A quarter of a percentage point is a small number. When a central bank changes one, long bonds can lose several percent of their value in an afternoon, currency pairs move a hundred pips in minutes, and the discount rate applied to every future corporate profit on earth changes at the same time.
The interest rate a central bank sets is the price of money at the shortest maturity there is. Everything priced in that currency — bonds, equities, property, the currency itself — is priced off it, directly or at one remove. That is why a technical adjustment to an overnight rate reaches into markets that appear to have nothing to do with lending.
What the policy rate actually is
A central bank does not set mortgage rates, corporate loan rates or savings rates. It sets one very short-term rate — in the United States a target range for the rate at which banks lend reserves to each other overnight — and lets everything else be priced relative to it.
That single anchor works because banks fund themselves at or near it. Change what a bank pays for overnight money and you change what it charges everyone else. The mechanism is arbitrage, not instruction.
The other half of the picture is expectations. A ten-year yield is not driven by today's overnight rate; it is roughly the average overnight rate the market expects over the next ten years, plus a premium for uncertainty. This is why the long end can fall on the day a central bank hikes: the hike was expected, and the guidance implied fewer hikes to come.
The transmission mechanism, stage by stage
Economists call the path from policy rate to inflation the transmission mechanism. It runs in stages, and each stage takes longer than the last.
Stage one, immediately. Money market rates reprice within hours. Bank funding costs move, and floating-rate loans tied to a reference rate reset at their next date.
Stage two, one to six months. New borrowing gets more expensive. Companies postpone capital spending and slow hiring. Households delay large purchases financed with credit. Existing fixed-rate borrowers are unaffected until they refinance, which is why the effect is gradual rather than a step change.
Stage three, six to eighteen months. Weaker demand starts to show in prices. Firms lose pricing power, discounting returns, and measured inflation falls.
The lag is the hard part. A rate set today has most of its effect on inflation a year to two years from now, and the central bank has to act on a forecast of conditions it cannot yet observe. That is the structural reason central banks overshoot in both directions: they tighten until something breaks, then ease until inflation returns.
Nominal is not real
One distinction separates people who follow rates from people who understand them. A 5% policy rate with 6% inflation is looser than a 2% rate with zero inflation. What matters for behavior is the real rate — the nominal rate minus expected inflation.
Borrowers care about real rates because inflation erodes the debt they owe. Savers care because it erodes what they hold. Central banks talk about nominal rates because that is what they set, but their target is the real rate relative to the neutral level: the rate that neither stimulates nor restrains, usually written r*.
Nobody can observe r*. It is estimated, revised, and argued about, and the arguments move markets, because a rate of 4% is restrictive if neutral is 2.5% and accommodative if neutral is 4.5%. When policymakers debate "how restrictive are we," this is what they are debating.
Who sets the rates that matter
Eight central banks account for almost all the currency volume a retail trader will ever touch, and each has a personality that comes from its mandate and its economy rather than from whoever is running it this decade.
- Federal Reserve (USD). Dual mandate, prices and employment. Sets the price of the world's reserve currency, so its decisions are a global input rather than an American one.
- European Central Bank (EUR). Price stability as the primary objective, one policy for a monetary union with no shared budget. Structurally slower to ease once inflation has been a problem.
- Bank of England (GBP). Inflation target with a published letter to the Chancellor when it is missed by more than a point. Sterling is unusually sensitive to domestic fiscal news alongside rates.
- Bank of Japan (JPY). Spent decades fighting deflation with rates at or below zero and yield curve control. The yen is the classic funding currency, which makes it strengthen violently when carry trades unwind.
- Swiss National Bank (CHF). Willing to intervene directly in the currency market to stop the franc appreciating. A safe haven with an owner that actively dislikes safe haven flows.
- Bank of Canada (CAD), Reserve Bank of Australia (AUD), Reserve Bank of New Zealand (NZD). Commodity-linked economies, so their currencies carry a growth and commodity beta on top of the rate story.
Knowing the mandate tells you how a bank will react to a given shock. A supply-driven inflation spike puts the ECB in a corner and gives the Fed room to weigh the employment side. That asymmetry is a tradeable difference between two currencies facing the same event.
Bonds: an arithmetic relationship
Bond prices and yields move in opposite directions, and this is not a tendency — it is arithmetic.
Buy a ten-year Treasury paying a 3% coupon at par. Yields then rise and new ten-year notes are issued at 4%. Nobody will pay par for your 3% bond when a 4% one is available, so the price falls until the total return to maturity matches the market. For a ten-year bond that repricing is roughly 8%.
Duration measures this sensitivity. As a working approximation, a bond loses its duration in percent for every one percentage point rise in yields:
- Duration 2: a 1% yield rise costs about 2% of price.
- Duration 7: about 7%.
- Duration 20: about 20%.
The 2022 tightening cycle demonstrated this at scale. Long-dated government bonds — the assets most retail portfolios classify as "the safe part" — lost 20% to 30% of their value. Nothing defaulted. The math simply did what it always does, and holders discovered that credit risk and interest rate risk are not the same thing.
The same math is what killed Silicon Valley Bank in 2023. The bank had not made bad loans; it had bought long bonds with short deposits and the yield curve did the rest.
Equities: three separate channels
The link between rates and stocks is looser than the bond relationship but still traceable, and it works through three channels that can pull in different directions.
Competition. When short-term government paper yields 5% with no credit risk, the bar for holding equities rises. Money leaves risk assets not because companies got worse but because the alternative got better.
Cost of capital. Companies that carry debt pay more on it. Refinancing at higher rates transfers profit from shareholders to lenders.
