The numbers explain why. Broad money, or M3, consists principally of currency and bank deposits. Its growth accelerated from around 9.3 per cent in 2024-25 to 15.3 per cent in 2025-26, while nominal gross domestic product (GDP) grew by less than 9 per cent in the latter year. Annual money growth remained at 16 per cent on October 15. Our compilation places M3 at approximately 91 per cent of GDP in March and around 93 per cent by September, using projected GDP. The ratio was approximately 86 per cent 18 months earlier. This represents substantial monetary expansion relative to income.
There is, however, a serious intellectual argument against assigning money growth an independent role in policymaking. In the New Keynesian framework associated with the recent Nobel prize winners, spending responds to current and expected future real interest rates. The relevant real rate is the nominal rate minus expected inflation. Lower real rates encourage spending today and weaken the incentive to postpone consumption. When prices adjust slowly, stronger demand eventually pushes up inflation.
From this perspective, monetary policy works through interest rates, expectations, and financing conditions. The central bank need not separately target money if these channels adequately capture the stance. Money may expand because banks lend more and create deposits, making it partly an outcome of economic decisions rather than an independent cause. Financial innovation also makes money difficult to define, while the central bank cannot precisely control broad money.
Moreover, creating bank reserves does not guarantee additional lending. Banks may retain liquid assets because they fear losses, lack capital or cannot find creditworthy customers. Borrowers may prefer repayment to fresh borrowing. The American experience after the global financial crisis illustrates why reserve creation need not translate proportionately into money, credit or inflation.
Monetarists would respond that this places too much confidence in policymakers’ ability to identify the correct interest rate. A nominal rate can look restrictive while being expansionary once expected inflation is factored in. Sustained growth in money beyond the public’s willingness to hold it encourages adjustment through spending and purchases of assets. Money, therefore, provides information that an interest-rate diagnosis can miss. The relationship is neither immediate nor mechanically proportional: Desired money holdings and the speed of spending also change.
India has two reasons to take this monetarist warning seriously.
First, the possibility that additional liquidity remains disconnected from lending is less reassuring when credit is already expanding strongly. Bank credit reached ₹227.43 trillion on October 15, growing 18.2 per cent over the preceding year, compared with 11.6 per cent a year earlier. This does not establish that reserve creation caused every additional loan. It establishes that we are not observing abundant reserves alongside stagnant bank lending. Money and credit are expanding together.
Second, and more importantly, India lacks a sufficiently reliable measure of the inflation expectations driving spending, saving and price-setting. Household survey expectations persistently exceed subsequent national consumer price index-based inflation. Some discrepancy reflects households’ different consumption baskets. Survey movements can still be informative. But using these numerical expectations to calculate the real rate remains problematic.
Professional forecasts help predict reported inflation, but they cannot substitute for understanding the expectations governing wage negotiations, rents, school fees and contract renewals. Those decisions determine how anticipated inflation enters future prices. We need better surveys and evidence from contracts committing parties to future payments.
Until then, monetary aggregates offer a valuable cross-check. Even someone convinced that money has no independent causal role should ask whether its growth improves inflation forecasts. Prediction does not require establishing that the predictor independently causes the outcome. If money growth helps anticipate inflation, ignoring it because causality is disputed would discard useful information.
This is compatible with a New Keynesian framework. If expected inflation is mismeasured, our estimate of the real rate is wrong. Rapid monetary and credit growth should prompt a reassessment: Is policy providing more stimulus than the nominal repo rate suggests?
Nor is all monetary growth dangerous. A growing economy needs more money, and financial deepening can justify a higher money-to-GDP ratio. The warning is persistent, substantial expansion beyond income growth without evidence of a corresponding increase in desired money holdings.
India’s historical experience suggests that 80-84 per cent of GDP is a useful reference range for broad money, not a scientifically established optimum. Returning towards it gradually would require money growth to lag nominal income growth, rather than an abrupt monetary contraction.
The repo rate should remain the principal instrument and inflation the objective. But money growth deserves an explicit place in forecasts, minutes and communication. When expectations are poorly measured, an interest-rate framework needs more cross-checks, not fewer.
The writer teaches at Indian School of Business
 "RBI, Reserve Bank of India")