How An Accounting Change At The RBI Helped The Government
Earlier this week, the Reserve Bank of India agreed to transfer a Rs 1.76 lakh crore surplus to the Government of India for 2018-19. As part of this, a transfer of Rs 52,637 crore stemmed from the Bimal Jalan Committee recommendations, which suggested a reduced level of contingency reserves for the central bank.
The transfer, included another surprise element — a change in the way the Reserve Bank accounts for the gains it makes on foreign exchange interventions it undertakes to keep currency volatility in check.
“During the year, exchange gain/loss has been computed using weighted average cost method resulting in an impact of Rs 21,464 crore,” said the RBI.
The accounting change, however, will have a far more lasting impact, explained Ananth Narayan, associate professor of finance at SP Jain Institute of Management and Research.
He explained the implications of the change in this conversation with BloombergQuint. Edited excerpts below:
At first glance, what did you make of the balance sheet, what did you make of the income increase this year?
There are two components which explained the increase in this year’s RBI surplus. One is the fact that they haven’t taken anything into the contingency funds. Last year they had set aside Rs 14,000 crore for the contingency funds. This year, because of the nature of recommendations of Bimal Jalan Committee, there is no need to increase the contingency fund. In fact, they’ve decreased it. I think the second element is clearly that last fiscal year, there were a lot of open market purchases of bonds conducted by the RBI -- Rs 3.5 lakh crore. Clearly, the interest that accrues as a part of that, against those bond holdings goes back to the government as revenues. So that’s another big chunk, I am estimating about Rs 20,000 crore.
The big change in the accounting sense is on the foreign currency accounting.
The way by which RBI had been accounting for its foreign exchange operations has been anomalous for the last many many years. Let me explain this.
When the RBI buys dollars as part of its intervention and reserve management processes, it buys it at various rates. Historically it has been buying it at Rs 40, Rs 50, Rs 60 so on and so forth. Now, normally, when the valuation of these forex reserves changes, obviously as the dollar-rupee goes up, the rupee value of this particular reserve goes up. But that is not taken into the profit and loss account, but directly into the balance sheet as part of the currency and gold evaluation reserves. So, it is not recognised as an income but as an accounting entry.
As of last year, that number was pretty large Rs 6.9 lakh crore. And here’s where the anomaly is. That Rs 6.9 lakh crore also had an element of recognised income, not just an notional income. Let me explain this. Now, RBI not just purchases dollars but also sells dollars as a part of its forex intervention. Now, you would expect that it purchases dollars at Rs 40, 50, 60 and is today selling it at Rs 70-72, there should be some amount of recognised profit and loss impact from that transaction.
But the RBI has not done that historically. It has always valued the sale of dollars against that week’s revaluation rate, which is obviously very close typically to where the RBI sold those dollars. In this sense, this actual revaluation reserves, the way in which it is accounted for, does not allow the RBI to ever recognise P&L gains.
What the RBI seems to have from this year onwards, is gone to a more conventional accounting method where it has said -- ‘listen, I am holding $400 billion of currency reserves. The average holding of this cost is, say, Rs 55 to the dollar. If I today sell the dollar at say Rs 72, for every dollar that I sell, I should recognise Rs 17 of profit. That’s a big change that has happened this year.
My own estimate is that, because of the change happening in this financial year, the RBI seems to have recognised up to Rs 40,000 crore of profit from the sale of dollars.
Remember, from July to end of the year, last year, RBI was selling dollars considerably in order to reduce the volatility in the currency market. So, every sale of dollars that they undertake, would have helped them recognise some of the currency P&L and that I estimate is, Rs 40,000 crore.
To summarize, the actual sale of dollars never really got them substantial profits, because they were valuing the sale of dollars against that week’s revaluation rate. Now, they’ve moved to an inventory kind of evaluation, where they’ve looked at the weighted average cost at which they purchased the dollars overtime and every sale seems as a recognition of profit when compared to the weighted average rate.
