Development Finance Institutions: Need to rethink proposal?

Q: It was in your time that the development finance institution (DFI) idea was junked and you went towards the universal bank model. Now, do you think that the current banks are incapable of funding infrastructure and we have to go back to DFIs?

Rangarajan: When we moved to the idea of universal banks, the understanding or the implication was that with a lowering of the cash reserve ratio (CRR) and statutory liquidity ratio (SLR), there will be more funds available with the banks and therefore they will be able to meet both the short-term and the long-term credit. Along with it, there was also the assumption that both the equity market and the bond market will expand and that will provide the additional finance required.

However, the situation now, is while the equity market is expanded, the bond market has not. It is not vibrant enough. Second, we have reached a situation as far as the commercial banks are concerned where the flow of credit, both short-term and long-term credit, has choked.

The major problem which everybody is addressing is how to take care of the stock of non-performing assets (NPAs) and what can be done. Therefore, in retrospect perhaps our decision to move away from DFI was perhaps a little premature and we really need to get the long-term credit going because ultimately it is a new investment that will give jobs, that is a new investment that will generate faster growth. Therefore, in this situation, a fresh look at what can be done to provide long-term finance is very much needed.

Q: In your time Reserve Bank of India (RBI) created IDBI as its 100 percent subsidiary and gave it some cheap loans. The IFCI and ICICI also got tax-free bonds. So now also when you have the gen-next wholesale banks or DFIs, do you want some kind of a government guarantee?

Rangarajan: Two things, once the concept is accepted, then we can work through the details. First of all, it need not necessarily be totally fully owned government banks. However, some amount of government participation, I think, at this stage would be preferable. After all, the government itself in the Budget, if you look at it both this time and last time, have been providing for so many funds for infrastructure and so on and so forth. They are better channelled through an institution rather than setting up separate funds. Therefore, to answer your first question, yes, some participation of the government is desirable.

Second, yes, it is an interesting question and that needs to be addressed. In some way the DFIs were able to play the role they did because some element of, not subsidisation, but some element of funds at a lower cost was also provided. Basically we should move towards an idea or towards a situation where there is a mix of funds for these institutions. Some of which raised from the market, but some maybe coming at a concessional rate either through for example as it was done earlier allowing the bonds of these institutions to be treated as part of the SLR. This was one mechanism that was done, SLR.

However, also even now you do look at National Bank for Agriculture and Rural Development (NABARD), the way in which now NABARD is also being funded is that through the infrastructure reconstruction fund where banks pay when they are not able to fulfil the priority sector targets to give the money to NABARD at a concessional rate - that will not work here, that particular principle will not work here, but  the point really is some element of concession will have to be built into it, perhaps they may be asked to also raise funds from abroad, and maybe some institutions like the German institution and so on can provide credit at a concessional rate. Some element of that will have to be built-in.

Q: Finally regulatory issues, will the current regulatory framework suffice or do you think we need a new regulatory framework?

Rangarajan: The regulatory framework that we now have is in relation to banks. The regulatory framework that you might have to frame in relation to the DFI type institutions may have to be somewhat different because the banks take money from innumerable number of depositors and so on and therefore the safety becomes a major issue and therefore we have quite an intricate framework for that.

However, we need to evolve an alternative framework for that, because as I said, we did not have any framework, or we did not have any regulatory framework. We cannot go to that situation; we need to go to a situation in which separate regulatory framework is adopted.

Q: C Rangarajan’s main thrust is that the jump to universal bank was probably premature, that a space is needed for long-term financing with a new institution. What is your sense, do you think the universal bank model will not work for long-term infrastructure lending?

Sriram: To my mind, the concept has two broad issues to address. One is of course that the appraisal skills and the requirements of long-term funding, especially in the infrastructure space, needs very specialised training and analytical skills and that is where the long-term financial institutions probably would have brought in better expertise is one sector of thought which the commercial branches or rather the universal banks have not been able to do so. So, that is one of school of thought in that particular area.

The other area is of course that these institutions have issues with financing cost or charges, both from the lender side and from the borrower side. Their resources are quite high cost and that is one reason why number of DFIs actually went into the universal banking mode where they were getting access to a lot of low cost deposits and thereby making infrastructure funding and even corporate funding a little bit much cheaper than what it would have entailed. If you see the interest rates that happened in those days for DFIs, it was pretty high even after the government support by way of credit enhancement or tax-free bonds and so on and so forth. However, even after that, you would find that the cost of financing is pretty high.

