What Central Banks Actually Control

Central banks hold a unique position. They issue central-bank money, operate the settlement infrastructure of the monetary system, influence the cost of short-term funding, provide liquidity to banks and communicate the expected path of policy. Through these channels they influence financial conditions across the economy. One of the most common misunderstandings, however, is to assume that a central bank directly controls everything that happens to money, credit, interest rates and asset prices. It does not.

Part I built the architecture: the economy as a network of balance sheets (Lesson 1), money (Lesson 2), credit (Lesson 3), banks (Lesson 4) and sectoral balance sheets (Lesson 5). Part II adds the institution at the top of that system and asks how it changes the conditions under which every balance sheet operates.

1. Balance sheets
   ↓
2. Money
   ↓
3. Credit
   ↓
4. Banks
   ↓
5. Sectoral balance sheets
   ↓
6. Central-bank policy

The monetary system is not a machine in which a central bank presses a button and the economy responds. It is a transmission system. The central bank changes the conditions under which the financial system operates; institutions and markets respond; households and firms adjust; asset prices, exchange rates and credit conditions move; and the effects eventually reach spending, investment, employment and inflation. The question for this lesson is what a modern central bank actually controls, and how that control travels through the system.

A central bank’s balance sheet differs from that of a commercial bank. Its assets typically include government securities, foreign-exchange reserves, loans to financial institutions and other financial assets; its liabilities include currency in circulation and commercial-bank reserves. It occupies the highest monetary layer of the domestic banking system, because banks settle payments among themselves in central-bank money. That position is special, but it does not mean that the central bank determines every monetary or financial variable.

A useful way to organise monetary policy is to sort variables into three groups, under a conventional operating framework.

The third group responds to monetary policy but is determined by a much larger system of markets, households, firms, banks and governments. This distinction is fundamental.

The most visible conventional instrument is the policy interest rate. Depending on the institutional framework it may be an overnight target rate, a deposit facility rate, a refinancing rate, a target range or another administered short-term rate. Terminology differs; the principle is the same. The central bank sets, directly or indirectly, the price at which very short-term money-market transactions take place, and that price is the foundation of a much broader pricing system.

Two policy settings as of early November 2026 show the different conventions. The ECB steers policy through the deposit facility rate, with two further administered rates around it. The Federal Reserve sets a target range for the federal funds rate.

The policy rate is only the starting point. The important question is how a change in it reaches the rest of the economy, which is the problem of monetary transmission.

A central bank can set or guide its own rate. It does not announce the other rates in the economy. A ten-year government yield, a corporate bond yield, a mortgage rate and an equity valuation each depend on a different mix of factors, of which the policy rate is only one.

The policy rate is therefore an anchor; the financial system builds the rest of the price structure around it.

Central banks have their strongest influence at the short end, because they control or administer the conditions under which banks settle payments and obtain or place central-bank liquidity. When the policy rate rises, money-market rates generally follow, banks face a different funding environment, short-term government securities and derivatives reprice, and the effect spreads to floating-rate loans, money-market instruments, short-duration bonds and corporate financing. The farther along the curve, the greater the role of expectations and markets.

A ten-year rate is not today’s policy rate. As a simplification, a long-term rate can be thought of as the expected path of short-term rates plus a term premium. If markets expect policy rates to stay high, long yields can rise; if they expect rapid easing, long yields can fall. The relationship is not mechanical: a central bank can cut while long yields rise, if investors expect higher inflation, larger deficits, a higher term premium, resilient growth or eventually higher policy rates. The statement “the central bank cut, so yields must fall” is therefore not necessarily true. Markets price the future, not only the present decision.

Commercial banks hold accounts at the central bank, and the balances are central-bank reserves. They are an asset of the commercial bank and a liability of the central bank, used mainly to settle payments between institutions and to meet certain liquidity and regulatory needs. As Lesson 2 and Lesson 4 stressed, reserves are not household deposits: a household holds a deposit at a commercial bank, and a commercial bank holds reserves at the central bank. They are different layers of the monetary system.

