Heterogeneour Homebuyers, Mortage Choice and the Use of Mortage Brokers.

He e ogeneous Homebuye s, Mo gage Choice and he use o Mo gage B oke s Da id Du ya*, Mau ice J. Rocheb aDa id Du y, The Economic and Social Resea ch Ins i u e, 4 Bu ling on Road, Dublin 4, I eland. Tel: 353-1-6671525, Fax: 353-1-6686231, Email:da id.du y@es i.ie * Co esponding au ho bMau ice J. Roche, Depa men o Economics, The Na ional Uni e si y o I eland Maynoo h, Maynoo h, Co. Kilda e, I eland. Tel:353-1-7083786, Fax: 353-1-7083934, Email:m[email p o ec ed] 1 Abs ac Choosing a mo gage p oduc in he ace o labo income isk, in e es a e isk and bo owing cons ain s is one o he mos impo an decisions acing a household. This pape in es iga es he choice be ween a a ie y o ixed a e mo gages and adjus able a e mo gages. We ind ha households wi h a high loan- o- alue a io, isky income and high isk a e sion a e mo e likely o choose a ixed a e mo gage. Choosing a mo gage p oduc elies ma ke sea ch and in o ma ion. The pape inds ha in gene al i s - ime homebuye s and hose wi h a high loan- o- alue a io a e mo e likely o use a mo gage b oke . JEL Classi ica ion: E40, G21, R51 Keywo ds: Mo gage choice, Fi s - ime homebuye , Mo gage b oke , In o ma ion, P eda o y Lending 2 1. In oduc ion Home pu chase ep esen s a majo inancial commi men o he homebuye . Fo he majo i y, he inance equi ed is bo owed om a lending ins i u ion. The mo gage ma ke is highly compe i i e and lende s now o e a wide a ie y o mo gage p oduc s wi h di e en ixed, adjus able and acke in e es a e e ms. In addi ion, lende s o e discoun ed adjus able and ixed a e p oduc s o a ac new business and e ain exis ing cus ome s. In his pape we ocus on wo aspec s o mo gage choice; household isk managemen and in o ma ion sea ch. In a ecen heo e ical pape on mo gage choice and household isk managemen , Campbell and Cocco (2003) show ha in a li e-cycle model wi h bo owing cons ain s and income isk, an adjus able a e mo gage (ARM) “is gene ally a ac i e, bu less so o a isk- a e se household wi h a la ge mo gage, isky income, high de aul cos , o low mo ing p obabili y.” In his pape he choice o mo gage p oduc is examined using a mul inomial logi model. Ou mic o da ase allows us o analyse he mo gage choice o wo he e ogeneous g oups, namely, i s ime buye s and epea buye s. Ou esul s indica e ha i s ime buye s p e e ixed a e mo gages (FRM) ha ha e he longes e m. This is in acco dance wi h he p edic ions o he heo e ical model o Campbell and Cocco (2003). In o ma ion plays an impo an ole in he mo gage choice ma ke , gi en he ange o lende s and a ailable p oduc s ha ace he bo owe 1. I he chosen p oduc is o be he mos e icien o he consume hen he consume needs o ha e ull knowledge o he ange o mo gage p oduc s a ailable and hei p ices. Homebuye s can unde ake an in o ma ion sea ch hemsel es. Al e na i ely, a mo gage b oke will p o ide in o ma ion on he ange o mo gage p oduc s a ailable in he ma ke , o a mo gage agency will p o ide in o ma ion on p oduc s o e ed by he i ms ep esen ed by ha b oke . Gi en he ange o lende s and mo gage 1 Miles (2004) p esen a use ul diag am on he low o in o ma ion and unds in he UK mo gage ma ke . 