The catalytic effect of blended finance
Ugo Panizza is one of the few economists with long-standing interest in development finance. His 2019 paper Smart Development Banks anticipated the recent trend towards thinking about impact at a market level, by arguing that development banks should see their role as providing intelligence to the wider development community, about what problems are holding markets back. His latest paper is The Catalytic Effect of Blended Finance.
This is a theory paper intended to generate some foundational principles. As with most theory papers, it is not easy to bridge the gap between a few lines of algebra and blended finance in practice. This blog discusses the theory from a practitioner’s point of view.
The paper explores the performance of blended finance as a response to three underlying problems: there are positive externalities (social benefits that are not captured in the financial returns of private investors), which results in private markets under-investing from society’s point of view; there are financial frictions that mean financing (loans) is too expensive so there is too little investment; there is “credit rationing” so that lenders restrict the quantity of finance they are willing to provide, so there is too little investment.
The first thing that will strike practitioners as odd is that the theory starts from a position where the private sector will invest in a project, it just won’t invest enough. Blended finance is deployed to make the project larger. In practice, although some project parameters can be changed, as a rule the size of a project is already decided (the GW capacity of the solar park) and blended finance is used to persuade commercial actors to finance it.
The second thing that will strike practitioners as odd is how mobilization ratios are calculated. Professor Panizza defines it as the amount by which the project gets bigger divided by the fiscal expenditure required to deliver that change.[1] Practitioners measure it as the ratio of the public to private contributions in a transaction, without asking how much larger the transaction is thanks to the public contribution.[2]
The ultimate objective of blended finance is to increase the volume of investment, and the impact that creates, so measuring how much investment increases relative to how much that cost to achieve is in theory a better way of defining the mobilization ratio. However, it cannot be done in practice for reporting purposes because we have no way of knowing how much less investment would have happened in the absence of a blended finance intervention, all we can observe in the act is what happens within the financing of a transaction.[3]
Professor Panizza also make a crucial assumption of diminishing returns. That means for every extra dollar put into a project, the incremental effect on the output of the project gets a little smaller each time. That assumption “guarantees that the optimal project has a finite size” (because after a while, making it any larger is not worth the cost) but it is another assumption that might not chime with practitioners.[4] Many businesses exhibit increasing returns to scale, at least over a range of potential sizes. There is no reason to think that adding a second line to a green ammonia plant will cost more per unit of output than the first (it will probably cost less).
This combination of assumptions lies behind one of the papers main results: the catalytic effect of blended finance declines as market failures becomes more severe. This result applies across (almost) all the market failures the paper looks at, but it is easiest to explain for positive externalities. If positive externalities are very large, that means the optimal size of the project is much larger than the private sector would deliver. But because of diminishing returns, the further you try to increase project size the more expensive that gets, so the fiscal cost of blended finance starts to catch up with how much the size of the project has increased. In extreme cases to reach an optimal size you might spend $11 to increase the project size by $10, so the multiplier falls below 1.[5]
Whether a blended finance transaction is a good idea depends on how much social welfare increases, relative to the fiscal cost, not on the size of the mobilization multiplier. So Professor Panizza’s conclusion that high multipliers should not be interpreted as better than low is correct and important, and one that practitioners emphasize.[6] But I do not think that practitioners should take away the lesson that larger positive externalities imply greater fiscal costs (and hence lower multipliers) – at least, not at the level at which practitioners operate. As Andrew Warner pointed out many years ago, in A Framework of Efficient Government Investment, it is perfectly possible for an investment to have large positive externalities but also offer private returns that are high enough for private investors to undertake it with no or little public intervention.
Providers of concessional blended finance should only be willing to supply a large subsidy when the resulting impact is high, which will create a correlation between the fiscal cost of blended finance and the size of market failures in a portfolio of transactions, if you think impact is also correlated with the size of the underlying market failures. But large market failures do not imply a large subsidy is necessary. It could be the case, for example, that the positive externalities from green ammonia are enormous but the viability gap financing required to get projects off the ground could be small (if the technology matures from where it is today).
I think the insights in this paper make more sense once we think of a “project” as referring to total investment (consisting of many projects). Suppose the positive externalities from green ammonia are very large but the subsidy required to get a project off the ground is small. Because the positive externalities are very large, it would be socially optimal to have lots of green ammonia projects. So even if the subsidy per project might be disconnected from the size of the market failure, the total amount society ought to spend to increase the quantity of green ammonia projects will be large.
The same is true of diminishing returns. There is no reason to think individual projects exhibit diminishing returns. But if we look across the universe of potential projects, and we want to push the volume of total investment higher, then once we have exhausted all the projects where output can be increased most efficiently, with only a small fiscal cost, we must start selecting less efficient projects. And so on.
The rest of the paper discusses the efficiency of different instruments, subsidised loans and guarantees, in different settings. This is one of the most urgent research topics in blended finance. There is a lot of loose talk about one instrument being more effective than another, without first making sure we are comparing apples to apples by measuring effectiveness after equalising the fiscal cost to the public of each instrument. This section of the paper is also tricky for practitioners to interpret. For example, it is not easy to see who is benefiting from the subsidy – the project sponsor who gets a lower cost of capital so they are willing to borrow more, or the lenders who receives a premium so they are willing to lend more. However, this blog is long enough already, so I will hope to return to those questions another time.
[1] Professor Panizza also argues that the relevant quantity for basing mobilization ratios on is the fiscal expenditure cost (grant equivalent) of the instrument involved not the face value of the instrument. He is quite right. The cost to the public of a $1m grant is much greater than the cost of a $1m low-interest loan or low-fee guarantee. It is frustrating the industry continues to more or less ignore this fact.
[2] That does not mean those questions are never asked, merely that the answer does not affect how the mobilization ratio is calculated. When blended finance is applied according to agreed principles, it will often be the case that the transaction in question would not have happened at all without the public contribution. However, that does not mean the underlying volume of investment would have been zero without that transaction – we don’t know what else might have happened.
[3] It is also complicated by the fact that many blended finance transactions are not stand-alone projects but are funds or other instruments where the effect on the ultimate volume of investment is indirect and hard to attribute, and there is usually some substitution for other forms of finance. For more discussion of that point see: What is holding blended finance back.
[4] Economic theory that includes increasing returns to scale is more challenging because it is harder to pin down when firms stop growing. This is a long-standing issue – here is a book from 1999 compiling papers on the topic. Since the 1970s, trade theory has incorporated the idea that trade might occur to exploit increasing returns.
[5] A reminder – this is not how multipliers are reported in practice. When a DFIs invest $60m in a project and the private sector invest $40 so the reported multiplier is 0.66, that is based on the face value not the fiscal cost of the $60m and $40m is not a measure of how much total investment has increased over a counterfactual. Although Professor Panizza says a blended finance investment with a multiplier below 1 could be justified on welfare grounds, he says it would be ineffective because at that point a straightforward grant would increase project size at a lower cost. I am not sure I understand that argument. In theory a project sponsor should be indifferent between a $10m grant and a concessional loan with a fiscal cost (grant equivalent) of $10m.
[6] See: Beyond leverage ratios: a strategic approach to blended finance
