154 away from AAA, compared to the other rating assigned for the same tranche. We ft ihned5s%i msi il ga nr irf ei csaunl tcse wl ehveenl (wc oel uu ms eno2u)r amn ad rckoent ssihs at er ne tmh ei gahs luyr sei gbny i ff ri ceaqnuternecsyu, lat sl bwe iht eant we use our combined market share measure for large CRAs (columns 3 and 4). Next, we study the tranches rated by both DBRS and S&P in columns (5) to (8) of Panel A. In column (5) we find that the odds ratio of MS S&P by Balance is negative, but only significant at the 10% significance level. We do find negative significant results, at the 5% level, for our MS S&P by Frequency measure of market share in column (6) of Panel A. We find that a one standard deviation increase in MS S&P by Frequency increases the odds of experiencing a worse rating by DBRS than by Ssh&aPrefomr ethaesusraemfoertrlaarngceheC.RCAosn,sistent results are found for our combined market MS Large CRAs by Balance, with the odds of −3.89, statistically significant at the 5% level (z-statistic of −1.97), column (7). Ionf as ul amr g, et hCeRrAe sr ue sl tusl tos f, oPna na ev le rAa go ef , Ti na bal eh i4g.h4e sr ul igkgeel si ht ot oh da totf hae whoi gr hs ee rr amt i anrgkbe yt sDhBaRr Se trha tainn gt hs ,eorna tai vnegraasgsei ,gtnheadn bMy oMo odoy d’ syw’ s hf oe rn tMh eo os admy ’ es itsr ai nn cc rheea. sSion,gDi Bt sRmS ai sr kg ei vt i snhgawr eo. r s e Wt o e( 4f i)nodf oPpa pn oe sl iBt eorf eTsaubl tl es 4f o. 4r . t rI annccohl ue smrna t(e1d) bo yf PKaBnRe Al Ba, nwdeMf i onoddtyh’ as ti nt hceool udmd sn rsa(t1i o) of MS Moody’s by Balance is positive, but only statistically significant at the 5% lienvel (z-statistic of 2.22). This indicates that a one standard deviation increase MS Moody’s by Balance increases the odds of experiencing a better rating by Kr eBsRu Al t st h( aagna fi nr o amt tMh oe o5d%y ’ ss, i fgonri fti hc aenscaeml eevterla) nfcohr et. hWe emfai nr kdect osnhsai rs et e mn te saisgunrief isc a n t MS Moody’s by Tranche in column (2), and for our combined market share measure for large CRAs in columns (3) and (4). However, if we move to the tranches rated by KBRA and S&P in column (4) to (8) in Panel B, we find no significant results at all for our market share measures.
RkJQdWJsaXNoZXIy MTk4NDMw