144 mm aa rj okreitt y( i o. ef. ,ttrhaen cUhSeasni nd Eo uUrms aamr kpelte) saer ep ai srsaut eeldy. bPya nr eellaCt iov fe Tl ya lbalreg4e. 2i ssshuoewr ssi tnhtaetrtmh es of both principal balance size (60.70%) and frequency (53.41%). 4.3.3 Control Variables We include several control variables to capture characteristics of the underlying dy ee aa rl , os fu icshs ua as nt rcaen. Wc hee rceopuonrtt, tcha ep idt ae ls ca rl liopct ai vt ei osnt ,attri as tni sc as catni odnvvaar li ua be ,l ec rdeids it tr irba ut itni og n, asni nd Panels A to C of Table 4.2. Tranche Count equals the total number of tranches in a corresponding RMBS deal. In our total sample, the mean tranche count is 15.86. Log Tranche Value equals the natural logarithm of the face value of a tranche at issuance. The mean Log Tranche Value over the whole sample is 17.40. Capital Allocation is the level of credit support for a tranche and the mean is about 44% over the whole sample. Log Transaction Value equals the natural logarithm of the tt rr aa nn sc ahcet ii os na vpaalrute) (mi . ee a. , stuhreefda ci ne mv aillul ieo, nast oi sf sUuSa ndcoel,l aorfst. hTehteo mt a el aRnM B S o f w h i c h t h e Log Transaction Value of the sample is 21.12. Coupon is the coupon rate assigned for each tranche at issuance, the mean coupon rate of our full sample is 2.60%. Number of ratings is the number of ratings assigned to a tranche at issuance; the number of ratings range from 1 to 3 and the mean number of ratings over the whole sample is 1.59. Wvaeriaablsleo, control for the market share of the issuer. We include a dummy Top Ten Issuer, that equals one if the issuer is among the top 10% of iasmsuoenrgs tmheearseumreadinuinsgin9g0t%he. global RMBS market share, and zero if the issuer is 46 We also control for the geographic location of the underlying collateral for each tranche (Geography of Collateral). If the geographic location of the underlying collateral is widespread (i.e., 40% in California and 46vaWriaebdloe ninotthinactlmudoedtehl.is variable to test our third hypothesis since we use the issuer size measure as an independent
RkJQdWJsaXNoZXIy MTk4NDMw