Personal Website of R.Kannan
Students Corner - A Decade of Economic
Reforms in India - A Review

The changes in the operating procedure of monetary policy and the growing openness of the economy during the 1990s impacted both on the size and composition of reserve money. A distinct feature during the 1990s was the growing influence of capital inflows on the conduct of monetary policy. With capital flows in excess of the current account deficit, the Reserve Bank absorbed the surplus foreign exchange in the market. As a result, the share of net foreign assets (NFA) in reserve money increased to 78.1 per cent as at end-March 2002 from 7.8 per cent as at end-March 1990, with implications for monetary expansion. As on March 14, 2003, the ratio stood at 97.0 per cent. Nonetheless, the Reserve Bank could effectively sterilise such capital flows and keep reserve money under control by trading the surpluses on the external account with the fiscal deficit and thus, neutralise its own balance sheet. In this context, it may be noted that for a part of the accretion to NFA in the form of aid receipts, revaluation and income from foreign assets, there is no monetary impact and hence no need for sterilisation. For instance, such flows amounted to 64 per cent of incremental NFA during 1998-2002; in 2002-03 (up to March 14, 2003), the ratio was 31 per cent.

The share of the net Reserve Bank credit to the Centre in reserve money declined over the 1990s. More recently, the net Reserve Bank credit to the Centre has recorded a decline even in absolute terms in the face of a sustained surge in capital flows. In the process, the monetisation of the Centre’s fiscal deficit (as a proportion to GDP) declined from 2.1 per cent during the 1980s to 0.7 per cent during the 1990s. As a result, reserve money expansion could be contained to 14.3 per cent during the 1990s as compared with 16.5 per cent during the 1980s.

An additional factor behind the lower reserve money growth was a sustained cut in cash reserve requirements since the late 1990s, in line with the stated policy objective to reduce it to the statutory three per cent. As a result, the M3 multiplier jumped to 4.4 as at end-March 2002 from 3.1 as at end-March 1995 thus amplifying the monetary impact of primary liquidity in the recent years. The overall M3 growth rate, therefore, worked out to 17.3 per cent per annum during the 1990s, close to 17.2 per cent in the previous decade although reserve money expansion decelerated during the 1990s.

A stylised fact of the 1990s, as noted earlier, was a sharp rise in net foreign assets of the Reserve Bank with a corresponding decline in net domestic assets reflecting sterilisation by the Reserve Bank. An issue that needs to be examined is whether it was the reduction in net domestic assets that caused subsequent capital inflows or whether they offset the previous capital inflows. For most of the countries, both lines of causality could be at play depending upon the degree of capital account liberalisation and sensitivity of foreign flows to interest rate differentials. An examination of the offset coefficient – the response of net foreign assets to net domestic assets - for a sample of six countries studied by Schadler et al (1993) showed that all, except Thailand, had scope for effective sterilisation. For India, Granger causality tests indicate a uni-directional causality from changes in NFA to net domestic assets (NDA).8 Moreover, the sterilisation coefficient - the response of change in NDA to that in NFA - was (–) 0.84 during April 1994-December 2002, i.e., an increase of Rs. 100 in NFA attracted a policy response of sterilisation that drained away NDA worth Rs. 84 from the system9. In other words, the results do not suggest that capital inflows were in response to domestic monetary conditions. This indicates scope for effective sterilisation. In theory, surges in capital flows should ease domestic money market conditions and lower interest rates. Given the estimated sterilisation coefficient for India, the authorities were largely able to steer interest rates consistent with domestic macroeconomic conditions

The sustained increase in the ratio of net foreign assets to net domestic assets during the 1990s reflected both the scale effect (the CRR cuts) and the substitution effect (sterilisation) of the changes in the monetary policy framework and structural changes. This has two implications. First, the reduction in NDA limits the Reserve Bank’s ability to sterilise capital flows in future. Second, this also impacts the Reserve Bank’s income in terms of the differential between domestic and international interest rates.

While the Reserve Bank is increasingly able to ensure orderly conditions in the financial markets, the critical issue in determining the effectiveness of the signals emanating from the changes in the policy rate is the degree of ‘pass-through’, i.e., the speed and the magnitude of the response of the market interest rate spectrum to the monetary policy signals (RBI, 2002d). Between March 1998 and February 2003, the Bank Rate and the repo rate were cut by 425 basis points and 250 basis points, respectively. In addition, the CRR was reduced by 550 basis points over the same period. The repo rate was cut by a further 50 basis points in March 2003. The easing of the monetary policy stance was mirrored in a general softening of interest rates in the money markets (with call rates declining by almost 325 basis points) and in the government securities markets (with the yield on 10-year government securities declining by almost six percentage points). However, the prime lending rates of major banks remained sticky. This suggests a low level of pass-through of the changes in the policy rates on to the lending rates, thereby blunting the efficacy of the monetary policy. Over a medium-term horizon, the spreads of lending rates of commercial banks over their average costs of deposits reveal a marginal narrowing down since 1996-97. The spreads, however, still continue to be fairly large. The spread of the Central Government’s market borrowings over the banks’ average cost narrowed from 5.2 per cent during 1990-96 to 4.4 per cent during 1996-2002 while that of other borrowers declined from 9.2 per cent to 8.1 per cent over the same period.

