Apr 02, 2016 01:18 PM IST | Source: CNBC-TV18

Permanent liquidity key for solving banks' cash deficit woes

The RBI will need to figure out the best time to either ease or tighten liquidity and decide on the right tools for injecting more money into banks, say Hitendra Dave, Head of Global Banking & Markets for India at HSBC and Neeraj Gambhir, Managing Director & Head-Fixed Income at Nomura India

The Reserve Bank of India (RBI) is scheduled to announce its first bi-monthly monetary policy for 2016-17 on 5 April. While stock markets are hoping for a 50 basis points (bps) rate cut, bond markets expect easier liquidity.

The key reason why liquidity is a talking point is that since January, the liquidity deficit with banks has been as high as Rs 1.5 lakh crore to Rs 1.8 lakh crore versus the RBI's stated policy that the deficit should be around Rs 80,000 crore. Post advance tax payments on March 15, this deficit has gone up to Rs 2.5 lakh crore.

While the RBI says it is sorting out the deficit liquidity issue through fixed rate repurchase agreements (repos) and numerous variable repos, bank treasurers say banks will chase deposits to get cash. 

In order to restore liquidity in banks, three issues will have to be addressed, namely, 1) how should the RBI go about reducing the cash deficit 2) How much should the deficit be and finally 3) should the RBI keep the banks in deficit when it is cutting rates?

Below is the verbatim transcript of Hitendra Dave and Neeraj Gambhir's interview with Latha Venkatesh on CNBC-TV18.

Q: Do you think the RBI has got its liquidity management all wrong?

Dave: There are two parts to it. One is the generic thought around what is an appropriate liquidity condition, whether it is a surplus, whether it is a deficit in the context of a given monetary policy situation, whether you are tightening, whether you are being accommodative etc. So, I think, that is one part of the debate and I am very pleased to note that the RBI after the last policy mentioned that they are undertaking a review of what is an appropriate level of liquidity because that has been a matter of significant debate in the market place about whether we should be 125 bps lower on the policy rate side but in terms of liquidity, we are more or less unchanged and as this discussion will show possibly we are a bit tighter.

The other part around the numbers that you mentioned from March of last year to March of this year, the loss of permanent liquidity on account of increase in the currency in circulation is roughly of the proportion of Rs 200,000 crore and there is another about Rs 40,000 crore higher, which the banks are maintaining by way of higher balances with the RBI. So that is permanent liquidity loss of about Rs 225,000-250,000 crore.

Against this, it is a simple set of how much has been added by way of permanent injections. We do know that RBI has only three tools of permanent injection, one is cash reserve ratio (CRR) which they currently don’t have a preference to experiment with.

Second is increase in their foreign exchange (Fx) assets.

Third is increase in their rupee assets and at least the numbers that we can figure out would suggest that they have added about Rs 56,000 crore by way of increase in rupee assets, which is the open market operations (OMOs) and the similar number by way of increasing the assets which is about USD 9-10 billion of purchases. So that leaves the gap of between Rs 70,000 crore and Rs 100,000 crore and the liquidity adjustment facility (LAF) number has gone that much higher.

Q: Would you broadly agree with these numbers that RBI's addition of liquidity has been less than the loss of liquidity because of Fx and other reasons?

Gambhir: The keyword here is permanent liquidity. This is a tautological statement that if the banking system is short of the central bank's primary liquidity then RBI has to provide that liquidity otherwise the banking system will be defaulting on its CRR requirements, which RBI does by way of its term repo operations because these operations are temporary, the system has to go back again to RBI for borrowing this money, which creates the sense of a lack of liquidity in the system. So, I think the keyword here is how much permanent liquidity that you are creating in the system or if not permanent, at least more enduring liquidity you are creating in the system and how fast are you creating that liquidity.

These calculations can be done at the beginning of the year with a reasonable degree of accuracy, what is the amount of permanent liquidity creation that the system will require but the question is whether you do it linearly through the year or you wait until a very late time in the year and then you create that liquidity. That is the key issue here in my mind.

Q: Normally, forex flows are healthy in January-March but this year probably because of the Chinese yuan devaluation and the scare about emerging markets, there was an outflow of capital. The RBI could not have done the OMOs earlier and if it did do the OMOs now, it runs the risk of further rise in M3 and therefore endangering its inflation position. Is that a fair argument?

