevariationinITadoptionamongbanksisthatourresultsareunlikelytobeexplainedbyregulatoryarbitrage,whichhasbeenshowntobeanimportantdriverofthegrowthoffintechs. Findings Webuildamodelinwhichbankscanscreenfirmseitherbyacquiringinformationaboutfirmsandtheirprojectsorbyrequiringcollateral.Crucially,ITmakesitcheaperforbankstoanalysehardinformationandthusrelyoncollateralisedlending.Thisbenefitsstart-ups,astheyhavenotyetproducedsufficientinformationandhavetobescreenedthroughtheuseofcollateral.ThemodelthuspredictsthatITinbankingwillspurentrepreneurship–andthemoresowhencollateralvaluerises.Consistentwiththemodel'simplications,wefindthatjobcreationbystart-upsisstrongerinUScountieswithIT-intensivebanks,especiallyduringperiodsofrisingcollateralvalues. 1Introduction Theriseofinformationtechnology(IT)inthefinancialsectorhasdramaticallychangedhowinformationisgathered,processed,andanalyzed(e.g.Vives2019).Thisdevelopmenthasimportantimplicationsforcreditsupply,asoneofbanks’keyfunctionsistoscreenandmonitorborrowers.Financingforyoungfirmsislikelytobeespeciallysensitivetosuchchangesinlenders’technology.Theyhaveproducedlimitedinformationavailabletolendersandoftenrelyonbanksasasourceoffunding(RobbandRobinson,2014;Babina,2020). ThispaperstudieshowtheriseofITinthefinancialsectoraffectsentrepreneurship.Ourempiricalanalysisisguidedbyasetofhypothesesderivedfromaparsimoniousmodelinwhichbanks’screeningstrategydependsontheirlevelofITadoption.WefirstestablishthatUScountiesinwhichmoreIT-intensivebanksoperateexperiencemorejobcreationbystartups.Guidedbythemodel’spredictions,wethenpresentevidenceconsis-tentwithbanks’ITadoptionfacilitatingentrepreneurshipthroughacollateralchannel:thepresenceofIT-intensivebanksstrengthenstheresponsivenessofjobcreationbyen-trepreneurstochangesinlocalrealestatevalues.Thispatternisespeciallypronouncedinindustriesthatrelymoreonrealestatecollateralandinmarketswhereevaluatingcollateralismoredifficult.FurtheranalysissuggeststhatITadoptionalsoreducesthecostandtimeofdisbursingcollateralizedloans. WemeasureITadoptionatthebank-branchlevelastheratioofPCsperemployee,followingseminalpapersonITadoptionamongnon-financialfirms(Bresnahanetal.,2002;BrynjolfssonandHitt,2003;Beaudryetal.,2010;Bloometal.,2012).ThissimplemeasureofITadoption,whichisbasedonlyonhardwareavailability,isastrongpredictorofalternativemeasures,suchastheITbudgetoradoptionoffrontiertechnologies,buthasmuchbetterdatacoverage.Wepurposelyfocusonbanks’generaladoptionofIT,rather thanspecifictechnologies(e.g.ATMsoronlinebanking),becauseofthemulti-purposenatureofIT. FollowingPierriandTimmer(2022),wedecomposeITadoptionintoabankfixedef-fect,whichconstitutesourmainmeasureofITatthebanklevel,andaborrower-countyfixedeffect.Thisadditivemodelallowsustopurgebank-levelITadoptionfromunob-servablelocalfactorsthatcouldaffectboththeITinten