3 p oduc s i migh be expec ed ha buye s, o o e come in o ma ion de iciencies, would use a mo gage b oke o ind hem he mos a ac i e op ion. Howe e , a ea u e o he mo gage ma ke is asymme ic in o ma ion. This in o ma ion asymme y can ake wo o ms – mo al haza d and ad e se selec ion. In he case o mo al haza d he bo owe (p incipal) canno obse e he e o being made on his/he behal by he mo gage b oke (agen ). Wi h ad e se selec ion he bo owe is unable o judge he e ec i eness o he mo gage b oke . Gene ally mo gage b oke s ecei e hei commission om he lende (see Anglou and A no (1991) o di icul ies associa ed wi h commission con ac s o es a e agen s, pg 112). Thus, he homebuye manages o a oid he cos o unde aking an in o ma ion sea ch in he expec a ion ha he b oke will sou ce he mos e icien mo gage o hem2. Howe e , Miles (2003, 2004) sugges s ha his may no always be he case as “in e media ies ha e some inancial incen i e o sell sho - e m discoun ed p oduc s wi h he p ospec o a esale in he nea u u e”. Gu en ag (2001) dis inguishes be ween up on mo gage b oke s and con en ional mo gage b oke s. The la e end o engage in p eda o y lending. Ou da ase iden i ies i he homebuye used a mo gage b oke o no . Ou expec a ion would be ha bo owe cha ac e is ics would be impo an de e minan s o whe he o no a b oke is used. Fi s - ime homebuye s wi h li le o no expe ience o he housing ma ke a e conside ed mo e likely o use a mo gage b oke as hey a e seeking o o e come in o ma ion asymme ies.3 Wi h mo gage b oke s o e ing an oppo uni y o bo owe s o o e come in o ma ion de iciencies and choose a mo e e icien mo gage he cha ac e is ics ha in luence he decision o use a mo gage b oke a e examined in his pape . In a logi model we examine he in luence o household cha ac e is ics on use o a mo gage b oke . The analysis inds ha , 2 O cou se his assumes ha he b oke is no in luenced by di e en commission a es o encou age he ake-up o pa icula p oduc s o p oduc s om a pa icula lende – he p incipal agen p oblem. 3 The da ase we use in his pape indica es ha jus o e hal o bo owe s use a mo gage b oke . 4 in gene al, he bo owe s ha a e mo e likely o use a mo gage b oke con o m o p io expec a ions, pa icula ly i s - ime homebuye s o hose wi h a high loan- o- alue a io. In he mul inomial logi model o mo gage choice we ind ha he use o a mo gage b oke as an explana o y is s a is ically signi ican and educes he likelihood ha all ypes o homebuye would choose any mo gage o e a discoun ed one-yea ixed a e mo gage. The la e ype o mo gage happens o be he cheapes in ou da ase . I is sugges i e ha mo gage b oke s a e no engaging in p eda o y lending. The es o he pape is o ganized as ollows. In Sec ion 2 we p esen a e iew o he li e a u e. Ou mic o da ase is desc ibed in Sec ion 3. Sec ion 4 p esen s ou esul s. The las sec ion o e s conclusions. 2. Rele an li e a u e One o he ea ly pape s o examine he issue o mo gage p oduc choice is Alm and Follain (1987), who de elop a wo-pe iod heo e ical model o mo gage choice. The household chooses he amoun s o housing, non-housing consump ion, isky in es men asse s and non- isky in es men asse s, wi h he eal app ecia ion a e o house p ices being one o a numbe o andom a iables. They use wo app oaches o mo gage choice in his en i onmen . In he i s app oach he consume compa es u ili y le els be ween an adjus able and ixed a e mo gage. In he second app oach he in e es a e di e en ial ha makes he consume indi e en be ween he wo mo gage ypes is compu ed. One o he main conclusions is ha as households a e assumed o be isk a e se and an inc ease in unce ain y abou he mo gage a e inc eases he p obabili y ha a consume chooses a ixed a e p oduc . They also ind ha a la ge mo gage inc eases consume p e e ence o a ixed a e p oduc as a la ge mo gage implies a g ea e 5 amoun o in es men in a isky asse i an adjus able a e mo gage is chosen. Alm and Follain (1987) also include he ole played by expec ed capi al gain om housing and exis ing household asse s. B ueckne (1986) also de elops a wo-pe iod heo e ical model o mo gage choice and ake accoun o in e es a