Gambhir: I would say that any liquidity management of this nature requires -- given the institutional set-up that we have in India -- a considerable amount of judgement on behalf of the RBI to see when that permanent liquidity injection is required. What is the most appropriate time for it and if there are any changes in the real world as against that judgement then obviously it leads to either tightness of the liquidity or excess liquidity.

What I would say is that if you are in an accommodative monetary policy stance then the preference should be to give system more liquidity rather than less liquidity. So that even if you err a little bit in your calculations, if the system is surplus then the ability of the system to sort of carry forward the accommodative monetary policy stance into the rates of economy is that much more higher. That is the difference here. The point is that the RBI has been waiting a little longer in terms of providing that liquidity than probably the accommodative monetary policy stance would have warranted.

Q: I assume that this is your argument as well -- it is not a question of how deficit it is, the point is the RBI should never be in a deficit mode that we should be operating at the reverse repo window, at a time when the RBI is in accommodative stance, is that your point?

Dave: I think there are schools of thought whether you can have surplus liquidity without the operating rate becoming the reverse repo rate and now the RBI has various tools. Earlier on three-four years back the RBI did not have tools. So either it was reverse repo rate or repo rate or MSF rate, now they do have this thing called a variable rate repo and variable rate reverse repos.

So as I mentioned earlier, once you are relatively clear that the stapermate of the monetary cycle merits a much easier liquidity condition as in defined as not a deficit of 0.50 to 1 percent but a surplus of half a percent but you don’t want that half a percent to go to reverse repo rate for whatever reasons. That is a point of a separate discussion.

To come back to the point you were making earlier, I think at least looking at the data that I have gone through it does appear that there was a certain degree of assumption that there would be an increase in the forex assets because that is how it had been working for most of 2014 and first half of 2015. That switched off from the original Chinese devaluation date of August and that is what is causing this sudden widening of the gap between the total borrowing of the system from the RBI, which is around Rs 2,30,000-2,40,000 crore and the government balance, which is around Rs 160,000 crore. This differential to me is simply the loss of permanent liquidity over the last 12 months or so.

Q: Will it not be a problem in terms of the RBI endangering the system with too much of liquidity? If it were to bring the system to a square position, that is banks are borrowing maximum Rs 5,000 crore net or lending the system Rs 5,000 crore then you will have to provide liquidity to the extent of Rs 1 lakh crore. The Urjit Patel report indicates very clearly that when the system was printing notes or doing OMOs with greater regularity as it did in the period from 2010 to 2012, we ended up having double digit inflation, so while the impact may not be immediate, we may pay the price in the coming year, is not that a fair argument?

Dave: I cannot agree with you on this point. I think we are talking about a system, which will borrow when the government\\'s balance is zero about Rs 80,000-90,000 crore, which will borrow Rs 30,000-40,000 crore when RBI is at the peak of its Ways and Means Advances (WMA) drawdown -- the 50,000 limit which has been set two days back or so.

So, we are far from our system which is going to be flooded with liquidity. Flooded with liquidity is if there is so much money in the interbank system and there are not enough assets and we are giving it back at the reserve repo rate, if the money goes back to the central bank counters everyday through a variable rate reverse repo operation, I am not able to understand at least how that suddenly -- that in a banking system of Rs 90 lakh crore if you have Rs 40,000 crore which is less than half a percent or something which is given back to RBI, how that tail will wag the inflation dog so much. I am not entirely aligned to that school of thought.


Q: You were saying that we are very far from a surplus liquidity situation. My point is that if the RBI were to go ahead and do OMOs to the extent of the deficit, which is 1.5 lakh crore, by the time it works its way into the system, it may be very difficult to control the M3 at a later stage. Your point.

Dave: The point I was making was that -- I don’t think it is Rs 100,000 crore. I think as a starting point, let us agree what is an appropriate level of liquidity. Is it the system not having any surplus and not having any deficit? If that be the case, it appears that currently the RBI needs to do anywhere around Rs 40,000-50,000 crore more of OMOs but you have to be slightly forward looking in these aspects.

So over the next 12 months, the system is once again -- this is unique to a country like India where the system loses something like Rs 200,000-250,000 crore by way of increasing the currency in circulation. That means if you do this Rs 50,000-60,000 crore, which I am talking about today and don’t do anything then next March when government balances go once again to Rs 150,000 crore, you are looking at total dependence of RBI of almost Rs 400,000 crore.

