The theory is that loan and other debt capital flows respond to interest rate differentials. So if country A were to raise its interest rate with respect to another, capital would tend to flow to such country A. If these flows were in excess of what was needed to finance the trade deficit (assuming one existed in A) then As currency would begin to appreciate. This of course quietly assumes that country A is normal by developed country standards. That is, lenders have some faith in the credit of country A and it has complete capital convertibility.
The latter assumption powerfully affects our context. For exchange rate purposes, it is the short-term rate that matters. Since banks do most of the exchange rate business, the relevant rate is the one at which banks can borrow. Internationally, the reference rate is commonly the London Inter-Bank Rate (LIBOR), which hovers around the key policy rate in the concerned currency, eg, the US Fed rate for US dollar LIBOR. In India it is the call money rate, which hovers around the RBIs repo rate. So, traditionally the differential between LIBOR and repo rate has been seen as important for the level and direction of the rupee.
Since the end of 2000, when the RBI set about lowering the yield on gilts, it was unwilling for the short end to keep pace with US Fed or LIBOR rates. As recently as June 2002, the rupee had always gone down and the RBIs constant concern was to avoid disruptive declines. Now, if keeping repo rates much higher (say 5 per cent against 1 per cent on the dollar) helped the rupee keep at levels from which it would otherwise have declined, the inverse should also hold. With the current account in balance and other capital flows strong, rupees excessive hardening is clearly undesirable.
Sorry, but cutting the repo rate does not work. In the absence of capital account convertibility, the only access of foreign funds to Indias gilt market is capped at some $1.5 billion. Once that cap is reached, the repo rate ceases to matter. If the repo rate is not the operative one, what is The answer is the rate offered on Non Resident Indian (NRI) deposits by our banks. FCNR rates, which deposits are denominated in foreign currency, had been aligned with global rates a long time ago. As a result, FCNR deposits have stagnated at $10 billion. The rates on rupee denominated NRI deposits (NRE) were however much higher, in line with our domestic rates. In early January 2002 the rate for 3- to 4-year deposits was as high as 8 per cent; in January 2003 the rate for 1-year deposit was still 6 per cent.
On 17 July 2003, the RBI stepped in to cap interest rates at 250 basis points (bps) over LIBOR. On 15 September, it reduced the cap to 100 bps and finally on 18 October, it slashed it to a mere 25 bps. NRE deposits aggregate some $22 billion, with average maturity of between 1.5 to 2 years. That is, each month some $1 billion worth mature. With current interest rates at 1.35 per cent for 1-year deposits, and with the risk of rupee depreciation upon the depositor, it is unlikely that much of the maturing deposits would be renewed. A monthly rate of $1 billion is somewhat larger than the total of other capital flows and the current account surplus. Hence the measure completely takes the upward pressure off the rupee.
The great thing is that, if need be, it is easy to reverse course by simply raising the cap. Most importantly, changes in this do not impact domestic interest rates whatsoever. Given the large current stock of NRI deposits and the increasingly larger and wealthier NRI population, this particular lever should do excellent duty for several years to come.
The author is economic advisor to ICRA (Investment Information and Credit Rating Agency)