e caps and ma gins. He eaches a numbe o conclusions: bo owe s who place a high alue on u u e consump ion a e likely o op o a ixed a e p oduc as hey p e e a igh in e es a e cap4; bo owe s wi h a apidly ising income s eam a e likely o a o an adjus able a e p oduc , as a e bo owe s who make la ge downpaymen s as bo h hese bo owe ypes ha e a p e e ence o a loose in e es a e cap. Dhillon, Shilling and Si mans (1987) examine empi ically he impac o p icing and bo owe cha ac e is ics on he choice o mo gage con ac . A p obi model is used wi h he choice being limi ed o be ween one ype o ixed a e mo gage and one ype o adjus able a e mo gage. They ind ha he mo gage p ice a iables a e all signi ican . Gene ally, bo owe cha ac e is ics ei he ha e a weak impac o a e insigni ican in de e mining he mo gage in e es a e choice. Households wi h co-bo owe s, ma ied couples o a sho expec ed enu e ha e a endency o p e e adjus able a e mo gage p oduc s. O he cha ac e is ics such as age, educa ion, i s - ime homebuye and sel -employmen a e insigni ican . In con as o he B ueckne (1986) he empi ical indings also sugges ha bo owe s wi h g ea e weal h would seem o p e e adjus able a e mo gages B ueckne and Follain (1988) include egional dummies in hei p obi model o mo gage choice. They also deal wi h he issue o he unknown al e na i e in e es a e. In gene al da a is a ailable o he ype o mo gage chosen by he bo owe . Da a is no a ailable on he ange o al e na i es conside ed and ejec ed. They a gue ha his migh be a sou ce o 4 B ueckne e e s o an in e es a e cap as he maximum inc ease be ween pe iods in he adjus able mo gage a e. 6 po en ial selec i i y bias. This po en ial bias is due o he ac ha he bo owe can be assumed o ha e chosen a a o able mo gage ype. They ind li le e idence o selec i i y bias and main ain ha his is o be expec ed “when consis en p icing policies lead o li le a ia ion in he FRM-ARM a e di e en ial wi hin ma ke s”. Bo owe cha ac e is ics a e gene ally no ound o be signi ican , wi h only income and a a iable iden i ying in e ci y mo e s, being close o signi ican . I is wo h no ing ha hei income e ec is con a y o hei expec a ion wi h he empi ical esul s sugges ing ha high-income bo owe s a e mo e likely o choose an adjus able a e mo gage han low-income bo owe s. Howe e , hey no e ha income a iable exhibi s a - a io ha is only close o being signi ican and he s eng h o he income e ec is modes . The impo an a iables explaining mo gage choice a e he di e en ial be ween ixed and adjus able a es and he le el o ixed a es. Phillips and Vande Ho (1991) ex end he basic model by decomposing he di e en ial be ween ixed and adjus able a es o ake accoun o ini ial discoun ed a es. The ini ial discoun ed a e is gene ally a p omo ional a e on o e by lende s o a ac new cus ome s. Usually his is o a se ime pe iod i.e. a one yea educed a e. They ind ha mo gage choice is p ima ily de e mined by ela i e mo gage cos s wi h he ini ial discoun being he mos impo an ac o in luencing adjus able a e choice. Phillips and Vande Ho (1994) conclude ha ela i e p icing and local a ea economic and housing ma ke condi ions a e he main ac o s de e mining mo gage choice. Si mans and Fe ei a (1995) examine he p icing o housing and mo gage se ices, using a mul iple logi model o de e mine he p obabili y ha he homebuye is a i s - ime o a epea homebuye . On he basis o hese esul s di e ences in housing and mo gage cha ac e is ics a e examined o epea and i s - ime homebuye s. They ind ha i s - ime homebuye s exhibi a 7 sligh ly g ea e p e e ence o ixed a e mo gages. Howe e , he e may be a p oblem o endogeni y wi h hei esul s. Type