I find it quite fascinating that there is almost nothing in the social impact bond literature- from the insightful Brookings report to the amazing Social Finance database – on how to price a social impact bond.
There are some figures shared here and there, but without any substantiation or figures helping calculate the IRR. In my opinion, this is mainly due to the following reasons: i) most investors in Social Impact Bonds are foundations and/or investors with concessional money that are mainly motivated by testing out the instrument, not setting the perfect pricing ii) SIBs do not follow a standardized structure, some are designed as unsecured loans, others as an equity investment in an SPV etc. It is hard to get transparent pricing data out of these iii) SIBs returns might depend on a large number of scenarios and outcomes, many are still being designed while launched, and hence a straightforward pricing data is challenging iv) it is just confidential as most private transactions
I was asked at a panel how I would price one, and here what I broadly said:
I would look at it more or less as an unsecured loan to the service provider ( IRR of ~15%) and then I would add a premium and/or discount based on the following factors: i) country risk ii) outcome payer risk iii) implementation risk iv) modelization and outcome risk v) service provider risk, and I would also look at additional ones such as: availability of a first loss, length of tenor, comprehensive default clauses etc. It is definitely a complicated task.
It would help a lot to start building a SIB pricing benchmark: i) building a “risk score” based on the factors above for every SIB ii) calculate the corresponding average IRR iii) generate the resulting pricing curve.
Easier said than done, but I hope the Social Finance folks could look at this as a next step.