How the Federal Reserve Actually Changes Interest Rates

When the Federal Reserve raises rates, nothing about that announcement compels a single bank to charge more. The Fed does not set the rate on your mortgage, your savings account or a corporate loan. It sets a target for one narrow rate, and then makes that target true through its own operations.

Understanding the difference between announcing and enforcing explains a lot about why policy works with a lag, and why it sometimes does not work as expected at all.

The rate being targeted

The federal funds rate is what banks charge each other for overnight loans of reserves — the balances they hold at the Fed. It is an unusually specific thing for the most-watched number in economics to be.

It matters because it anchors the short end of the curve. Almost every other short-term rate is priced relative to it, and longer rates embed expectations about where it will go. Move the anchor and the structure above it shifts.

How the target is enforced now

Before 2008 the Fed kept reserves scarce and adjusted the supply in small amounts to nudge the funds rate. That method depended on scarcity, and quantitative easing ended scarcity by flooding the system with reserves.

The current framework works differently. The Fed pays interest on reserve balances, which sets a floor: no bank will lend to another below what it can earn risk-free at the Fed. A second facility, available to a wider set of institutions, catches lenders who cannot hold reserves directly. A standing repo facility caps the upside by lending against Treasuries at a fixed rate.

The result is a corridor. Rather than manipulating quantity, the Fed sets the prices at which it will borrow and lend, and the market rate settles between them. This is why the Fed can now change rates without changing the size of its balance sheet.

Why the distinction matters

It means rate policy and balance-sheet policy are separate levers. The Fed can raise rates while still holding a large portfolio, or shrink the portfolio while holding rates steady. Commentary that treats these as the same thing usually gets the reasoning wrong.

Transmission: the slow part

A change in the overnight rate reaches the real economy indirectly, and unevenly.

Each channel operates on its own schedule. The conventional estimate is that most of the effect arrives over roughly a year to eighteen months, but that range is wide and depends on the structure of debt in the economy at the time. In a country where most mortgages are fixed for thirty years, rate changes bite far more slowly than where they reset annually.

The forecasting problem

Because the lag is long, policy has to be set on a forecast. The Fed is responding not to today's inflation but to what it expects inflation to be once today's decision has worked through.

This is genuinely hard, and it is worth being honest about how hard. Forecasts are revised. Data arrives late and is revised too — initial employment estimates routinely move by amounts that would have changed the interpretation. A central bank steering by a forecast built on provisional data is doing something closer to navigation in fog than to control engineering.

The dual mandate

US law directs the Fed to pursue maximum employment and stable prices. Most of the time these point the same way. Occasionally they conflict, and the conflict is the whole difficulty of the job: an economy with rising prices and weakening employment offers no policy that improves both.

The mandate does not rank the two, which leaves the trade-off to judgment. That is why individual policymakers can look at identical data and reach different conclusions without either being unreasonable.

The balance sheet as a second instrument

Alongside the policy rate, the Fed holds a portfolio of Treasuries and mortgage-backed securities. Changing its size is a separate tool with a separate transmission mechanism.

Buying long-dated assets pushes their prices up and yields down, which compresses long-term borrowing costs even when the short-term policy rate is already near zero. That was the rationale for quantitative easing: with the conventional instrument exhausted, the central bank moved further along the curve.

Reducing the portfolio works in reverse and is usually done passively — allowing bonds to mature without reinvesting the proceeds. This is slower and less disruptive than selling, though it means the pace is determined by the maturity profile rather than chosen.

How much effect balance sheet policy has is genuinely disputed among economists, and the honest summary is that estimates vary widely and depend on assumptions that cannot be tested directly.

Forward guidance, and the credibility problem

Because long rates reflect expected future short rates, a central bank can influence long rates today by shaping expectations about tomorrow. Saying rates will stay low for a considerable period can lower long yields immediately.

The catch is that guidance is only as good as the belief that it will be honoured. A central bank that signals one path and takes another finds its next signal discounted. This is why guidance has grown more conditional over time — tied to economic conditions rather than to dates, which preserves flexibility at the cost of some precision.

It also explains the care taken with language. A word changed in a statement is not stylistic; it is the instrument being adjusted.

What to watch beyond the decision

The rate decision itself is usually well anticipated. More information sits in the accompanying material: the projections showing where officials expect rates to go, the vote split, and the changes in wording from the previous statement.

Markets frequently move more on the projections than on the decision, because the decision was priced in and the projections were not. A meeting where the rate is unchanged can be the most consequential of the year.