Also, if you add to that, there is hardly any income to the bank other than the transaction that happens whether in terms of the fund based or non-fund based limits that they do. If you talk about other areas where today the universal banks get incomes from corporate or from infrastructure by way of selling cash management products, insurance - both general and life, salary accounts, retail business, and so on and so forth, all of that adds to the income from a particular relationship thereby making the cost of the core funding becoming much cheaper than otherwise it entailed. So, these are two broad areas which I feel that needs to be discussed.

Q: These are two different issues that B Sriram is raising. You were there when ICICI got reverse merged into ICICI Bank. The first question is that is the universal bank at the moment unable to fund long-term needs, should the answer lie in restricting long-term finance to only say two or three universal banks, maybe SBI, ICICI, variety. Would that be the answer or would it be an altogether new experiment with DFI?

Gambhir: I think either the historical perspective of merger of ICICI with ICICI Bank or for that matter the three development institutions that we had at that point in time, the context was quite different. The context was that the banking sector was quite fragmented and at that point in time there was a vast funding difference between where ICICI could raise funding versus where, let us say, SBI could raise its liabilities. In the context of domestic banks becoming more active in corporate lending and not just in let us say working capital financing, but also in term financing. The financing cost of the development financial institutions was quite high and they found it very hard to compete.

So, that to my mind is a very important issue when you think about a long-term financing institution in the context of Indian economy where most of the liabilities of the banking system are actually pretty short-term. Where is the long-term liability going to come from, that was the issue then, and that is the issue that probably is going to be discussed even now when we think about these long-term financing banks.

To my mind, the concept of a specialised wholesale institution is a very interesting concept and it needs to be thought about, it needs to be dealt into further, but a copy-paste of the earlier 1960s, 1980s DFI model into today’s time is probably a wrong thing to do. I think what we need to think through is a model which is a hybrid model between an investment bank and a commercial bank where a bank which is let us say a wholesale long-term financing bank is not only just doing a classical borrowing and lending transactions, but is also developing the market, the capital market along with it by doing stuff like securitisation of these assets, different forms of credit enhancement, raising financing through the cross-border mechanisms.

So, the suite of products that this kind of a bank that needs to today deal with to effectively manage its liabilities and effectively manage its asset side risk is entirely different and that is something that needs to be developed and probably development of that suite of products in the context of a very universal bank is a little bit more difficult than in a specialised institution which is more dedicated towards capital markets, more dedicated towards cross-border activity, etc.

Q: I think we are confusing issues both between commercial bank and long-term finance. Give me first a monosyllabic answer – do you think that the current commercial bank cannot do long-term financing or are you saying that some commercial banks cannot, like for instance HDFC Bank has chosen not to and RBL Bank perhaps will choose not to, but some people like SBI and ICICI could continue to have the skills. So, is the universal bank model wrong or is that some of them should do but some of them should not?

Gambhir: I think the universal bank model is not wrong. I don’t think the commercial banks cannot do it, but I think that in every marketplace there is a need for specialised institutions for a particular activity and that kind of a specialisation sometimes is hard to develop in a more generic universal commercial bank and hence -- otherwise you would not have investment banks in the first place.

So, the fact that you have these kinds of institutions exist means that there is a scope, there is a place for more specialised institutions which understand some of the dynamics of these kinds of financings and these kinds of market activities a lot better than probably a generic commercial bank would understand. Now, that is not to say that SBI and ICICI cannot do it, I am sure they have the capability, but probably you cannot have 26 or 40 of these banks doing the same thing.

Q: Let us assume that we have to try and work a long-term financial institutional model. How would that institution become viable?

Shah: I think what RBI calls wholesale finance is actually looking at more corporate accounts and all which will, I guess, have to have both the long-term assets and the short-term assets also. I would think that a lot of these institutions will end up playing in the bond market, will also be part of the primary dealership and all. I think it is viable and this is all part of RBI, who is trying to make the Indian financial system more heterogeneous because currently it is very homogenous. So, I think that objective is obviously achievable.

