Central Banks & Monetary Policy

One overnight rate, set by a committee, propagating into every price on this site. The macro layer sitting under the yield curve.

What a central bank actually controls

• It sets one price: the rate on overnight money between banks, steered by the rate paid on reserves and the rate charged to lend against collateral.
• Everything else — mortgage rates, corporate spreads, currency levels, equity valuations — is a consequence, transmitted through markets that can amplify, delay or ignore the signal.
• That single lever is why the yield curve matters so much: it is the market's statement about where that one rate goes next, extended out for decades.

The corridor: how the rate is actually enforced

• A ceiling — banks can borrow from the central bank against collateral, so nobody lends above that rate in the market.
• A floor — banks earn interest on reserves held at the central bank, so nobody lends below it.
• The market rate sits between them, and in an ample-reserves system it sits essentially on the floor, because banks have more reserves than they need.
• The OIS market prices the expected average of that overnight rate, which is why OIS is the cleanest read on policy expectations available and the standard discounting curve for collateralised trades.

Transmission: five channels, all imperfect

• The lags are long and variable — commonly cited as twelve to eighteen months for the real economy. A committee is therefore always setting policy for conditions it cannot observe.
• The asset-price channel is instant and the real-economy channels are slow, which is why markets reprice on the announcement and the economy responds a year later. Most commentary confuses the two.
• Structure changes potency. An economy of long fixed-rate mortgages transmits rate rises far more slowly than one of floating-rate or short-fixed loans — the same policy move has materially different force in different countries.

The Taylor rule: a benchmark, not a recipe

A simple rule that captures much of what committees actually do: start from the neutral real rate, add inflation, then lean against the inflation gap and the output gap.

How markets price the path

• Forward rates from OIS, STIR futures and FRAs together give a dated path for the policy rate — meeting by meeting, months ahead.
• Financial media translate this into "the market prices two cuts by June". That translation is arithmetic: the change in the forward rate divided by the size of a standard move.
• A forward is a price, not a forecast. It contains a term premium and reflects hedging demand as much as expectation, which is why forwards have historically been poor predictors of the actual path while remaining the correct rate to hedge at.

Quantitative easing, and what it did

• Mechanically: the central bank buys bonds and creates reserves to pay for them. The seller ends up holding a deposit instead of a bond; the bond sits on the central bank's balance sheet.
• It does not "print money" into the economy. Reserves circulate only between banks and the central bank. Broad money grows if banks lend more — which is a consequence, not a mechanism.
• The intended channels: compressing the term premium at the long end, signalling that policy will stay easy, and pushing investors out of government bonds into riskier assets.
• The measured effect on long yields was real but modest per programme; the effect on asset prices was larger and faster. Whether it reached the real economy proportionately remains genuinely contested among economists, and honest summaries say so.
• Quantitative tightening reverses it, usually by letting bonds mature rather than selling. It is slower, less predictable in its market impact, and interacts with government issuance in ways that showed up in the 2022–23 bond markets.

Forward guidance and the credibility problem

• Because expectations are a transmission channel, telling markets what you intend to do is policy. Guidance moved long rates without any rate change at all.
• Its cost is optionality. Committing to a path constrains a committee that may need to change course, and breaking guidance damages the credibility that made it work.
• Several central banks moved from calendar-based guidance ("rates on hold until 2023") to state-based ("until inflation is durably at target") and then to explicit data-dependence — a retreat driven by exactly this trade-off after 2021.
• Credibility is the actual asset. An institution believed to be committed to its target gets tighter financial conditions from smaller moves; one that is not must move further to achieve the same result.

Practitioner rules

• Read the path, not the decision. A rate move already priced changes nothing; the surprise is in the projection and the language.
• Separate the two lags. Asset prices respond on the announcement, the economy in a year. Most disagreement about whether policy is "working" is really disagreement about which lag is being discussed.
• Forwards for hedging, not forecasting. Trade at the forward; do not believe it.
• Watch real rates, not nominal ones. A 5% policy rate with 6% inflation is accommodative. The stance is the gap to neutral, which is why the inflation page and this one are the same subject seen from two sides.
• Two of the key inputs are unobservable. Anyone stating the neutral rate or the output gap without a confidence interval is stating an estimate as a fact.