A V Rajwade: Sterilisation & CRR - A thought experiment

WORLD MONEY

A CRR hike, to the extent it neutralises the increase in money supply, is not a tax on the banking system.

It is quite feasible to calibrate the intervention so as to leave interest rate and liquidity conditions roughly where they were before the specified dollar purchase-cum-sterilisation.

""Shankar Acharya, "Midsummer Madness", August 9, 2007

As readers may recall, I have more than once argued the case for using changes in the CRR as an instrument of sterilising the growth in money supply in the market, arising from central bank intervention. I believe that this is a far superior system to the creation of securities under the market stabilisation scheme (MSS), the cost of which is borne by the tax payer. The issue is important as the cost of sterilisation (the difference between the coupon on the created securities, and the return foreign exchange reserves earn) seems to be a major deterrent to keeping the exchange rate competitive. There are of course significant costs attached to an appreciating currency. The major ones are:

The translation losses for the central bank from appreciation of the domestic currency.

Profit compression/losses in the tradables sector resulting into lower direct and indirect taxes, and hence an increase in the fiscal deficit.

The disincentive to investment, growth and employment creation, and loss of existing jobs.

After the introduction of the MSS, the RBI tried to avoid using CRR as a sterilisation measure. This policy stance was probably taken for two reasons:

Increasing CRR militates against the long-term policy goal of bringing it down to 3 per cent.

It is a "tax" on the banking system.

The first argument has already lost whatever strength it had as, in recent months, CRR has been increased more than once. As for the second, this article argues that an increase in CRR to the extent it neutralises the increase in money supply resulting from intervention, is not a tax on the banking system; that it would leave the system's earnings and expenses unchanged.

I must confess that I have not found too many takers for my argument, and hence the "thought experiment" described in the following paragraphs. (I am of course aware how passe thought experiments look in modern economic debates: these require empirical analysis based on a dozen equations, with an even larger number of Greek letters thrown in, as manifestations of the rigour of the argument.)

At the heart of my thought experiment is a foreign exchange market consisting of just two players, Bank A and Bank B. On a given day, Bank A needs to sell $100 million, while the underlying demand from Bank B is only for $80 million. We consider two possibilities:

No intervention:
The domestic currency appreciates sufficiently to attract discretionary buyers to bridge the gap of $20 million between supply and demand (such discretionary demand could come, for example, from Bank B wanting to take a trading position, going long in the dollar, at the new price; or from a customer of that bank wanting to buy dollars for an underlying commercial transaction, to take advantage of the better price. Either way, Bank B transfers rupees equivalent to $100 million to Bank A, which uses them to pay the beneficiaries of $100 million received by it. We need not consider what the beneficiaries do with the rupees "" this is identical whether there is intervention or not.

The central bank intervenes:
The alternative scenario is that the central bank intervenes to buy the excess supply of $20 million and the rupee does not appreciate. But in that case, Bank B now has surplus rupees to the extent of $20 million, as it has paid out rupees only to the extent of $80 million, the demand at the unchanged exchange rate. The position of Bank A is the same in both the scenarios.

If the central bank now increases CRR by a percentage sufficient to mop up rupees equivalent to $20 million, Bank A may need to borrow the money needed for the increased CRR, from Bank B; the latter's surplus (rupees equivalent to $20 million) is used partly to fund its own increased CRR, and the balance for lending to Bank A. The latter's costs go up to the extent of the interest on the borrowed funds "" but Bank B's income goes up exactly to the same extent, leaving the system's income unchanged, under the two scenarios.

In short, the sterilisation through a corresponding increase in CRR is not a tax on the banking system; contrarily, sterilization through either open market operations or MSS is a subsidy to the banking system! It earns interest on money created by the central bank through intervention.

avrajwade@gmail.com