Monetary policy transmission mechanism

How does a change in one overnight rate, decided by six or twelve people in a room, propagate out into the ₹300-lakh-crore or $28-trillion real economy and eventually nudge the price of a haircut? That chain is the thing an analyst most needs to feel rather than merely recite. It is the monetary-transmission mechanism, and it runs, with leaks and lags at every joint, like this:

Step 0: the decision. The MPC/FOMC changes the policy rate (say, a 25-bp cut), or changes the stance/guidance, or the quantity of liquidity (OMO/QE).

Step 1: money-market rates (days). The overnight rate (WACR/EFFR) moves to the new policy rate within the corridor almost immediately, because arbitrage and OMOs enforce it. Short-term wholesale rates (T-bills, commercial paper, certificates of deposit) follow within days. This first link is fast and reliable: it is the one part of transmission the central bank genuinely controls.

How do bank deposit and lending rates respond?

Step 2: bank deposit and lending rates (weeks to quarters). This is the leakiest, most-studied joint, and India has engineered it hard. Banks reprice loans and deposits toward the new policy rate, but how fast depends on institutional plumbing. India's key reform: since October 2019 the RBI has mandated that new floating-rate retail and small-business loans be linked to an external benchmark (EBLR), almost always the repo rate itself, so that when the repo rate moves, EBLR-linked home, auto, and MSME loans reprice automatically, within a quarter. This dramatically sped up transmission on the lending side (before it, banks set rates off the opaque, sticky internal MCLR/base rate and dawdled on passing cuts through). The deposit side is stickier, because term deposits are locked at old rates until they mature. That is why a rate cut compresses bank margins in the short run (loans reprice down faster than deposits) and a hike expands them (the asset-sensitivity point from M5.01). The US analogue: the policy rate flows to the prime rate (which moves lockstep with fed funds), to floating corporate loans priced off SOFR, and, much more loosely, to mortgage rates, with a subtlety worth the analyst's attention: the US 30-year fixed mortgage tracks the 10-year Treasury yield (plus a spread), not the current fed funds rate, so US mortgages are driven by the expected path of rates (M7.03), which is why they can fall while the Fed is still hiking if the market expects cuts ahead. Transmission is not one pipe; it is a bundle of pipes of different lengths.

How does the bank lending channel work?

Step 3: the cost and availability of credit (quarters). Two things move, and beginners forget the second. The price of credit changes (rates up or down). But so does its availability, through the bank-lending / credit channel: when policy tightens, banks facing higher funding costs and thinner capital cushions ration credit (tighten lending standards, shrink loan books), and marginal borrowers are cut off entirely, not just charged more. This quantity effect is often bigger than the price effect, especially for small firms with no access to bond markets, and it is why credit growth and lending-standards surveys (the Fed's SLOOS; the RBI's credit data) are leading dashboard indicators.

Step 4: spending and investment (several quarters). Cheaper, more-available credit → households borrow and spend more (homes, cars, durables: the interest-sensitive sectors move first), firms find more projects clearing the (now-lower) hurdle rate and invest more (M3.03: the WACC just fell), and existing borrowers on floating debt have lower payments and more free cash to spend. Working alongside the credit channel are: the asset-price / wealth channel (lower rates lift bond, equity, and house prices → households feel richer and spend, see why it all matters), and the exchange-rate channel (lower rates → capital flows out → the currency weakens → exports cheaper and imports dearer → net exports rise and imported inflation rises; the RBI weighs this heavily, M7.06).

Step 5: the output gap, then inflation (a year or more). Higher aggregate demand closes the output gap (pushes actual output toward or past the economy's capacity), the labour market tightens, and, eventually and only if demand runs ahead of supply, wages and prices rise toward or through the target. This last link is the slowest, the loosest, and the one the central bank is actually aiming at, which is the whole problem, stated next.

What are Friedman's long and variable lags?

Friedman's long and variable lags. Milton Friedman's most durable practical insight, and the reason monetary policy is genuinely, permanently hard, is that the lag between a policy action and its peak effect on the real economy and inflation is long (commonly estimated at 12–24 months, sometimes longer) and variable (its exact length changes from cycle to cycle and cannot be known in advance). Sit with what that means for the person setting the rate. You are steering a vehicle whose response to the wheel comes a year to two years later, and you do not know whether it will be twelve months or twenty-four this time. Three consequences follow, and they explain almost everything about how central banks behave and misbehave:

1. Policy must be forward-looking. You cannot set today's rate for today's inflation. Today's inflation is the result of decisions made 1–2 years ago, and today's decision won't bite until 1–2 years hence. So central banks must act on forecasts of where inflation and growth will be, which is why their projections (the SEP, the RBI's forecasts) are the actual object of policy and why the forecast, not the current print, is what moves markets.
2. Over- and under-shooting are near-inevitable. Because the effect is delayed and uncertain, a central bank that keeps tightening until it sees inflation fall has almost certainly tightened too much (the medicine already in the pipeline will keep working after they stop). That is the classic mechanism by which central banks cause the recessions that end inflations. "Long and variable lags" is why the safe-looking policy of "wait for the data" is itself a way to go wrong.
3. Attribution is treacherous. When inflation falls, was it this central bank's hikes, or the last cycle's, or a supply shock reversing? The lags make cause and effect nearly impossible to disentangle in real time, which is why honest central bankers are humble and dishonest ones claim credit for whatever happened. The analyst's posture (M6.02's Marks, M7's humility thread): read the central bank as steering with a delay under deep uncertainty, not as an omniscient controller. Never assume a rate change has "already worked" or "done nothing" based on the last few months of data. It is still in the pipe.