Using private capital to fund impact delivery is an idea that has been around for a long time but never found a practical model that could scale, in spite of interest and resources from large incumbents across public and private sectors.
What we have instead is a large and noisy Impact Investment space, packed with traditional business models that operate in a field that has inherent impact and/ or traditional models operated by businesses that take a commendable ethical position on elements of their implementation.
Meanwhile, the absolute majority of impact delivery happens using traditional donor funds & traditional implementation models.
What we miss is a model where private capital is invested in impact delivery at scale AND investors can expect to make a decent return on their investment.
I believe developing such a model at scale is possible today.
Impact delivery Today
First, a very quick look at the way funding works in traditional impact delivery (head here for more details).
- A Donor makes available funds for a specific area;
- Implementers bid for these funds;
- An award is made to one implementer/ a consortium;
- The donor pre-funds activities & carries all the risk;
- A monitoring frameworks tracks mostly inputs/ processes throughout implementation;
- At the end of the project, an external Evaluation is completed;
- This whole cycle takes 5–10 years
In the model above, ALL funding comes from public sources (government/ taxpayer and/ or foundations); You can find some some private money in these models — i.e. when the funder is a private company with a corporate social responsibility fund — but this rarely changes the way impact is delivered.
In this model:
- The funder carries ALL the risk;
- The funder (or anyone else for that matter) has no idea of how the implementation progresses;
- Because of that, implementers are fundamentally trust mediators — the best of them are geared for managing donors, rather than delivering impact;
Over the last decade or so, a few innovative models have been articulated that blend private and public resources in the funding of impact. One of the more interesting among them is the Social Impact Bond. Here is the theory:
- Private investors purchase a bond governed by an impact condition (“within the next 5 years, child mortality in Community X will be reduced by Y%”).
- The capital thus raised is invested in activities that are likely to contribute to achieving the outcome (“reduction in child mortality in community X”);
- If the condition is fulfilled, the donor will pay the private investors a return;
This is a fantastic model. However, in practice, there are some significant limitations:
- Implementation/ Impact delivery remains traditional — traditional implementers get contracted to deliver outputs, exactly like in the donor-funded model;
- Because of that, there is no way to gauge progress towards outcome;
- Because of that, the bond is not liquid — no regular investor will buy it as there is no way for them to build a risk profile;
- Additionally, agreement on evidence that the outcome has been achieved is highly complicated, which means impact bonds take years to negotiate;
- Finally, gathering that evidence/ completing the evaluation is both expensive and time consuming;
Because of the above, the impact bond remains a highly exotic, illiquid instrument in which only a precious few large investors are willing to invest and which takes a long time to structure — making it useless for any rapid deployment of capital;
Today we have the means to address the above limitations and structure a new type of asset that would serve the same purpose as the impact bond but that would be (i) liquid; (ii)accessible to any investor; (iii) very fast to structure; and (iv) profitable to investors;
Here is how it would work:
- We start with a theory of change/ assumption. For example, we propose the assumption that 1 million assisted deliveries over 5 years in Country X is likely to lead to a 5% decrease of newborn mortality;
- We define as reference for newborn mortality the country’s Demographic & Health Survey (DHS) which is published every 5 years;
- Every time an assisted delivery happens in Country X, a token is issued that corresponds to that unique, individual delivery. That means that total tokens will be capped to 1,000,000 (corresponding to the no. of assisted deliveries in our theory of change);
- These tokens are made available to investors on exchanges worldwide — the initial valuation is simply the cost of delivery;
- Because they are available on exchanges, these tokens are liquid — people can buy, sell or “hodl” this token as they please/ need;
- The day when the next DHS comes out (within the next 5 years), a snapshot gets taken of the token blockchain ledger. Every address that contains vaccination tokens will receive a dividend payment (for example 10%) from the government of country X/ a donor; This means essentially that (impact) investors receive dividend payments;
This model serves the same purpose as an impact bond, but it has none of the limitations of the traditional impact bond:
- Impact delivery is funded with private capital / risks are privatized;
- This capital is not “do-good” capital; Investors stand the opportunity to make a decent return ;
- Donors only pay for success; they can reduce their back-office costs significantly;
- Investors get rewarded for investing in impact delivery;
- Implementers and local communities can start innovating, since the only metric that matters is the pre-agreed output/ outcome (assisted deliveries in our example, leading to a measurable decrease in child mortality);
- Significant savings come from a reduction in the need for Monitoring & Evaluation;
- Token investments are accessible to wide segments of investors — regular individuals can invest a few dollars and institutional investors can invest millions — impact investment could become part of any investment portfolio;
- Finally, such models can be deployed pretty much over night, at scale, which means they can be used to effectively raise private capital as a response to emergencies;
We are currently looking at structuring this model in a concrete use case. Obviously, the trickiest part is defining an acceptable model for the distributed verification of the individual impact events (the proof of impact). Once that challenge is resolved — and it will — models like this can go to scale super fast.