Project on Assessment of Key Issues Related to Monetary Policy
[Source: RBI Report on Currency & Finance 2003-04]

Module: 3 External Sector Openness And Conduct Of Monetary Policy : The Indian Experience

Transformation of Reserve Bank's Balance Sheet Due to Accretion of Forex Reserves (Contd)

Notwithstanding these measures, the overall balance of payments surpluses not only persisted but also widened in each of the years since 2001-02. Accordingly, a key instrument of managing capital flows has been sterilisation through outright open market sales of Government of India securities by the Reserve Bank from its portfolio. In recent years, these have been supported by sucking of liquidity through operations under the Liquidity Adjustment Facility. The scale of sterilisation operations - in particular, in the fourth phase (2001-02 onwards) - can be gauged from movements in the outstanding net Reserve Bank credit to the Central Government (NRBICG). Although balance of payments had been recording an overall surplus in each of the years since 1993-94 (with exception of one year, 1995-96), outstanding NRBICG also increased in all years except one even as open market operations were undertaken. In contrast, in the most recent phase, NRBICG has recorded sharp declines. Illustratively, the NRBICG declined from Rs.1,67,308 crore at end-May 2001 to Rs.4,626 crore by December 10, 2004.

The finite stock of the Government of India securities with the Reserve Bank brought into sharp focus the limitations on the Reserve Bank's ability to sterilise capital flows in future and its implications for liquidity management and inflation. For instance, the Reserve Bank's Annual Policy Statement for the year 2004-05 observed: "the lagged effect of persistence of excess liquidity on aggregate demand cannot be ignored as it could have some potential inflationary impact". At the same time, the challenge of sterilisation, in the Indian case, is not very acute, per se, because the large order of fiscal deficit allows the banking system to park the surplus liquidity emanating from capital flows in gilt-edged paper.

Against this backdrop, the Reserve Bank constituted an Internal Working Group on Instruments for Sterilisation (RBI, 2003d). The Group recommended, inter alia, that use of CRR as an instrument of sterilisation, under extreme conditions of excess liquidity and when other options are exhausted, should not be ruled out altogether by a prudent monetary authority ready to meet all eventualities2 . The Group stressed the need to take into account the consolidated balance sheet of the Government and the Reserve Bank as it is immaterial as to whether sterilisation costs are borne by the Reserve Bank or the Government since, in the consolidated balance sheet framework, the net cost is the same. Accordingly, the Group strongly recommended against the introduction of central bank bills/bonds, in the context of current fiscal situation and considerations of market fragmentation. A key recommendation of the Group was the introduction of Market Stabilisation Bills/Bonds (MSBs) for mopping up liquidity from the system. The Group also recommended that surplus balances of the Government may be maintained with the Reserve Bank without any payment of interest so as to release securities for open market operations.

Following these recommendations, a Market Stabilisation Scheme (MSS) was introduced in April 2004 wherein Government of India dated securities/Treasury Bills are being issued to absorb liquidity. These dated securities/Treasury Bills are the same as those issued for normal market borrowings and this avoids segmentation of the market. By December 10, 2004, the outstanding issuances under MSS were Rs.51,334 crore. The issuance of securities under the MSS is expected to enable the Reserve Bank to improve liquidity management in the system, to maintain stability in the foreign exchange market and to conduct monetary policy in accordance with the stated objectives. Moreover, the MSS scheme will bring transparency in regard to costs associated with sterilisation operations. Hitherto, the costs of sterilisation were fully borne by the Reserve Bank in the first instance and its impact was transmitted to the Government in the form of lower profit transfers. With the introduction of the MSS, the cost in terms of interest payments would be borne by the Government itself in a transparent manner.

Sterilisation operations, as discussed above, have been a key instrument of managing capital flows. Reflecting these operations, the net Reserve Bank credit to the Central Government has declined in the recent years while net foreign assets have been increasing. In this context, a critical issue is: whether it is the reduction in net domestic assets (NDA) that caused subsequent capital inflows or whether the reduction in NDA offset the previous capital inflows. The former view would suggest that capital inflows reflect higher money demand by residents and, if so, sterilisation operations are ineffective. This would occur if sterilisation operations place upward pressure on interest rates and the assets are perfect substitutes. In this case, even a small rise in domestic interest rates would attract large capital inflows rendering sterilisation operations ineffective. For India, evidence suggests that sterilisation operations have been effective. Over the period 1995-2004, close to two-thirds of capital flows were sterilised by the Reserve Bank through open market sales/repo operations³. This enabled the Reserve Bank to keep base money growth close to the desired trajectory.