Whether a central bank controls the quantity of reserves depends on its operating framework. Historically many central banks operated with scarce reserves, so that the quantity played a direct role in keeping the overnight rate near target. In a modern abundant-reserves or floor-type system the central bank steers short-term rates mainly through administered rates, including the remuneration of reserves and standing facilities. The equation “more reserves means easier policy” is therefore not universally valid.

The Federal Reserve’s own data illustrate the point. In the week ended 21 October 2026, reserve balances at the Federal Reserve Banks averaged about $2.96 trillion, around 44% of the $6.81 trillion of total factors supplying reserve funds. Policy had been tightened in September through the administered rates, with reserves still abundant. After large-scale asset purchases, reserves can rise enormously without a proportional rise in household spending or bank lending, one of the main lessons of the monetary system after the global financial crisis.

Consider a central bank buying a government bond from a commercial bank. The central bank records the bond as an asset and creates the corresponding reserve liability. The commercial bank gives up the bond and receives reserves.

The reserves do not become cash in the hands of households. The operation changes the composition of financial assets in the banking system, and may affect yields, liquidity, portfolio allocation, risk premia and expectations. It does not mechanically create an equivalent amount of household spending. As Lesson 2 noted, when the seller is a non-bank investor, the seller’s deposit also rises, but even then the central bank does not credit households at large.

Central banks are providers of liquidity: the ability to meet payment and funding obligations without selling assets at severely depressed prices. A bank can be solvent but illiquid, owning valuable assets while lacking immediately available funds. The central bank can provide liquidity against eligible collateral, which is one reason central banks exist as lenders of last resort. Liquidity support does not, however, solve insolvency, a distinction introduced in Lesson 4 and important here because it defines what monetary policy can and cannot do.

One of the most powerful tools of modern monetary policy is not a balance-sheet item. Financial markets price the future, and households and firms also decide on the basis of expectations about inflation, rates, income, growth, asset prices and financing conditions. If a central bank signals that rates will stay restrictive for an extended period, bond yields, the currency, equity valuations, mortgage rates and corporate borrowing costs can move before the policy rate does. The central bank has influenced financial conditions through communication, which is why modern monetary policy is partly an exercise in expectations management.

Forward guidance is the communication of the likely future path of policy: an expected path, the conditions under which it might change, the central bank’s reaction function, its assessment of inflation risks or the likely duration of restriction. It has a built-in limitation. Markets update continuously, and if inflation surprises, expectations can be revised at once. Central-bank communication influences expectations but does not dictate them.

Markets also try to infer the reaction function, how the central bank will respond to inflation, activity, employment, financial stability and, in some jurisdictions, the exchange rate. They ask what would make it tighten and what would make it ease, and the answer is priced today. This creates a feedback loop: data shape expectations of policy, expectations shape asset prices and financial conditions, those shape behaviour, and behaviour generates new data.

Financial conditions describe the overall ease or tightness of financing in the economy: short-term rates, government yields, corporate spreads, mortgage rates, equity prices, the exchange rate, credit availability, lending standards, volatility and liquidity. A central bank does not set all of these, but its policy influences them collectively. It can be restrictive without moving its rate, and easy without cutting.

The stance of monetary policy cannot always be inferred from the policy rate alone. The ECB’s meeting of 29 October 2026 offered a current example: it left its three key rates unchanged while noting, according to its statement, that financing conditions had tightened since its previous meeting, a period in which euro-area government bond yields had risen. Policy rates were constant; conditions were not.

The transmission can be summarised as a chain from the central bank through policy rates, liquidity and communication to short-term rates, the yield curve and expectations, to bond yields, credit spreads, equities and exchange rates, then to financial conditions, borrowing, saving, spending and investment, aggregate demand, and finally inflation and activity. The sequence is not mechanical, the channels interact and can pull in different directions, and their strength varies over time. Six channels are worth distinguishing.

On the asset-price channel, a simple valuation relationship is that the value of an asset is approximately its future cash flows divided by a discount rate, so a higher discount rate lowers present value, all else equal. The longer the duration, the greater the sensitivity. A long fixed-rate bond falls in price when market yields rise so that its yield becomes competitive with new issues, and the same logic applies to growth equities, real estate and private assets whose value depends on distant cash flows. The effect is not identical across assets, because expected earnings, risk premia, inflation expectations and fiscal policy can change at the same time. How these effects play out is the subject of Part II’s next lessons.