Whether we like it or not, that is a reality. So over the next 12 months, after doing this Rs 50,000-60,000 crore the RBI needs to do another Rs 200,000-250,000 crore. It has to choose whether it is comfortable with an assumption that it will be able to buy USD 30-40 billion or whether it wants to split that a little bit with increase in INR assets or it might even have to explore the CRR tool at some point of time because I agree with you, you just cannot keep buying bonds endlessly just to arrive at this balance.

Q: The problem is that RBI has three imponderables, which it cannot predict. Forex flows, government balances and currency leakage and currency leakage also is now becoming increasingly unpredictable, we are getting into an election season so we don’t know how much of currency leakage will happen. There is this new committee, which the government is going to appoint to look at government balances, one would assume that it will also have RBI officials in that committee. Should it be that every time the deficit in the system goes beyond a point, beyond that 40,000-50,000 mandatory one percent of deposits then the RBI should cancel the T-bill auctions?

Gambhir: Yes, you do need a certain amount of dynamic cash management by the government. Especially now that the government is trying to collect cash proactively for its next year's liabilities and making sure that it has enough cash going into the next year. Both treasury bills as well as the cash balances of the government are meant for making sure that they can manage their payments smoothly. So you can play between the outstanding cash balance that the government has, the amount of outstanding treasury bills that they have, the overall issuance of treasury bills to some extent.

I am not sure you can deal with the entire Rs 1,60,000-1,70,000 crore problem by just managing the treasury bills. The challenge is that a system which needs about Rs 4 lakh crore of RBI balance as CRR, the moment it starts going into a more than Rs 1-1.5 lakh crore of dependence on RBI for overnight liquidity or for daily liquidity, it starts having its own problems which then need to be addressed through a large amount of injection via the RBI. So we need a little bit more management by the government of its cash balances than just doing -- treasury bills obviously are required but it does require a little bit more than that.

Q: What would be your message or request to this new committee that is going to be formed?

Dave: What the RBI is doing currently is quite okay, whenever they find that the government has taken money away from the banking system of Rs 50,000 crore, they announce Rs 50,000 crore of repos. So what the system is losing is uncollateralised deposits and replacing it with collateralised borrowings and presumably, I think the challenge has become because the government in the last four-five months has run consistent balances, these so called frictional or temporary outflows have acquired a semi-permanent kind of characteristic.

However, I think from this committee, because even if they were to auction it out, we would have to bid for it etc, I think it would be very helpful to study on seasonality so that the system can prepare in advance, which is already in place but that can be made more and more updated in terms of information regarding government balances and likely significant chunky outflows and inflows which we do not know of because we only can see bond market related flows etc.

I would make the point that the challenge for the system right now is not the frictional tightness. The frictional tightness can be very well managed with the way RBI currently provides repos at very short notice, there is no issue there. I think the bigger issue is around how to manage this persistence of outflow, which is acquiring a rate of around Rs 15,000-20,000 crore a month, of late it is like Rs 20,000 crore a month. So, you have done OMOs of Rs 30,000 crore in the last two weeks or so but that has been matched by currency outflow and therefore you are back to the same level of deficit as you were earlier, which is the challenge. I think that part needs to be addressed, this part I am quite comfortable with.

Gambhir: I just wanted to add to what Hitendra Dave is saying. My wishlist has been for a very long time that we need to take the government cash balances out of this equation. I think the core thesis, the core problem here is that we have something which is basically a treasury management, which is getting mixed up with a more strategic monetary policy operation. I think we need to separate the two. That to my mind is the real long-term solution here to start off with but I am not sure if we can implement that at this point in time, maybe it requires some amendments to the RBI act. So that government can start directly dealing with the banks rather than going through RBI's balancesheet. That is a very important aspect.

The second important aspect that needs to be addressed is the question that Hitendra started with which is that what should be the liquidity surplus or deficit in different cycles of monetary policy. Should you run both easing and tightening policy with deficit stance, which is the case currently or should you run easing policy with a surplus stance and a tightness policy with a deficit stance? That also requires a little bit of a debate and clarity in my mind.

Follow us on
Available On