DFI Is not the Solution for Infrastructure Projects

There have been suggestions, lately, of the need for one or more development financial institutions (DFIs) in India. This is expected to boost the financing of infrastructure projects in India.  However, the argument misses the root problem. Let us see why. 

Financing can be categorised into various types on several dimensions. It can be long term or short term, equity or debt, public or private, and so on. However, the underlying theme for any kind of financing is always that the financier bears some risk and expects a commensurate return on that risk. For any kind of financing, the financier must efficiently assess the risk and compute the appropriate price for that risk. 

Various financiers have differing levels of ability, capacity, or motivation to assess and price the risk. It is this difference between the financiers that leads to sub-optimal outcomes for various parties. If this simple framework is understood and adopted, a lot of our problems can be solved. 

Let us talk more specifically about the lack of financing for long term infrastructure projects. Why does a bank or a non-bank finance company (NBFC) (we call them lenders here onwards), today, not finance infrastructure projects?

ALM Mismatch 

One of the key cited reasons is the issue of asset liability management (ALM). In India, most lenders borrow funds with maturity under 5 years. The average maturity would be about 2 years for a typical lender. The reason is primarily the absence of a deep bond market to borrow from. Other than insurance companies and pension funds, the other financial institutions do not have access to long term funds. As a result, they lend to a project with a maturity of, say 20 years, with funds of 2-year maturity. This leads to a mismatch in the maturities of assets and liabilities for the lender, commonly understood as an ALM mismatch. After 2 years, when the existing liability matures, it must be refinanced afresh because the asset has not yet matured. The lender, therefore, suffers a re-financing risk on its liabilities in such a case.

But the ALM mismatch, if understood properly, can be managed by a lender. Let us see how this works from a lender's perspective.

Ideally a 20-year project should be financed by 20 yr money to avoid an ALM mismatch. But suppose a lender funds a 20-year loan with 2-year money. It can keep refinancing its 2-year liability till the 20-year loan matures. The problem occurs if the lender is not able to do so for various reasons. So, how does a lender mitigate this risk?

There are two key principles that every well-meaning lender must follow. Firstly, as is well known, it must develop strong credit underwriting and monitoring capability in the segment in which it lends. 

Liquidity Risk 

The second principle is not as well appreciated. It must prudently manage its liquidity risk by maintaining adequate amounts of liquid assets to deal with any risk of delay in refinancing its liability. It should be willing to go beyond the regulatory requirement. But this is not easy. Managing liquidity risk requires special expertise.  It is also perceived to be opposed to the short-term incentive of a lender, which is to maximise its profits. The inability to actively manage liquidity risk has proven to be fatal for many lenders in the last few years.

DFI as a solution? 

Thus, it is not the issue of ALM, per se, but its poor management that leads to trouble for the lenders. One way out for a lender is to not take any ALM risk. This is the equivalent of a DFI style of lending. Borrow 20-year funds and lend it for 20 years to a project. But there are two questions.

A) Where would the DFIs get 20-year money? There is no long-term bond market in India. The long-term Infrastructure bonds issued by the likes of REC and NABARD are implicitly backed by government guarantee. If this model is adopted, the DFIs would also be effectively lending on behalf of the government. This has, in the past, led to adverse outcomes linked to moral hazard. We have moved away from government backed financing of infrastructure projects. There are issues to be fixed in private sector financing but going back to government funding is not desirable. 

B) If the DFIs operate with no ALM mismatch, their cost of borrowing will rise. This will affect the shareholders (through lower profits), the projects (through higher cost), or both. Since DFIs are expected to fund projects on a large scale, this issue can be significant. 

Another reason cited in support of DFIs is that the existing lenders are currently not willing to lend to infrastructure projects. That is true, but the real underlying reason needs to be understood and addressed.

Distorted Lending System

While talking of ALM risk, we mostly think of the maturity mismatch. But the real (as against contractual) maturity of a loan is heavily influenced by the movement in interest rates. Most infrastructure projects have clauses for interest rate reset, typically at an annual frequency. If the interest rate in the market falls after a project has been financed, the borrower receives the benefit of the fall in rate. But if the rate goes up, lenders have often found it difficult to pass on a rate hike to the borrower, if it is performing well. This is because our banking system lacks atomistic competition and is dominated by a few large banks. 

The public sector banks (PSBs) as well as privately owned lenders have systematically under-priced the credit risk for infrastructure projects. The PSBs have under-priced mainly on account of misaligned incentives and weak credit assessment capability. As a result, the private lenders have also had to under-price risk on account of competitive pressure from the PSBs. 

The negative effect of the aforesaid mispricing of risk is not visible when the economy and the loan book keeps growing fast. A rising tide lifts all boats. This was the case during the period 2004-12. However, the lenders become exposed to the accumulated credit losses whenever the loan book growth slows down, as happened in 2012 onwards. The result has been that a huge amount of infrastructure loans has turned into non-performing assets. 

Summary

The current situation is that private lenders are not keen to underwrite the risks that are not priced in fully. They are also in a risk averse mode. That will not change unless the fundamental reasons like sanctity of contracts, time bound recovery of dues, and overall ease of business are credibly addressed. 

To summarise, the setting up of DFIs is neither necessary nor desirable to solve our problem of project finance. It will effectively put the onus of financing in the court of the government. The real problems of our financing sector are related to governance and risk management at macro as well as micro levels. The government needs to strengthen the incentive framework and the lenders need to improve on their risk management capabilities. These need to be recognised and then addressed for sustainable benefits.