A 100-basis-point increase does not have the same effect in every economy, because the financial structure differs, which connects directly to Part I.

This is why central banks monitor the structure of the financial system as well as inflation and employment. Policy does not need to control every variable. It changes the pricing environment in which the balance-sheet network of Part I operates: debt-service costs, asset valuations, bank profitability, bond prices, exchange rates, corporate investment, household saving, government financing costs and capital flows.

Central banks can also act through the balance sheet. Under quantitative easing a central bank buys financial assets, mostly government bonds. The operation can increase reserves, change the maturity composition of the assets held by the private sector, reduce the supply of certain securities available to investors, influence term premia, affect market liquidity and signal a stance. QE is not accurately described as the central bank printing money and giving it to households. The central bank changes its own balance sheet and the composition of assets in the private sector, and the effects on prices and spending run through financial-market and portfolio channels. The financial asset has not disappeared: it has moved between balance sheets, from the private sector to the central bank, with reserves in exchange.

Quantitative tightening is the reverse: the balance sheet shrinks through maturing securities that are not reinvested, asset sales or reduced lending. The effect is again not simply “fewer reserves means less money means less spending”. What matters is how the adjustment changes market liquidity, term premia, asset supply, bank funding, expectations and broader financial conditions. Lesson 7 examines why central banks moved to zero rates and to QE, and Lesson 10 examines the reversal.

The limits of monetary policy matter as much as its reach. A central bank cannot directly control:

• Long-run growth: it depends on technology, labour, capital, institutions, demographics and innovation.
• Fiscal policy: governments set taxes and spending within their institutions.
• Private credit demand: households and firms decide whether to borrow.
• Global capital flows: international investors reallocate portfolios continuously.
• Commodity prices: they depend on global supply and demand, geopolitics and physical constraints.
• Exchange rates and asset prices: it influences them, but markets set them from expectations of cash flows, risk and policy.
• Inflation: it can influence inflation conditions, but inflation is the outcome of demand, supply, expectations, wages, commodities, fiscal policy and monetary conditions together.

These limits prevent an overly mechanical view of policy, and they matter in the current environment, in which both the ECB and the Federal Reserve have raised rates this year while an energy shock, which no central bank controls, continues to work through prices.

If central banks control so little directly, why are they so powerful? Because they anchor the monetary base, short-term interest rates, the settlement system, liquidity provision and monetary expectations, and because financial markets price assets relative to expected future short-term rates. The central bank is not the whole financial system; it is the anchor around which much of the system is priced.

Modern capitalism is decentralised. Millions of households, firms, banks and investors take independent decisions and markets aggregate them into prices. The central bank does not replace that process. It sets the short-term monetary environment; markets set the price of longer-term claims; banks decide how much credit to extend; households and firms decide how much to borrow; investors allocate portfolios; governments set fiscal policy; foreign investors determine capital flows. The macroeconomic outcome emerges from all of these together, which is why monetary policy is best understood as a system of incentives, prices and expectations and not as a command mechanism.

It also suggests a deeper reading. Monetary policy is not only about the quantity of money. It sets and communicates the price of short-term central-bank money and influences the wider financial pricing system: the opportunity cost of holding money, the cost of borrowing, the return on short-term assets, expectations of future rates, discount rates and financial conditions. In that sense the central bank influences the price of time, since an interest rate is the price assigned to transferring purchasing power across time. A central bank does not control the economy. It controls, or strongly influences, the monetary conditions under which the economy operates.

When analysing monetary policy it helps to keep seven variables distinct. They are related but not identical, and the usual analytical mistake is to treat them as one.

A practical framework for analysing any central bank then asks five questions, so that the analysis does not stop at the policy-rate announcement.

• Ben Bernanke, 21st Century Monetary Policy, in the Core Canon, and Ulrich Bindseil and Alessio Fotia, Introduction to Central Banking, in the Reading Room.
• The central-bank publications listed under Institutional Reading.