As we have already spoken, the key challenge is going to be on the liability front and what we have seen with wholesale funded institutions having long-term sticky liability is always going to be some kind of a challenge unless we have very robust bond market. So, I think the way this will work is, most of the long-term banks or the wholesale banks will be aligned to the current universal banks and it will be some kind of an alignment so that you can get the asset side expertise on one side, but have the stability of the liability through your association with a universal bank.

Q: If a long-term finance institution is looking only for bond market money, today its cost of finance has to be northwards of at least 7 percent or 7.5 percent. So, what C Rangarajan was saying was that that you have to give some kind of government aid to this institution so that at least some of its money is cheap, some foreign money, some SLR status, or some tax free bonds. You think that is the way out?

Sriram: Partly yes, there has to be something in it for these banks. For these banks to become viable, to my mind, there are two broad reasons. One is that on a standalone basis, the cost of resources is going to be very important for these banks and second is of course that they were going to be anyway competing with the universal banks which are not going to give up this space. In that context, the pricing power of specifically the better rated corporate is going to again face a challenge there.

So, the point here is that on a standalone basis, if the cost of resources is going to be high and I would imagine that without a credit enhancement, or without any sort of -- even today if you see the cost of resources of an IFCI for example or IIFCL, it still requires a lot of working to cost cheaper for the asset, for the person who is going to take a loan. To that extent, for a person who starts a bank and then tries to raise resources, the resources can come only from two-three sources. One is of course the public deposits from the bonds, from maybe external commercial ECAs or from sister banks which are there.

Now all of them are surely going to take upwards of 8 percent in terms of resources and if you add the CRR cost to it, it is definitely and with the margins and the sort of risk that is taken by these banks and the infrastructure lending especially prior to COD. I think the viability aspect as I said earlier, it is very important today that the financial cost, both for the person who is borrowing and for the person who is lending, have to be pretty good. There are problems in terms of project completion. Some sort of government help could be needed for such banks initially.

Q: How would you design this institution in terms of liabilities?

Gambhir: I think there are two parts to the liability cost. One is the direct cost that you pay to your lenders which is the bond holder or the depositor. The second is the regulatory overburden that comes on to this liability. For example, CRR, SLR, priority sector lending, these are all things that add to the cost of the liability.

Now, is it possible for us to design an institution given the fact that this institution is likely to be funded largely from bond market or funded largely from long-term sources, can you actually minimise the burden of let us say CRR and SLR on these banks, can you completely do away with the need to do priority sector lending given the fact infrastructure is one of the big focus.

Q: Then it will be viable?

Gambhir: I think that can take away something close to 50-75 basis points of course, and also bear in mind that by design these institutions are likely to be light weight in terms of their physical infrastructure. They are not supposed to have thousands of branches all over the country. There will be few institutions, lesser number of people, lesser overhead in terms of manpower. So, I think it is possible to come up with a financing structure where the cost of liabilities -- it may not be as competitive as let us say SBI's cost of financing is, or cost of liabilities is, but it can be pretty reasonable and also, bear in mind that most of the infrastructure financing is not AAA rated. It is somewhere close to BBB minus. They are all project risk, so, you need to charge for the project risk. So, it is possible that you can design a liability base which is more suited towards infrastructure financing, more suited towards long-term financing, but then carries minimal amount of regulatory overhang in terms of cost.

Q: Wasn’t the old IDBI and ICICI also free of this level of CRR and SLR responsibilities? What is your sense, the costing only borrowing from the bond market will not work, what is your final winding up comment on how to make this work and more importantly won’t this face concentration risk?

Shah: Obviously some innovation on both the asset side and the liability side. My understanding is that the wholesale banks will not be subjected to SLR, CRR and the priority sector. So, in a way, they will have that small advantage of not being burdened with that. I think they will be able to borrow in Indian market, international markets; they will not be as highly geared as banks I think because they will have a much higher chunk of equity their overall cost of funds will be lower. In the sense, their NIMs should be protected.

I think the larger thing to make the business model viable will have to be work on the asset side and I do believe that if these institutions are conceived as only project financing institutions, then it will be hard to make it work because the asset quality, the inherent risk and the fact that on that they will still compete with the existing universal banks, will make it highly unviable. However, if they do a lot of other things like structured credit, if they do some other M&A financing, and other wholesale activities also, then they will have enough innovation possibilities on the asset side to make the yields really work for them because they need to compete in a few areas that currently banks are not able to compete because of either capital market exposure or other reasons and allow them to also provide another specialised form of financing for the corporate in India.