Infrastructure DFI must be a fully public entity

This post will try to explain why DFIs that have private investors will not be able to achieve national objectives on infrastructure financing.

I have written here explaining the problems with infrastructure finance, categorised the infrastructure sector into those where projects need to be de-risked to attract commercial investors, and how DFIs can perform the de-risking role.

Governments across the world have experimented with various DFI structures to de-risk and crowd in private capital to finance infrastructure projects. All the Western DFIs are fully government-owned entities which use different concessional financing instruments (concessional or subordinate debt, guarantees and credit enhancements, and equity) to crowd-in private capital into individual projects. In simple terms, while the project entities that the DFI invests or provides finance are privately owned, the DFI itself is fully government-owned. The DFI seeks to crowd-in private capital at the project level. All of them address the risks of capture and abuse arising from government ownership by having good governance systems.

India has taken a different path, seeking to crowd in private capital at the DFI level itself. This has been a consistent endeavour from the IDFC in the nineties to the NIIF. The theory is that the government will supplement every rupee that the private sector brings in, thereby making both go longer, besides expanding the pool of private capital available for infrastructure finance.

Let’s unpack this model. The private (or non-government) institutional investors in the DFI’s Asset Management Company (AMC) want to maximise returns from the investments made in the portfolio projects. This introduces a natural selection bias towards the nature of projects that the DFI will want to fund.

The government instead seeks to maximise its additionality - de-risk and expand the envelope of projects that are amenable for commercial funding. It wants the DFI to invest in sectors like urban water supply and sewerage, solid waste management, electricity distribution, urban mass transit, energy saving projects (ESCO), health and education PPPs, etc., where private investors are deterred by the commercial unviability of the projects.

There’s an inherent and irreconcilable conflict between these two objectives. Since money is fungible and corporate shareholders cannot have such conflicting goals, this model will struggle to satisfy either side. I cannot see how an AMC that brings together both private and public shareholders can meet the government’s objectives.

In the battle of the conflicting objectives, the private capital’s interests will invariably prevail. Or else, they will exit the entity. There’s no scenario under which the government’s additionality objective can be realised.

It can be argued that even without providing concessional capital, the DFI is de-risking projects by having the government as a co-investor. This would have been the case with infrastructure investments in India a decade and a half back and would be so even today in several countries. But in the current Indian context, such de-risking is minimal. This can be easily verified by examining the portfolio (and pipeline) of NIIF and scrutinising them for the additionality of scarce public finance in its portfolio of investments. It’ll reveal that the DFI has competed with other private investors and crowded out private capital.

It’s possible to pretend otherwise and kick the can down the road by rationalising the DFI’s current investments as arising from a need to build a portfolio of commercially viable investments to establish its track record. This rationalisation glosses over the fact that no track record can incentivise the entity to invest in risky projects of the kind that the government has in mind.

The only justification for government funds flowing into an institution with such a corporate structure is that of a Sovereign Wealth Fund, where the objective is to deploy the government’s reserves and surpluses and earn returns higher than from traditional risk-free Treasury Bond investments.

This raises the question of why India has persisted with structuring DFI’s in this corporate form that’s different from the practice elsewhere.

The primary reasons for structuring DFIs without majority government ownership and in PPP mode are to have an arms-length relationship with the government, to hire professional investment managers from the market, and to avoid the scrutiny of public oversight agencies like the Comptroller and Accountant General (CAG) and the Central Vigilance Commission (CVC). The latter two in particular have been critical determinants. Unfortunately, this reasoning introduces a critical design flaw in the corporate entity's commercial incentives, making it impossible to meet the government’s fundamental infrastructure finance objective.

It’s, therefore, hard not to argue that such entities are a very inefficient use of scarce government financing. It provides free public capital to private investors with little in return for the government (apart from returns if the objective is to be a SWF). In the nineties and early noughties, the IDFC was found free-riding on public funds and lost the trust of the government. It’s difficult to foresee anything different with NIIF.

The only alternative would be to embrace the tried and tested models of fully government-owned DFIs that have the flexibility to deploy a basket of differentiated instruments to de-risk and crowd in private capital. As I have blogged earlier, the government should channel all its different kinds of infrastructure finance initiatives, like viability gap funding, interest subvention schemes, capex support schemes etc., through this DFI. It should become the single point for all government initiatives that involve attracting private capital into infrastructure projects and should work with local, state and central government entities to achieve financial closure on their projects of such kind.

Yes, this entity will be constrained by public oversight and will not be able to pay premium salaries to its investment managers. But it’s possible to meet the objectives even with these constraints if there are good due-diligence protocols and governance systems, and they are complied with seriously. That has been the practice globally.