This relative downward inflexibility in the commercial bank interest rate structure can be attributed to a number of structural factors in the Indian banking system (RBI, 2002e):

• Average cost of deposits for major banks continues to be relatively high (6.25 to 7.25 per cent). This could be attributed partly, in turn, to the returns on alternative savings instruments in the country. Despite reductions in the administered interest rates on small savings and provident funds in the recent period, these instruments still yield higher returns than bank deposits, which are made even more attractive by tax benefits. This, therefore, constrains the banks’ ability to reduce deposit rates. Furthermore, in the absence of a well-developed pension market, bank deposits continue to be the dominant source of income for senior citizens. This hinders sharp cuts in interest rates on bank deposits. Moreover, although the inflation rate has declined in the recent period, its year-to-year variability, as noted earlier, remains high. This might have prevented a commensurate decline in inflation expectations and thus, keeping nominal rates high.

• Substantial portion of bank deposits remains in the form of long-term deposits at fixed interest rates. This limits the flexibility available to banks in reducing their lending rates in the short-run without adversely affecting their return on assets.

Relatively high overhang of non-performing assets (NPAs) further pushes up the average cost of funds for banks, particularly public sector banks.

• Non-interest operating expenses of banks work out to 2.5 to 3.0 per cent of total assets, putting pressure on the required spread over the cost of funds

• In view of legal constraints and procedural bottlenecks in recovery of dues by banks, the risk-premium tends to be higher resulting in a wider spread between deposit rates and lending rates

• In the context of the current cyclical slowdown of the economy, especially the industrial sector, the probability of defaults and late payments by borrowers increases; this could lead to an increase in risk premium and hence a stickiness in the PLR.

• The large borrowing programme of the Government, over and above SLR requirements, gives an upward bias to the interest rate structure.

In order to overcome these rigidities and to impart more flexibility to the interest rate structure, the Reserve Bank has recently initiated a number of measures. Banks were encouraged to introduce a flexible interest rate system option for all new deposits and urged to review and announce the maximum spreads over PLR. To have a greater degree of transparency with regard to actual interest rates for depositors as well as borrowers, banks were asked to provide information on deposit rates for various maturities, the effective annualised return to the depositors and maximum and minimum interest rates charged to their borrowers. These factors are expected to increase the degree of pass-through over time. For instance, the median range of interest rates at which at least 60 per cent of business is contracted narrowed from 12.75-14.00 per cent as at end-June 2002 to 12.00-14.00 per cent as at end-September 2002.

It would be useful to compare the Indian experience of pass-through with that of a developed economy. In contrast to the Indian experience, the pass-through in the USA is almost instantaneous and complete. For instance, between January 2001 and January 2003, the Fed Funds rate declined by 474 basis points; the prime rate over the same period declined by 480 basis points. The correlation coefficient between the two rates was, therefore, nearly unity. The spread of the prime rate over the Fed Funds rate was almost constant at 300 basis points in the US as compared with almost 600 basis points over the Bank Rate in India at present. The pass-through is relatively lower on to the long-rates. For instance, the correlation coefficient of 10-year Treasury yields with the Fed Funds rate was 0.70 over the period January 1997 to January 2003. On the other hand, in India, although long-term yields fell in line with the monetary easing, the prime lending rate did not show much co-movement. This brings into sharp focus the importance of a faster pass-through for an effective monetary policy.

The low interest pass-through, in the context of a decline in the inflation rate, is reflected in a hardening of the real interest rate (Table 5.9). The impact differed across various sectors in view of divergent trends in the pass-through as well as the relevant inflation rates. In particular, the appropriate real interest rate for borrowers increased significantly as manufacturing inflation declined more sharply than the headline inflation while the nominal rates remained relatively sluggish. The sluggishness of nominal rates, in turn, emanated from a combination of downward rigidity of deposit rates, low pass-through and risk-aversion by banks in view of the prudential norms. In the case of the Central Government, the increase in real interest rates could be attributed partly to the fact that the administered interest rate mechanism continued during the first few years of the 1990s. For the Central Government, real interest rates, in fact, fell sharply in 2002-03 on the back of a substantial decline in nominal yields. The key issue, in terms of the growth objective, is the impact of the structure of real interest rates, especially as the interest cost as a proportion of sales of corporates is much higher as compared with many emerging market economies (Mohan, 2002).

For India, econometric evidence suggests that a 100 basis point rise in real interest rates depresses the real GDP and hence widens the output gap (actual less trend output) by six basis points in the short-run and almost 60 basis points in the long-run.10 Available empirical evidence also suggests that the response of output to monetary policy signals is asymmetric, with a tight monetary policy being more effective than an easy monetary policy. For the US, the short-run response of output to increases in the Fed Funds rate is estimated to be more than twice of the response to decreases in the Fed Funds rate (Piger, 2003).