Emerging markets and developing economies (EMDEs) clearly want to attract more private capital. What’s less clear is how different investors think about that prospect—their return expectations and risk perceptions—especially when it comes to participating in blended finance structures that are design to attract them to EMDE assets.
A new survey from three researchers who study blended finance sheds lights on those questions and the conditions under which de‑risking works.[1] This blog highlights some of the results that caught our eye (there is much more in the paper) and what we think explains two provocative facts about measuring additionality and choosing risk mitigation tools.
The authors surveyed 141 senior leaders—CIOs, CROs, and CEOs—across a broad mix of investors involved in blended finance. Results are reported along two dimensions:
- Market exposure: investors active in both EMDEs and developed markets (DMs) vs those who are DM‑only;
- Capital type: concessional (below‑market) vs market‑rate investors.
Strikingly, no respondents cited the usual suspects—insufficient returns or limited risk tolerance—as the primary reasons why they would be unwilling to invest in EMDE blended finance funds. Instead, the main constraints are institutional: concessional capital providers say the main thing that holds them back is limited resources, while commercial market‑rate investors put most blame on the onerous process of agreeing blended finance structures. More than half of commercial investors also report a lack of in‑house blended‑finance expertise and mandates that restrict EMDE allocations.
Among investors who operate in both EMDEs and DMs, the study finds no statistically significant difference in their return expectations or risk tolerances across the two markets. The survey also asks how investors perceive various risks in EMDE investing, from low to very high, most of which are rated at medium or worse, but these responses are not compared to risk perceptions in DM markets so it is hard to piece together whether investors perceive the risk-adjusted returns on offer in EMDE markets to be superior or inferior to those in DM markets. Nonetheless, the authors suggest that the equality of return expectation implies that low allocations to EMDE assets may represent an inefficient portfolio choice, forgoing viable opportunities and diversification benefits.
It would be interesting to see further disaggregation amongst private investors. Blended finance is often a question of designing a structure to meet the requirements of asset managers who are pursuing distinct investment strategies on behalf of different types of asset owner. Last year BII and BCG published Scaling Blended Finance: Practical tools for Blended Finance Fund design, where we identified a set of archetypes that most of the blended finance funds that we see in the market fall into. To help fill in some of the blanks about the requirements of different types of investors in those archetypes, more fine-grained information would be useful.
Two provocative facts
The survey uncovers two provocative facts about the measurement of additionality and the nature of risk mitigation instruments which we think reflects the nature of most blended finance funds. Let’s take additionality first. They write: “One clear pattern arises: the majority of respondents fail to isolate the marginal contribution of their interventions” and “do not identify the counterfactual level of investment and impact in the absence of their intervention”. They call for more rigorous assessments of financial and impact additionality.
We suspect that investors do not know how to respond to survey questions about how they measure additionality because there is nothing that they can measure. In the right circumstances researchers can collect data and estimate the difference between what happens after an intervention and what would have happened in its absence.[2] But that comparison against a counterfactual is not something practitioners can measure in the act. Instead of measuring additionality, practitioners can try to supply reasons to believe that additionality is present.[3]
A motivation for blended finance, or for some of its archetypes, is the belief that there are markets with unmet demands for capital that are large enough to warrant the attention of the big pools of capital controlled by asset managers in wealthy countries, but where the investments on offer do not quite meet the requirements of those asset managers.[4] Blended finance can close the gap between investors and investments and, in so doing, smooth away some of the “frictions” that impede investment so that capital will flow more freely in the future. That suggests there are two claims that we should try to substantiate to provide reasons to believe additionality is present: that there is an unmet demand for capital and that the investment opportunities on offer do not satisfy the requirements of global asset managers.
If you are negotiating with a large asset manager who says they would be unwilling to invest in the assets in question without a blended finance structure, they could be fibbing and opportunistically trying to receive a subsidy. But if they are being truthful then there is financial additionality at fund level – meaning that the asset managers would not be allocating capital to buying these assets without the risk-return adjustment conferred by the blended structure.
Measuring additionality
What is there for anyone to measure? Practitioners could describe – and research could uncover – the characteristics of the assets that the fund proposes to invest in and then compare them to the reasonable requirements of asset managers, with reference to market benchmarks. That is the best we can hope for, if we want a reason to believe in additionality at the fund level. But that is not measuring additionality in a broader sense – what we might call “market level” additionality – which is the question of whether creating this fund with capital from these asset managers will deliver what we really want: an increase in the volume of investment in the targeted market.
We won’t know how much investment would have occurred in the market in the absence of the fund (the counterfactual) because some of the assets financed by the fund would probably have received some finance from elsewhere, in its absence. Measuring impact additionality is therefore equally difficult. We might be able to form an idea of the impact of the underlying projects in question, but we cannot say how many of those projects would have gone ahead in the absence of a fund that has attracted asset managers from wealthy countries.
Providing evidence of unmet demand for capital is difficult. We are all familiar with ubiquitous “SDG financing gaps” but those are not evidence of unmet demand for capital. There is evidence that persistent barriers to cross-border capital flows exist (here is one example), which result in lower levels of investment in neglected countries, but these investment gaps are not something practitioners can measure. It would be difficult enough for anyone to keep track of firms and projects that want capital but fail to find it, but nobody can observe how many more firms and projects would be raising capital if it was more easily obtained at a lower cost. If we want reasons to believe that participants in blended transactions are making a marginal contribution to impact on the ground, the existence of unmet demand must be the foundation of that belief. That is not something we can ask asset managers to measure, but researchers might.
Choosing risk mitigation tools
The authors also notice that the two risks in EMDEs that investors are most worried about are currency and political risks and write “surprisingly, when inquiring about the type of risk-sharing tools and concessional capital used in blended finance projects, the large majority of respondents highlight the (near) absence of risk management provisions designed to address currency risk (e.g., cross-currency swaps) and political risk.” The most common risk mitigation tool in blended finance is a “first loss” tranche.
We think this probably reflects that fact that the most common variety of blended finance fund is a multi-country long-term debt fund. If you are worried about political and currency risk in that context, a first loss tranche has you covered (up to a point) whatever the reason for default. There can sometimes be a gain from splitting out specific risks and allocating them to specialists who can charge a lower fee than implicit in the returns flowing to the investors in a first-loss tranche, but because most blended finance funds are multi-country, political risk insurance would probably be complicated to arrange.[5] Many blended finance structures are also for long term hard currency loans. Currency risk is still a worry – devaluations can result in dollar shortages and defaults – but I don’t know whether there is a currency swap solution to that problem. For long term local currency loans, currency swaps can also be very expensive or unavailable.
This survey was published as a working paper, and the research is still in the exploratory phase. We look forward to the next iteration and encourage blended finance practitioners participate in events hosted by the Sustainable Investing Research Initiative and have their say.
[1] These researchers – Professor Caroline Flammer, Geoffry Heal and Thomas Girouz, have also published analysis of concessionality in blended finance and a survey of private biodiversity finance.
[2] It not easy to do with data that has simply been collected before and after an intervention, hence the popularity of randomised control trials to create “treatment” and “control” groups.
[3] This was the conclusion we reached in The Elusive Quest for Additionality: “DFIs should take a probabilistic approach to evaluating additionality, focusing on identifying the circumstances in which an investment is more or less likely to be additional.”
[4] There are other archetypes, such as those that attempt to attract risk-bearing venture capital to young and innovative enterprises.
[5] Or maybe not – please let me know if I am wrong about that. Another problem is that investors maybe prefer the automatic protection of first-loss, rather than waiting to see if an insurance claim is successful.