o mo gage is one o he a iables being used o explain whe he o no a homebuye is a i s - ime o a epea homebuye . Howe e , he decision o choose a ixed o adjus able a e mo gage may also de e mined by whe he he homebuye is a i s - ime o a epea homebuye Sa-Aadu and Megbolugbe (1995) ex end he analysis o mo gage choice by using a mul inomial logi model o ake accoun o di e ences in he leng h o he in e es a e ixed e m. They ind ha he impac o mo gage p ice a ies ac oss he al e na i e mo gage p oduc s. Di e ences in bo owe cha ac e is ics, pa icula ly mobili y and a o dabili y, ha e an in luence on he ype o mo gage con ac chosen. The ole o in o ma ion in he housing ma ke is an impo an one. Some homebuye s, such as i s - ime homebuye s o hose who ha e ecen ly mo ed o an a ea, can ha e di icul y accessing in o ma ion abou he local housing ma ke . The ole o in o ma ion on house p ices is examined by Tu nbull and Si mans (1993), who used homebuye cha ac e is ics as p oxies o he le el o homebuye in o ma ion and sea ch cos s. They ind no e idence o signi ican di e ences in house p ices ac oss di e en ypes o homebuye and conclude “exis ing ins i u ions, such as mul iple lis ing se ices, success ully amelio a e many o he po en ial p ice e ec s o asymme ic in o ma ion and cos ly sea ch”. In a ecen pape on he issue o mo gage choice Campbell and Cocco (2003) de elop a heo e ical model o iden i y household cha ac e is ics ha should lead he household o p e e one mo gage ype o ano he . The pape places emphasis on mo gage choice as pa o household isk managemen . Households a e assumed o be isk a e se and o ace bo h income and in e es a e isk. The pape p esen s a ange o esul s based on assump ions abou 8 bo owing le els and he sou ces o isk and unce ain y. They ind ha i is op imal o households wi h s able income and a small mo gage o choose a FRM. In an applica ion i his model o he UK, Miles (2004) inds ha “a signi ican p opo ion o households – hough p obably no a majo i y – migh be expec ed o ind ha he ad an ages o e y long- e m ixed a e mo gages make hem a ac i e.” 3. The da ase Ou analysis is applied o he housing ma ke in he Republic o I eland. The e has been massi e housing boom in he las decade wi h eal house p ices inc easing by 9.6% pe annum be ween 1993 and 2003 and he numbe o mo gages aken ou wi h inancial ins i u ions inc easing by 6.4% pe annum o e he same pe iod. S a is ics o he I ish mo gage ma ke sugges ha he popula i y o adjus able in e es a e mo gage p oduc s has luc ua ed o e he pas decade. The mid-1990s saw adjus able a e p oduc s accoun ing o jus a hi d o he mo gage ma ke , wi h ixed a e p oduc s accoun ing o o e 67 pe cen . Howe e , lowe in e es a es associa ed wi h I eland’s en y o Economic and Mone a y Union in 1999 has seen a e i al o o unes o he adjus able in e es a e p oduc which cu en ly accoun s o o e 52 pe cen o mo gage p oduc s on mo gage loans paid, Figu e 1. [FIGURE 1 HERE] The da a o his pape is d awn om a single lende , pe manen sb. Pe manen sb is a na ional lende and was p e iously a building socie y be o e con e ing o a bank in 1994. The company is he la ges mo gage p o ide and he sample is ep esen a i e o he I ish mo gage ma ke . The da ase o mo gages paid con ains de ails abou mo gage p oduc , e m and in e es a e a issue, as well as bo owe and some house s uc u e cha ac e is ics. The o iginal 9 ime homebuye s and 45% o epea homebuye s used he se ices o a mo gage b oke . This is simila o he UK whe e Miles (2004) ound ha “ he p opo ion o i s - ime buye s using in e media ies is consis en ly highe han o o he ca ego ies o bo owe s and is cu en ly close o 60 pe cen .” In a compe i i e ma ke like he mo gage p oduc ma ke he e exis s a wide ange o mo gage ypes p o ided by a numbe o lende s. Thus, any homebuye en e ing he ma ke , be hey a i s - ime homebuye o a epea homebuye , aces a wide a ie y o in e es a es and p oduc s. In o ma ion on he a ailable in e es a es by ype o mo gage p oduc is published weekly in he p ope y supplemen s o he na ional newspape s. Howe e , he po en ial homebuye s ill has o asce ain how much hey can bo ow gi en hei cu en income and exis ing le el o ou goings. In o he wo ds, while in o ma ion on p ices is eadily a ailable in o ma ion on mo gage “quali y” is mo e di icul o access. Salop and S igli z (1977) de elop a heo e ical model o consume s in he insu ance ma ke whe e consume s ace un o eseen in o ma ion cos s. Those who know he dis ibu ion o p ices will buy ba gains while hose wi hou in o ma ion will buy andomly. In hei conclusions hey pu o wa d he no ion ha “in he p esence o some in o med consume s, unin o med consume s ough o ‘buy wi h he ma ke ’; p ice will e lec quali y and ma ke sha es will e lec he o e all ‘bes buys’”. In o ma ion asymme y in he mo gage and housing ma ke p obably ep esen s mo e o a challenge o he i s - ime homebuye who is lea ning abou he housing ma ke in i s en i e y, whe eas he epea homebuye has he bene i o expe ience gained in p e ious ansac ions. In he ace o impe ec in o ma ion he homebuye s ill wishes o pu chase he mos e icien mo gage p oduc o hei needs. Howe e , unde aking an in o ma ion sea ch ep esen s a cos o he homebuye in e ms o ime and income o egone. Howe e , i he sea ch is no 16 unde aken hen selec ion o a sub-op imal mo gage p oduc may mean highe mon hly epaymen s. I may also esul in a lowe le el o bo owing, es ic ing he homebuye ’s house pu chase abili y. One op ion is o he homebuye o unde ake an in o ma ion sea ch hemsel es in he expec a ion ha he in es men o he ime and ene gy equi ed will allow hem o make he bes choice. Al e na i ely, he homebuye can use he se ices o a mo gage b oke o unde ake he sea ch wi h a iew o maximizing he amoun o bo owings hey can unde ake, o o minimize he mo gage se ice cos on hei bo owings. A logi model is used o empi ically measu e he impac o di e en homebuye and mo gage p oduc cha ac e is ics on he choice o use a mo gage b oke o no . The pape hen examines he in luence o such a iables on he decision by a homebuye o use a mo gage b oke . The model akes he o m:    1 exp 1 exp , 1,...... ,         J jh h Py j j J  xx x (2) whe e y is he dependen a iable, wi h a alue o 0 o 1, and x is he se o explana o y a iables. The model is es ima ed using non-linea maximum likelihood me hod. The esul s o he logi model es ima ing he likelihood ha a homebuye will use a mo gage b oke a e gi en in Table 7. A Wald es ejec s he null hypo hesis ha he coe icien s a e all equal o ze o. All he a iables a e signi ican excep household a e - ax income and he numbe o bo owe s. High loan- o- alue a ios inc ease he p obabili y ha he homebuye will use a mo gage b oke . One possible explana ion is ha as he loan- o- alue a io inc eases he need o access he mos compe i i e in e es a e possible means ha bo owe s use a mo gage b oke o unde ake a mo gage p oduc sea ch. As he mo gage e m inc eases he likelihood o using a mo gage b oke also ises. This may e lec a o dabili y issues wi h bo owe s using mo gage b oke s o access bo owing o e a longe mo gage e m o keep epaymen s 17 manageable. Being male, in pe manen employmen and buying a dwelling loca ed in Dublin all inc ease he likelihood o using a b oke . The impac o he Dublin loca ion a iable may well e lec easie access o mo gage