ADVERTISEMENT

Don’t Double-Dip Into RBI’s Books

A standalone RBI dividend payment, without a government bond buyback, is tantamount to monetising the deficit, plain and simple.

A man throws a fishing net into an inlet from the Arabian Sea. (Photographer: Simon Dawson/Bloomberg)
A man throws a fishing net into an inlet from the Arabian Sea. (Photographer: Simon Dawson/Bloomberg)

In continuation of my earlier article on the subject of the Reserve Bank of India’s balance sheet, its capital and reserves, here are a few additional questions worth probing.

First, if the Reserve Bank of India was to hypothetically transfer some ‘excess’ capital to the government, would this necessarily require a sale of foreign currency or domestic assets by the RBI? I make the case that this would not be automatically necessary.

Second, we know foreign currency revaluation gains do not pass through RBI’s profit and loss statement, they reflect directly into Currency and Gold Revaluation Reserves on RBI’s balance sheet. Can some of this reserve now be realised into the profit and loss account? I make the case that given the nature of accounting followed by the RBI, it is virtually impossible to ‘realise’ this reserve in the current context.

Third, most importantly, my previous article made the case that any reduction in the Contingency Fund—RBI’s retained earnings—should at best be used by the government to buy back its bonds with the RBI. Is there any further logical basis to this?

I make the case that other things being constant, any increase in the contingency funds would translate into higher lending by the RBI to the government, through RBI purchase or funding of government bonds.

Any standalone one-off dividend now, without a buyback of bonds by the government, would effectively imply that the same funds would have been both lent to and now separately transferred to the government.

Let’s take each of these up questions individually.

Transferring ‘Excess’ Capital To Government

RBI has the luxury of creating money by passing accounting entries. To transfer say Rs 2 lakh crore of its Rs 2.5 lakh crore retained earnings (contingency funds) to the government, RBI only has to pass the entry reducing its contingency funds by the amount and crediting the government’s account with it.

This entry would not impact the asset side of RBI’s balance sheet.

As the government then spends this Rs 2 lakh crore, the funds would move from the government’s account to individual bank accounts with the RBI. This too would only involve entries within the liability side of RBI’s balance sheet.

Once the funds reached the banking system, it would become part of the regular liquidity management operations of the RBI. In today’s context, the banking system is short on daily liquidity – banks start the day short of their required statutory Cash Reserve Ratio by around Rs 85,000 crore.

The addition of Rs 2 lakh crore would flip the start-of-day position of banks to a surplus of Rs 1.15 lakh crore.

This by itself would not force the RBI to sell any of its assets – it could well let the surplus remain in the banking system and absorb it through reverse repos. In the current context, it would only reduce the need for the RBI to buy bonds through open market operations or infuse liquidity through repos in the future.

In summary, the RBI can transfer money from its capital to the government by way of an accounting entry, without any asset side impact. This would eventually translate into a liquidity inflow into the banking system and become a part of the overall liquidity market operations of the RBI.

Opinion
In Government-RBI Spat, Reserve Bank Escapes With A Few Burns And Bruises

Realising Currency and Gold Revaluation Reserves

Can the reserves in RBI’s CGRA—a considerable RS 6.9 lakh crore as of June 2018—now be realised by the RBI selling down some of its reserves?

The annual accounts of RBI for fiscal year 2018 (July 2017 to June 2018), are worth considering in this regard.

  • In the nine months between July 2017 and March 2018, RBI purchased over $28 billion of foreign currency across spot and forward markets, when the dollar-rupee spot averaged around 64.00.
  • - Thereafter, between April 2018 and June 2018, as the rupee’s fortunes turned, RBI sold over $26 billion across spot and forward markets, with the dollar-rupee in the 68.00 range.
However, the purchase of dollars at 64.00 and sale at 68.00 did not result in realised currency gains of over Rs 10,000 crores during the fiscal, as one might have expected.

That’s because the RBI marks-to-market its currency assets every week, and takes this profit and loss into CGRA, outside the recognised profit and loss.

In other words, as the rupee depreciated in the last fiscal, the gain in RBI’s purchase of dollars at 64.00 was taken into CGRA over time, and currency reserves were already marked to 68.00 when the RBI sale intervention commenced. The entire Rs 10,000 crore alluded to earlier, therefore, accrued directly to RBI’s CGRA rather than recognised as profit and loss.

