T his compar es with C C Fs under the standardised approach of 20 or 50 for

E C B • T h e N e w Ba se l C a p i t a l A c c o r d • A u g u st 2 0 0 3 10 super visor should deter mine whether or not collater al meets cer tain standar ds, such as the existence of well-established liquid mar kets and publicly available mar ket pr ices paragraph 484. Although supervisors are in a pr ivileged position to assess the liquidity of mar kets, this may not always be feasible given the natur e of some types of collater al e.g. inventor ies such as r aw mater ials, goods, etc.. Accordingly, a reference could be added that banks should utilise available infor mation fr om other competent bodies and author ities and should be able to demonstr ate to their super visor s that their analysis is based on a pr udent evaluation of such infor mation. W ith r egar d to the super visor y slotting cr iter ia for specialised lending SL, banks ar e expected to map their inter nal r atings based on their own cr iter ia accor ding to five categories paragraph 374. Annex 4 provides gener al assessment factor s for each sub-class of SL exposur e. W ith a view to pr omoting a level playing field, fur ther guidance on the way this mapping should be per for med seems to be war r anted for two r easons. Fir st, ther e is the complexity of the mapping pr ocess, as the pr oposed tables with super visor y r ating gr ades intr oduce numer ous indicator s e.g. 18 indicator s for object finance that ar e to be classified accor ding to four assessment categories, namely strong, good, satisfactory, and weak. Second, ther e is the gener alised natur e of some of the elements against which the assessment is to be per for med. For example, in the case of object finance exposur es, the differ ence between satisfactor y and fair mitigation instr uments for assessing the political and legal envir onment or between satisfactor y and fair insurance coverage for assessing insurance against damages may r equir e fur ther clar ification. U nder the foundation IRB appr oach, the pr oposed cr edit conver sion factor C C F for commitments is set at 75 r egar dless of the matur ity of the under lying facility par agr aph

281. T his compar es with C C Fs under the standardised approach of 20 or 50 for

commitments with or iginal matur ities of up to one year or mor e than one year r espectively par agr aph 56. By contr ast, a consistent tr eatment for shor t-ter m self- liquidating tr ade letter s of cr edit is pr oposed under both appr oaches par agr aphs 58 and 284. T he r easons, if any, for applying lower C C Fs for commitments under the standar dised appr oach than under the IRB approach, should be explained. T he use of the word “may” in paragraph 216 is misleading. It should be made clear that banks that use the advanced IRB appr oach “must” pr ovide own estimates of PD , LG D and EAD . Although the heading of paragraph 208 r eads “definition”, the ter ms “r etail r eceivable” and “cor por ate r eceivable” ar e not in fact defined in par agr aphs 209 and 210. Securitisation T he EC B acknowledges the substantial pr ogr ess that has been achieved in the ar ea of secur itisation in par ticular , although some issues may war r ant fur ther consider ation. W ith r espect to the incentive str uctur e, the proposed framework does not always yield the desir ed r esults. For example, low r ated tr anches i.e. BB and BB- attr act a substantially lower capital char ge under the standar dised appr oach than under the IRB appr oach. 14 T his uneven r elation between capital char ge and degr ee of r isk sensitivity is even mor e pr onounced when the pr oposed simplified standar dised appr oach Annex 9 is used in the assessment. Adver se incentives may therefore prevail and high risk exposures may be taken on thr ough innovative and complex instr uments by banks that have r elatively less sophisticated r isk management systems. T his ar gument is r einfor ced by the fact that securities markets offer a much more flexible platfor m to r eact to such incentives when compar ed to the gener al choice of the appr oach to adopt i.e. the IRB ver sus the standar dised appr oach, which may be mor e of a str ategic decision. C ases wher e the 1 4 At the other end of the spectrum, high rated tranches attract a relatively higher weight. 11 E C B • T h e N e w Ba se l C a p i t a l A c c o r d • A u g u st 2 0 0 3 capital char ge is not commensur ate with the r isk sensitivity and sophistication may need to be r evisited in or der to ensur e that the appr opr iate incentive str uctur e is in place. T he pr oposed r ules on secur itisation have become ver y detailed. T his is due par tly to the need to establish a compr ehensive fr amewor k as well as the complexity of secur itisation tr ansactions. In this context, the complexity of these r ules could be r educed by r elying less heavily on income and expenses stemming fr om assets, namely “excess spread”, as defined in paragraph 512, and “futur e mar gin income” FMI, as defined in par agr aph 203. W ithin the fr amewor k of secur itisation, FMI shall be deducted fr om T ier 1 capital par agr aph 523 and is explicitly excluded from calculating a “cap” paragraph

554, which sets upper limits for capital char ges. In contr ast, the tr eatment of