256 To nh ea vr eesrua gl tes, iwn ocrosleu mf onr (t1r a) ,nPc ha ne se l wAi tohf aTaVbEl eS 6m. 5e, tshhoodwt ht ha na t ftohre t cr raendc iht ersa twi ni tgh i sa, FLT method. We observe positive odds ratios of -0.69 (z-stat= 6.18) for the VES method, statistically significant at the 1% level. We find consistent results when wa ned s pt hl iot soeu rwsi tahm pi nl ef r ebqe ut we ne te no rt ri ga inncaht eo sr swiint h cfor leuqmu enn t( 3o)r. i gTi hn ea st oe r sr ei snucl tosl ui mn dni c(a2t)e, that CRAs assign worse credit ratings for VES, on average, than for tranches with the FLT method. Interestingly, we find that only for tranches with frequent omrei gt hi noadt,ocros ,mCpRaAr es da st os i gt hne aF LwTo rms ee t hr aotdi n, gw iotnh apvoesri taigvee foodrd t rr aa nt icohs eos f w1 .i1t h3 t(hz e- s tFaLt E= 3.17) significant at the 1% significance level. We do not find highly significant rr ee spue lat ts tf ho er Oa nBaSl.yIsni so oufr Tr aobblues6t n. 4e sbsuat nnaol wy s ii sn, ccl ou lduemt nh e( 1f o) lol of wT ai nb gl eaId, Ad ipt ipoennadl i cxoInI , t wr oel variables; STS Compliant, Single Originator, GDP Growth Rate and Country of Risk cseovnetrroallsa.dWdietisohnoawl cothnatrtoolsurtoreosuurltmsoodfeTl.able 6.4 remain robust when including Given data limitations, it is difficult to determine the exact reason for our findings. It could be that the underlying pool is on average riskier for VES transactions.81 In the FLT method, the retainer takes a significantly larger share of the first losses compared to the VES method, which can be interpreted as a signal to markets for confidence of the retainer in the underlying pool. WC ReAosbespearrvaet es ll yi gihnt Pl ya nd ei flf eBr oe nf Tt arbe lseu 6l t.s5 .wThheenr ec souml tps asrhi no gw tthhea tc rMe odoi tdrya’ st i (ncgosl uomf ena1c )h, St r&a Pn c(hc eosl uwmi tnh 2t )h, ea Vn dE SDmB Re tSh (ocdo, l cuommnp3a )r eadl l taos tshi genF aLTwmo resteh or adt. i An gr, eos nu l at vceornasgi es t, ef on rt 81 Following discussions with major rating agencies, it seems that CRAs do not specifically consider the different rr ii ss kk i nr eetsesnot if ot nh emuentdheordl ys i ni ng tphoeri trf oc lrieod, ri ta trha et irntgh ma notdheel sr ias nk dr ,e at es nstui oc nh , mi te at hpopde ai trsse tl fhtaht atthi es scer ee dn iat sr ca rt ienagt i nr egphriegsheenrt sr i tshk es for investors (e.g. because of the return profiles and incentives associated to them).
RkJQdWJsaXNoZXIy MTk4NDMw