International SF Weekly: LR/RWA considerations for securitisation

Research evidence excerpt

International SF Weekly: LR/RWA considerations for securitisation

Commentary

In a previous weekly commentary we discussed the ECB working paper on synthetic

securitisation: we commended the researcher on the quantitative review of the

European synthetic securitisation sector and questions some of their qualitative

conclusions.

Let’s rehash the report’s conclusions and our reservations

The report arrives at three main conclusions and we find issues with all three of them:

First, banks select loans for SRT to strategically reduce their effective capitalization.

The capitalization of banks is determined by their capital amounts and by the risk

weights assigned to their loans. The report claims to causally show that the banks

synthetically transfer loans with high risk weights relative to their economic riskiness,

and redeploys the released capital in new lending with lower risk weight relative to their

economic riskiness.

In fact, the data presented in the report suggests to us that the SRT loans are of better

credit quality than non-SRT loans (as indicated by credit performance and pricing), but

they may attract similar RWs as non-SRT loans. The discrepancy may be a result of both

application of SA and F-IRB, rather than A-IRB, with the former two generating capital

farther away from the economic risk of the loans.

We believe that the banks are optimising capital usage, given high cost of capital, and

realigning capital with true economic risks rather than seeking to deliberately reduce

their effective capitalisation as the paper concludes.

Second, banks reduce their monitoring efforts of the firms whose loans they synthetically

transfer.

The report concludes that after a loan is synthetically transferred, the bank reduces the