RBI’s June 2018 CGRA balance of Rs 6.9 lakh crore against foreign currency and gold assets of Rs 27.8 lakh crore indicates an effective holding dollar-rupee cost of around 48.50, assuming all its currency and gold assets was held in dollars. To recognize any profit and loss now by the sale of foreign currency, RBI would have had to follow one of weighted average or first-in-first-out or last-in-first-out methods of currency inventory valuation. As described earlier, it follows neither of these – it marks the whole inventory to market and takes any gains from these into CGRA, outside the profit and loss.

In other words, any sale of foreign currency now—at market—will simply not release any of the CGRA.

The only time there would be a case to release CGRA would be if the currency reserves came down to zero – which is impracticable.

While this might sound strange, this does ensure that the RBI and the government do not manipulate currency markets solely to recognise profit and loss. Had the RBI been following a weighted-average method of currency inventory valuation, for instance, a simultaneous sale and purchase of dollars today would have allowed RBI to recognise profit and loss against the dollar sale transactions and restate the holding cost of reserves to a higher dollar-rupee level.

At a practical level, therefore, if the government was to try and extract the reported Rs 3.6 lakh crore of ‘excess’ capital on RBI books, this presents difficulties. The Contingency Fund is only to the tune of Rs 2.5 lakh crore.

The balance Rs 1.1 lakh crore cannot be extracted from the CGRA – not without significant changes in accounting policy.

At best, therefore, RBI would have to tolerate the anomalous situation of having a negative Contingency Fund of Rs 1.1 lakh crore, offset by a large unrealisable but positive CGRA.

Opinion
Why RBI’s Capital Standard Must Remain Very High

Link Between Contingency Funds And RBI Holding Of Government Bonds

As discussed in the previous article, RBI’s contingency fund as been built over time by retaining some of the seigniorage income of RBI.

The RBI earns seigniorage by earning returns on its foreign currency and government bond assets, funded by interest-free currency in circulation and very low interest-bearing government and bank balances with RBI.

To the extent the RBI pays back seigniorage dividend to the government, the coupon paid by the government to the RBI against government bonds held by the RBI remains within the banking system.

However, to the extent that RBI withholds this seigniorage income and adds to its contingency fund, that represents a withdrawal of liquidity from the banking system.

Everything else being the same, to keep the system liquidity neutral, the RBI would have no choice but to replenish the liquidity withdrawal.

Surpluses can be tolerated – but systemic liquidity deficits would have to be filled by the RBI.

The most obvious way of doing this would be by RBI purchasing government bonds from the market.

All else remaining the same, all accretions to the contingency fund by withholding seigniorage would have resulted in the RBI buying government bonds – in other words, lending a like amount back to the government. There is a connect, therefore, between the Rs 2.5 lakh crore of contingency fund and the Rs 6.3 lakh crore of government bonds on RBI books.

To seek a standalone one-off dividend payment out of the contingency fund now, without any buyback of bonds by the government would therefore be a double deployment of the same money towards the government.

The first time by way of a lending by the RBI to the government, and the second, a transfer of dividend. Seen in this light, the latter would tantamount to monetisation of the deficit – plain and simple.

Opinion
Don’t Need Extra Funds From RBI To Meet Fiscal Deficit Target, Says Jaitley

In Summary

Technically, the RBI can transfer ‘excess’ capital from its contingency fund to the government by way of a simple accounting entry, without necessarily entailing any asset sales by the RBI.

Given the accounting norms of the RBI, its CGRA cannot be realised into profit and loss without a significant change in its accounting policy.

Lastly, other things being equal, the creation and expansion of the contingency fund would have resulted in the RBI buying government bonds, i.e. lending money to the extent of contingency fund created to the government. Though feasible in an accounting sense, it would be disingenuous to seek a reduction in the contingency fund now by way of a one-off dividend to the government, without reducing the amount of lending by the RBI to the government – i.e., without the government buying back its bonds on RBI’s books. If it is still forced through without a concomitant bond buyback, it would smack entirely of a direct monetisation of the fiscal deficit.

Ananth Narayan is Associate Professor-Finance at SPJIMR. He was previously Standard Chartered Bank’s Regional Head of Financial Markets for ASEAN and South Asia.

The views expressed here are those of the author and do not necessarily represent the views of BloombergQuint or its editorial team.