How the Fed sets rates, actually.

It is a one-line formula John Taylor published in 1993 that tells you, given inflation and the unemployment rate, roughly what the Fed funds target should be. The rule takes the neutral rate, adds half the distance inflation is running above 2%, and adds half the distance unemployment is running above its own 2% reference. You get a number. That number is the rule's opinion on what the Fed ought to be doing right now. The FOMC is not legally required to follow it and rarely does exactly, but every economist on every central bank staff has the output on a chart, because it is the cleanest way to say 'policy is tight' or 'policy is loose' without arguing about it.

Twelve people sit in a room eight times a year and vote. That is the literal answer. The structured answer is that the FOMC stares at incoming data on inflation (core PCE), the labor market (U-3 and payrolls), and financial conditions (credit spreads, equity multiples, the dollar). They run that through a set of staff models, the Taylor Rule being one of them, and they triangulate. Then they ask what the market is already pricing, because surprising the market is itself a policy action. The decision is announced at 2pm ET, followed by a statement, followed thirty minutes later by the chair's press conference. The press conference usually moves markets more than the decision itself, because that is where forward guidance happens.

The neutral rate, which economists write as r-star, is the Fed funds rate at which policy is neither stimulating nor restricting growth. It is the setting that, in the long run, keeps inflation at 2% and unemployment near its natural rate. Nobody knows exactly where it is. The Fed's own median estimate has drifted from around 4% before the financial crisis to 2.5% today, and back up in the post-pandemic rebuild. The uncertainty matters because a Fed funds rate of 4% is restrictive if neutral is 2.5% and roughly neutral if r-star has actually moved to 4%. A lot of the argument on financial TV is really just an argument about where r-star lives.

Forward guidance is the Fed telling you what it plans to do before it does it. The idea is that the rate set today matters less than the rate you expect next year, because that is the rate that gets baked into every 10-year Treasury yield, every 30-year mortgage, and every discount rate in every DCF on the street. So the Fed telegraphs. It releases the dot plot, a scatter of individual officials' projections. It uses specific words. 'Patient' means on hold. 'Data-dependent' means no preset path. 'Meaningful further progress' means we need one more good print. The bond market treats these phrases as instructions, and the Fed knows it, which is why every syllable is negotiated over for weeks.

Through the discount rate, mostly. Every asset with future cash flows, which is to say every stock, every bond, every house, every VC round, is priced by discounting those cash flows back at some rate anchored to Treasuries. When the Fed hikes, that rate goes up, and the present value of every long-dated asset goes down. Long-duration tech gets hit hardest because its cash flows live furthest out. Short-duration value holds up better. In fixed income the bond you already own loses price and the bond you haven't bought yet offers more yield. In credit, spreads widen because refinancing gets expensive and defaults climb. Your 30-year mortgage tracks the 10-year Treasury, not the Fed funds rate directly, but the 10-year cares about where Fed funds is going, so the mortgage cares too. The whole system is a single discount rate problem wearing different costumes.