Following Kouri and Porter(1974), the ability of a central bank to carry out sterilisation operations effectively can be examined by analysing the relationship between a central bank's net domestic assets (NDA) and its net foreign assets (NFA). The "offset" coefficient - the response of net foreign assets to net domestic assets - measures the degree to which capital inflows offset the effect of a change in NDA on money supply. An offset coefficient close to unity would imply that the efforts of the monetary authority to tighten monetary policy would induce equal and offsetting foreign inflows leaving no scope for independent monetary policy. In contrast, an offset coefficient of zero would provide the monetary authority with complete discretion in the conduct of monetary policy. A number of studies have, therefore, attempted to estimate offset coefficients. In most of the studies, offset coefficients are less than (-) 0.5 which suggests that monetary authorities have some scope for sterilisation. Moreover, consistent with the hypothesis of increasing capital mobility in the 1990s and the consequent declining monetary policy independence, there is evidence that the (absolute) offset coefficients have increased. For India, the estimated offset coefficient [(-) 0.3 during 1993-97] suggests that sterilisation operations conducted during this period enabled sufficient independence for monetary policy to pursue domestic goals.

For India, available evidence indicates a uni-directional causality from NFA to NDA, i.e., capital inflows were not induced by domestic monetary conditions. The sterilisation coefficient - the response of change in NDA to that in NFA - is found to be (-) 0.92, i.e., an increase of Rs.100 in NFA induced a policy response of sterilisation that drained away NDA worth Rs.92 from the system (RBI, 2004a). All accretions to NFA do not have a monetary impact; for instance, aid receipts, revaluation and the Reserve Bank's income on its foreign assets contribute to NFA but have no monetary impact, obviating the need for sterilisation to that extent. As such, following RBI (2004a), sterilisation efforts can be gauged by examining the impact of the Reserve Bank's net market purchases/sales of foreign currency from/to authorised dealers (ADPURC) on the net Reserve Bank credit to the Centre (NRBICC), and not the entire NDA. For India, data on market sales/ purchases are available effective October 1995. As in the previous case, Granger causality tests indicate a unidirectional causality from changes in foreign exchange purchases to reduction in NRBICC⁴. The sterilisation coefficient is 0.63, i.e., Rs.100 increase in foreign currency purchases from ADs induces sterilisation operations involving sales of Government securities worth Rs.63 from the Reserve Bank .

DRBICCG and ADPURC are defined as before. DIIP and DCRRAVG are monthly variations in the index of industrial production and average CRR, respectively. In addition, monthly dummies for April, May, August, September and November turned out to be significant and were included in the estimated equation.

³ Such open market sales transferred the Government securities from the balance sheet of the Reserve Bank to that of the commercial banks. In an environment of softening interest rates, investment in government securities turned out to be relatively attractive and profitable for the commercial banks. At the same time, investments in government securities are subject to market risks arising from fluctuations in market rates of interest. With the upturn of the interest rate cycle, there could be an adverse impact on banks' profitability. In this context, it is relevant to observe that the Reserve Bank had advised banks to build-up Investment Fluctuation Reserves to meet such eventualities (see Module 5).

⁴ In a bivariate VAR of net monthly sales/purchases of foreign exchange from ADs (ADPURC) and monthly variations in net Reserve Bank credit to Centre (DRBICCG) over the period October 1995 to March 2004, the null hypothesis of Granger non-causality of DRBICCG can not be rejected (chi-square of 0.03 at p-value of 0.87). On the other hand, the null hypothesis of Granger non-causality of ADPURC can be easily rejected (chi-square of 5.90 at p-value of 0.02). The VAR was estimated with one lag based on Schwarz Bayesian Information Criterion (SBIC).