Discounting. A share is worth the present value of future cash flows, and a higher discount rate shrinks distant cash flows more than near ones. This is why growth names, whose value sits years out, fall harder than mature cash generators when yields rise. Duration is not only a bond concept.
Sector behavior follows from the channels. Real estate investment trusts, utilities and heavily indebted growth companies tend to underperform in a tightening cycle. Large banks can benefit, at least initially, because lending margins widen faster than deposit costs. Energy often holds up, since the inflation that provoked the tightening tends to lift its revenue.
Then there is the reflex the market calls the Fed put: the learned expectation that a central bank will ease when markets fall far enough. Whether or not it is real, the belief in it is, and it explains a pattern that confuses newcomers — equities frequently bottom and start rallying before the first cut, pricing the pivot rather than the policy.
Currencies: the differential, not the level
Foreign exchange cares less about the level of any single rate than about the gap between two of them. Capital flows toward higher risk-adjusted yield, and the exchange rate is the price that clears the flow.
The clearest expression is the carry trade. Borrow in a low-yielding currency, convert, and hold assets in a high-yielding one. The interest differential is the return. The risk is entirely in the exchange rate: the position earns steadily and loses violently, because the funding currency tends to appreciate in exactly the risk-off episodes when everyone unwinds at once.
The practical rules are simple to state and easy to misapply:
- A central bank tightening faster than its peers tends to support its currency.
- A widening yield spread in favor of one currency tends to pull the pair that way.
- Emerging market currencies come under pressure when developed-market rates rise, because capital leaves and their dollar debts get heavier.
The reason these rules disappoint in practice is that markets price expectations, not levels. By the time a hiking cycle is obvious, the currency has already moved. What moves it next is the change in expected path, which is why a hike delivered alongside dovish guidance can weaken a currency on the day it arrives. We look at this in detail in how Fed and ECB divergence moves EUR/USD.
Property: the most levered rate bet most people own
Housing responds to rates more mechanically than any other asset, because most buyers borrow and the loan payment is what constrains them.
Take a $300,000 loan over thirty years. At 3% the monthly payment is about $1,265. At 6% it is about $1,799 — a 42% increase on exactly the same house. Nothing about the property changed. The buyer who qualified at 3% simply cannot afford it at 6%.
That is why transaction volumes collapse quickly when rates rise while prices fall slowly. Sellers who financed at low rates have no incentive to move, supply dries up, and the market clears through fewer sales rather than lower prices. Property adjusts on volume first and price much later.
Where you are in the cycle
Rate cycles rhyme. They are not a schedule, and anyone selling a calendar is guessing, but the phases have recognizable characteristics.
Low and stable rates reward duration in every form: growth equities, long bonds bought earlier, real assets, and speculative assets generally. Cash earns nothing, which is the point.
The first hikes are usually absorbed calmly. The economy is strong, earnings are rising, and the tightening is read as confirmation of health.
Aggressive tightening is where correlations break. Bonds and equities can fall together — 2022 was the clearest example in decades — because both are being repriced by the same discount rate rather than by their own fundamentals. Diversification built on the assumption that bonds hedge equities stops working exactly when it is needed.
The peak is only identifiable afterwards. It is also the point at which long bonds offer their best entry yield in the cycle, which is why the market spends so much energy trying to call it early.
The first cuts are ambiguous, and this is the most misread moment in the whole sequence. A cut into a healthy economy is a tailwind for risk assets. A cut into a collapsing one is a warning, and equities usually fall anyway. The same action means opposite things depending on why it is being taken.
How to follow it without guessing
The useful discipline is not predicting decisions. It is knowing what the market has already priced, so you can recognize a surprise when one arrives.
Interest rate futures show the probability the market assigns to each outcome at each upcoming meeting. This is the benchmark: a hike priced at 95% is not news when it happens.
The yield curve aggregates the expected path. A flattening curve says the market expects tightening to end; an inverted one — short yields above long — has preceded most recessions, with a lead time too variable to trade directly.
The data that drives the decisions matters more than the decisions themselves. For the United States that means the monthly payrolls report, the CPI release and the PCE deflator the Fed actually targets. All of them, with consensus and previous readings, sit on our economic calendar.
The language. Hawkish and dovish are not moods, they are forecasts about the path. "Inflation remains elevated" and "risks to inflation are skewed to the upside" are hawkish. "The disinflationary process is well under way" and "risks are broadly balanced" are dovish. Traders read a new statement against the previous one, word by word, because the deletion of a single qualifier is a signal.
Five mistakes that cost money
Trading the decision instead of the surprise. If the market expected 25 basis points and got 25 basis points, the number is already in the price. What moves is the gap between expectation and outcome, and the gap in the guidance.
Ignoring the lag. Tightening today affects the economy in a year or more. Reading current data as the verdict on current policy gets the timing wrong in both directions.
Assuming every cut is bullish. Context decides. Insurance cuts and emergency cuts look identical on a chart of the policy rate and mean opposite things.
Looking at one central bank. Currencies are relative prices. A hawkish Fed against a hawkish ECB is not the same trade as a hawkish Fed against a passive Bank of Japan, even though the Fed did the same thing.
Confusing safe from default with safe from loss. A government bond will pay you back. It will not protect you from a rate cycle in the meantime, as 2022 reminded a generation of investors who had never seen one.
The rate is the frame
Interest rates are the one macro variable that touches every asset a trader can hold. They set the cost of leverage, the return on doing nothing, the discount rate on the future, and the relative attractiveness of one currency over another.
Central banks move slowly and telegraph their intentions on purpose, which means the work is interpretation rather than prediction. Knowing where the cycle stands, what the curve already assumes, and how far reality would have to move before the assumption breaks is worth more than any forecast of the next decision.