b oke s ha o hose li ing in u al a eas. The highe he na ional a e age adjus able in e es a e he highe he p obabili y ha he homebuye will use a mo gage b oke . The a iable iden i ying i he pu chase is a i s - ime homebuye o no is signi ican and sugges s ha i s - ime homebuye s a e mo e likely o use a mo gage b oke . This is line wi h he a gumen ou lined abo e ha i s - ime homebuye s a e, by hei na u e, new o he homeowne ship and mo gage ma ke s. Mo gage b oke s a e used in o de o o e come in o ma ion de ici s and o ensu e he mos app op ia e mo gage in e es a e, possibly mo i a ed by a o dabili y. [TABLE 7 HERE] The able also shows he esul s o he logi model on use o a mo gage b oke o he sample spli by i s - ime and epea homebuye s. A Chow es o he sample spli , epo ed in Table 7, allows us o ejec he null hypo hesis ha he coe icien ec o s a e he same o each ype o homebuye . The impac o he explana o y a iables is b oadly he same as o he ull da a se . In he case o i s - ime homebuye s a lowe numbe o household cha ac e is ics a e signi ican , while he pu chase p ice o he p ope y is much mo e signi ican han o epea homebuye s. The “Cases Co ec ” alue indica es he numbe o obse a ions o which he p edic ed alue ma ches he ac ual alue, ha is, whe e he p obabili y alue is 0.5 o be e . This measu e is sugges ed by Woold idge (2002) and indica es ha he model pe o ms qui e well. I is wo h no ing ha his measu e does no ell us any hing abou he quali y o he p edic ion. An al e na i e based on he log-likelihood is also calcula ed (see Pindyck and Rubin eld (1991) and 18 Woold idge (2002)) gi ing a pseudo R squa ed o .0575. Howe e , Woold idge (2002) no es “goodness o i is no as impo an as s a is ical and economic signi icance o he explana o y a iables.” 5. Conclusions This pape has examined he impac o homebuye and mo gage p oduc cha ac e is ics on he choice o mo gage p oduc . In con as o some o he s udies homebuye cha ac e is ics we e ound o be signi ican in hei impac . Howe e , in common wi h he exis ing li e a u e he mo gage p ice a iables ha e he bigges impac . Use o a mo gage b oke educes he likelihood o choosing an adjus able a e mo gage. The esul s would appea o con i m he sugges ion o Miles (2004) ha b oke s encou age he use o sho - e m p oduc s. Howe e , his is a en a i e conclusion and me ely poin s o an a ea o u he esea ch. The pape also examines he impac o di e en cha ac e is ics by ype o homebuye and inds ha he impac o di e en cha ac e is ics a ies by ype o homebuye ac oss he ange o mo gage p oduc s. The pape hen ex ends he analysis o mo gage choice o conside wha de e mines he use o a mo gage b oke . The conce n o he homebuye is o y and pu chase he mos e icien p oduc ha sui s hei needs. Like any la ge ma ke o e ing a wide ange o p oduc s his can be hampe ed by sea ch cos s o by lack o in o ma ion. Homebuye s equi ing a mo gage can use he se ices o a mo gage b oke o o e come a lack o in o ma ion. In gene al, he a iables ha inc ease he likelihood ha a mo gage b oke will be used con o m o p io expec a ions. Loca ion is an impo an wi h an u ban loca ion inc easing he likelihood ha a mo gage b oke will be used. Fi s - ime homebuye s o hose wi h a high loan- o- alue a io a e mo e likely o use a mo gage b oke . 19 Re e ences Anglou, P.M., A no , R., 1991. Residen ial Real Es a e B oke age as a P incipal-Agen P oblem. J. Real Es a e Finance Econ. Alm, J., Follain, J.R.,1987. Consume Demand o Adjus able Ra e Mo gages. Housing Finance Re . 6, 1-16. B ueckne , J.K., 1986. The P icing o In e es Ra e Caps and Consume Choice in he Ma ke o Adjus able-Ra e Mo gages. Housing Finance Re . 