That’s a big change, it has got a huge historical element, its not a profit that has come in today. It’s profit, which in a sense, was kept away for many many years and is now being recognised.
In the actual annual report, they give a number, They say that during the year, exchange, gain/loss has been computed using the weighted average cost method, resulting in an impact of Rs 21,464 crore. So that would be what they have made because of their forex intervention operations, compared to the weighted average costs, is what you’re saying?
I’ll have to reconcile this Rs 21,000-crore number but, the way I see it, the actual currency and gold revaluation account has declined from about Rs 6.9 lakh crore as of last year to about Rs 6.6-odd lakh crore this year. So, Rs 27,000 crore declined in the currency revaluation account at a time when rupee actually marginally depreciated, and therefore the currency and gold revaluation account should’ve gone up. So, if I look at the net change to what I would have expected, the CGRA or the currency and gold revaluation account, would have gone up by about Rs 40,000 crore.
But maybe, RBI is right, maybe it is Rs 21,000 crore. Either which way, the sale of dollars this year was compared versus the weighted average holding rate at which they purchased the dollars overtime. That has been recognised this year as profit. This is a huge departure from what has been followed so far where practically, very little profit, if any, would’ve been recognised against any sale.
Is this change kosher? Bankers say this is the way its done in the commercial world and the YH Malegam committee of 2013 had recommended it too.
I think so. So, lets break it up into two parts. I do think that recognising a P&L profit when you sell the dollars is correct. Look, RBI holds both the dollar assets as well as the rupee assets. On the rupee assets, it earns about 7 percent every year. On dollar asset, it earns only about 2 percent. Now, that’s not to suggest that the dollar asset is less valuable than the rupee asset. What one expects that overtime, that rupee would depreciate and that’s how therefore the value of dollar that you earn,even at a lower interest rate would equate to what you would earn in rupees.Otherwise the market would not be where it is.
So, overtime, historically the RBI has only been recognising that 1-2 percent of dollar interest income, it has not been recognising the fact that the value of dollar has itself gone up. So, actually recognising P&L gains is not a bad idea. It is a historical anomaly as I mentioned earlier which is being corrected now.
Where I do have a problem is that this anomaly has been built up overtime. If the weighted average holding rate for argument sake, I could be wrong, but I think it is 55 rupees per dollar. This is not the 55 rupees rate at which the RBI has accumulated the reserves this year. It has been done over history, over many many years and decades. So, in a sense, you’re correcting for this historical anomaly and paying out a lot of money today.
It is very similar to the debate we had with respect to contingency funds which has been built up over time and now giving out a big cheque today. I think it does merit a debate on whether it is right to correct the historical anomaly at one shot. That’s one aspect.
The second aspect is, this can be a tricky kind of a situation and commercial bankers will vouch for this. In a sense, if the RBI was silly and if the RBI was profligate and the government was profligate, this can becomes a tool for the government to generate extra P&L gains if and when required. Let me explain. Every sale would recognise a P&L gain from the weighted average of, let’s say, 55 to where it sold, say 72. Every purchase of dollars would add to the weighted average of costs or of the actual dollars being held. So, if during the course of the year, RBI bought $50 billion and sold $ 50 billion, on the $50 billion that is sold, it would recognise an enormous amount of profit. On the $ 50 billion that it purchases, the very similar rates, it would not recognise gain or loss. It would simply add to the weighted average cost of the particular reserves being held by the RBI.
I am sure the RBI has thought of checks and balances to ensure it doesn’t happen but you could have a situation where a weaker rupee helps recognise more P&L gains and the RBI starts to recognise those higher rates. Second, even buying and selling with a net impact of zero can help the RBI recognise P&L.
Now, I must add that the RBI is far most prudent. So, I am sure, there are checks and balances to avoid all of this. But this does get into a kind of a slippery slope where P&L gains can be recognised out of this large currency and gold evaluation account, if the government so chooses this to do so.
Watch the full conversation here: