32 2.3 Data and Methods We begin the process by manually collecting data obtained from Bloomberg, w$ 1h. 0i c5h tpr irlol ivoind,etshaa tc owme rpel ei tses uuendi vaenr ds es oo fl d8 ,i3n2t4h eC LUOS torra nEcUh ems awr ki teht saf rt oo tma l 1v9a9l u6e uopf tcoh a2r0a1c 5t e. rFi os tri cesa, c ph rCi cLeOddaet ea ,l , tthhee 3d-amt aosne tt hp rboevnicdhems da reka/l raenf de rterna cnec hi ne tne ar ems et sr, ai tses uf oe rr the floating-rate tranches, credit ratings, balance and primary issuance spread.14 ches are rated by either Moody’s or S&P, or both. There are an insufficient number oT fh eCrLeOf osr er ,a ti end obuyr Fd iat tcahs eot r woet hue sr e s mo nallyl e trr aCnRcAhse st ot heant a obbl et asi nt aetdi s tai c raal t iannga l fyrsoems . Moody’s and/or S&P, consistent with the dataset used by Griffin et al. (2013). We apply several filters to our dataset and remove tranches with incomplete it nh feo rnmu ma t bi oenr . oBf eccraeudsiet wr aet i an rges , i nwt ee roe ns tl ye di ni nc l ut hdee ei nf f eocut ro fs tCuLdOy s CdLeOa sl ct or amn pc hl eexsi t wy iot nh af rto lme a8s ,t3 o2n4 et oc r7e, 9d 1i t0r. aWt ien gf u dr ti hs celro ds ei sdc aartdi sasl luter. aTn hc hi se sr ewdiut hc ems ios us irn go rtirgai nn saal cstai omnpol er tranche size (14 tranches) and missing information on the funding cost at issue (305 tranches). This filtering resulted in a full sample of 7,591 CLO tranches, of wt hhe i cEhU 5m, 9a3r5k et tr.aPnac nh ee lss aAr et oi sDs uoefdTianb tl eh e2 .U1Sr me paorrktes t saunmdm1a, 6r 5y 6s ttar tains tcihc es sf oi srstuheed Ui nS and EU market, respectively. 14 If fixed-rate tranches were to be included in our study, then it would be necessary to determine the appropriate benchmark yield curve for each tranche in the sample in order to obtain primary issuance spreads that could be consistently compared across the sample. By restricting the tranches in our sample to 3-month floating-rate tranches where the reference rate is the same interest rate benchmark, we avoid this problem. Furthermore, in constructing the final sample, we had to eliminate some tranches due to errant data or metrics that represented v co as s t t l . y atypical observations. For our analysis, we want to have a consistent benchmark for assessing the funding
RkJQdWJsaXNoZXIy MTk4NDMw