5, 119-136. B ueckne , J.K., Follain, J.R., 1988. The Rise and Fall o he ARM: An Econome ic Analysis o Mo gage Choice. Re . Econ. S a is . Vol. LXX, Numbe 1, Feb ua y. Campbell, J.Y., Cocco, J.F., 2003. Household Risk Managemen and Op imal Mo gage Choice. Qua . J. Econ. No embe . Dhillon, U.S., Shilling, J.D., Si mans, C.F., 1987. Choosing be ween Fixed and Adjus able Ra e Mo gages. J. Money, C edi , Banking. Vol. 19, No. 1, Feb ua y. Follain, J.R., 1990. Mo gage Choice. AREUEA Jou nal. Vol. 18, No.2 Gu en ag, J., 2001. Ano he View o P eda o y Lending. The Wha on Financial Ins i u ions Cen e wo king pape numbe 01-23-B. Miles, D, 2004. The UK Mo gage Ma ke : Taking a Longe -Te m View. Final epo and Recommenda ions. The S a ione y O ice, UK, Ma ch Phillips, R.A., Vande ho , J., 1991. Adjus able- Ve sus Fixed- a e Mo gage Choice: The Role o Ini ial Ra e Discoun s. J. Real Es a e Res. Vol. 6, Numbe 1. Phillips, R.A., Vande ho , J., 1994. Al e na i e Mo gage Ins umen s, Quali ica ion Cons ain s and he Demand o Housing: An Empi ical Analysis. J. Ame . Real Es a e U ban Econ. Assoc. Vol. 22 No. 3, pp. 453-477. 20 Sa-Aadu, J., Megbolugbe, I.F., 1995. He e ogeneous Bo owe s, Mo gage Selec ion, and Mo gage P icing. J. Housing Res. Vol. 6, Issue 2, Fannie Mae. Salop, S., S igli z, J., 1977. Ba gains and Ripo s: A Model o Monopolis ically Compe i i e P ice Dispe sion. Re . Econ. S a is . Vol. XLIV, Numbe 138, Oc obe . Si mans, G.S., Fe ei a, E.J., 1995. The P icing o Housing and Mo gage Se ices o Fi s - ime e sus Repea Homebuye s. J. Real Es a e Res. Vol. 10, Numbe 1. S eine , F., 2000, Quan i ying Disc imina ion in Home Mo gage lending: Es ima ion o Loan P ice Elas ici ies ac oss P oduc s and Races, S an o d Ins i u e o Economic Policy Resea ch Discussion Pape No.00-15, Decembe . Tu nbull, G.K., Si mans, C.F., 1993. In o ma ion, Sea ch and House P ices. Reg. Sci. U ban Econ. 23, pp. 545-557. Woold idge, J.M., 2002. Econome ic Analysis o C oss Sec ion and Panel Da a. The MIT P ess, Camb idge, USA/London 21 Figu e 1: Mo gage loans paid by ype o in e es a e 0 10 20 30 40 50 60 70 80 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 % Fixed Adjus able Sou ce: DOELG, Annual Housing S a is ics Bulle in, a ious issues 22 Table 1 F equency o p oduc by epea homebuye , i s - ime homebuye and o al ma ke Repea Homebuye Fi s -Time Homebuye To al Ma ke Type o mo gage p oduc % % % 1 y ixed 53.7 61.5 57.8 2 y ixed 4.6 5.7 5.2 3 y ixed 3.7 4.8 4.3 4 y ixed 0.2 0.2 0.2 5 y ixed 2.0 2.4 2.2 7 y ixed 0.1 0.1 0.1 10 y ixed 0.5 0.4 0.5 Fixed e m unknown 0.1 0.1 0.1 1 y adjus able a e 26.1 20.9 23.4 Adjus able a e 9.0 4.0 6.4 To al 100.0 100.0 100.0 23 Table 2 Desc ip i e s a is ics by homebuye ype and o al ma ke Repea Homebuye Fi s -Time Homebuye To al ma ke Fixed in e es a e mo gage* 64.9 75.1 70.3 Adjus able in e es a e mo gage* 35.1 24.9 29.7 New House* 24.4 46.8 36.3 Loan amoun , € 103,708 115,040 109,715 House P ice, € 190,182 150,260 169,018 Loan- o-Value a io 57.1 77.7 68 House P ice/income a io 4.4 3.5 3.9 Loan/income a io 2.2 2.6 2.4 Mo gage epaymen / a e - ax income, % 19.5 17.7 18.6 Mo gage e m, yea s 21 25 23 Co-bo owe * 71.9 68.8 70.2 Ma ied* 56.7 18.4 36.4 Numbe o dependen s 0.8 0.2 0.5 Age 37 30 33.5 House size, sq 1,306 1,192 1,245 * % wi hin each homebuye ype 24 Table 3 Mo gage choice by ype o homebuye Dependen a iable o mul inomial logi model Y Repea Homebuye Fi s -Time Homebuye To al % % % 1 1 y ixed (discoun ed) 54.2 61.9 58.3 2 2 and 3 y ixed 8.4 10.6 9.6 3 4 and 5 y ixed 2.0 2.4 2.2 4 1 y adjus able a e (discoun ed) 26.4 21.1 23.6 5 S d. Adjus able a e 9.0 4.0 6.4 100.0 100.0 100.0 25