Reading the FOMC: Statement, Projections and Press Conference

Three documents, in the order they arrive

The statement is short and every word is deliberate: the target range, the characterisation of activity and employment, the sentence about the balance of risks, and any forward language about the path. It is a negotiated document produced by a committee, so a change in a phrase is a change in the consensus rather than a stylistic preference — reading two consecutive statements side by side and marking what changed is the whole skill. The most informative changes are usually about the labour market and about the balance of risks, because they are the inputs to the reaction function that the committee can describe without committing to a path. The projections, published quarterly, add three forecasts and a distribution: growth, unemployment and inflation, plus each participant’s view of the appropriate policy rate, summarised as the median path. The distributions matter more than the medians, because a wide spread on the appropriate rate means the committee itself is uncertain about the reaction function — which is the same point the reaction-function lesson made about the unobservable neutral rate. And the median path is not a promise: it is the rate that is appropriate given each participant’s own forecast, so if the forecast is wrong the path changes, which is why markets trade the gap rather than the dot. The press conference is where the committee explains, and it is where the largest reactions frequently happen, because an answer to a question about the criteria for cutting is more informative than the statement. The risk for a reader is that it is unscripted: a phrase in an answer carries less weight than a phrase in the statement, but a consistent theme across several answers is more informative than either. Watch the questions about the balance of risks, about the definition of restrictive, and about any tension between the projections and the statement. The same decision, three sources of information — The decision: as expected: no information; it is repriced through the futures contracts, not the announcement · The statement: one clause changed: the statement is negotiated, so a change is the consensus moving ← · The projections: median cuts two → one: the committee’s modal path, against which the market had two priced · The press conference: the criteria for future action, in far more detail than the statement can carry The dot plot is a set of individual forecasts under individual forecasts, not a plan. Treating the median as a commitment — or as a surprise when it changes — is the most common way to misread an announcement day, and it is why the second-order question is always what each participant assumed to reach their dot.

What the market is doing, in two parts

The expected path is what moves first, and it is visible in short-dated futures: the front end of the curve reprices within seconds of the statement, because a rate path is a set of expectations about the next few meetings. The term premium is the second part, and it moves for different reasons — the supply of duration, the amount of uncertainty in the policy outlook, and demand for the safety of government paper. Separating the two is what allows the apparently paradoxical outcomes: a hawkish decision can lower long yields if it resolves uncertainty and compresses the premium, and a dovish one can raise long yields if it suggests the committee expects a harder landing. Two other markets are worth watching because they carry the interpretation rather than the policy. Break-even inflation is the traded expectation of inflation; if it rises after a hawkish statement, the market is saying the committee will be behind the curve, which is a different statement from the one the committee intended. And the dollar is the relative-policy channel: a hawkish surprise strengthens it, which imports disinflation and is itself part of the transmission, which is why the committee’s reaction is conditioned on the currency doing some of the work. The practical discipline is to decide in advance what would change your view. Write down the two or three sentences whose change would matter — the balance of risks, the labour market assessment, and any reference to the restrictiveness of policy — and the data points in the projections you would compare against the priced path. Then the announcement day becomes a test of your expectations rather than a stream of noise, and the reaction you observe can be attributed to something specific rather than to sentiment. • The decision is priced; the statement, the projections and the press conference carry the information. • Read two statements side by side and mark what changed — the document is negotiated. • The front end reprices on the path, the long end on the path and the premium. • Break-evens and the dollar carry the interpretation, which can contradict the committee’s intent. A hawkish decision with a lower long yield is not a contradiction: it is the path rising and the premium falling, and knowing which moved is the difference between reading the market and narrating it.

The dots are not a forecast, and the statement is negotiated

The projections published four times a year contain a chart that markets treat as a policy commitment and that is not one. The so-called **dot plot** shows each participant’s assessment of where the policy rate should be at the end of the next few years, and the median of those dots is the number quoted in every headline. Three properties limit what it means. Each dot is **anonymous**, so the median can move without anyone changing their view — one participant moving is enough if the distribution is tight. Each dot is **conditional on that participant’s own forecast** for growth and inflation, so it is not a promise about what the committee will do if the economy behaves differently. And the median is a summary of a distribution whose width matters: when the dots are tightly clustered the median reads as a plan, and when they are spread across a percentage point the same median is an average of disagreement and the market is wrong to treat it as guidance. The second document is shorter and more consequential than it appears. The statement is a **negotiated text**, revised line by line by the committee, and this is why a change of a single word — from “modest” to “moderate”, from “further” to “some” — is the thing professional readers look for. The comparison that matters is with the previous statement, not with the economy: the market has already read the last one, so the information is entirely in the difference. Reading a statement without the previous statement beside it is reading a document with the meaning removed. Two structural features of the day are easy to overlook and both matter. The first is that **the balance sheet is a second instrument**, and a committee can tighten by letting securities run off while holding the policy rate, or do the reverse. Announcements about the pace of that runoff are policy, not plumbing. The second is the **blackout period** in the days before the meeting, during which officials do not speak publicly. Its value is that the final week’s silence means the market’s positioning going into the meeting reflects the last thing officials were willing to say, which is why the pre-meeting drift and the post-meeting reaction are studied as a pair rather than separately. • The dot plot is anonymous, conditional on each participant’s own forecast, and its dispersion matters as much as its median. • The statement is a negotiated document — read it against the previous statement, not against the economy. • The balance sheet is a second instrument; runoff pace announcements are policy decisions. • The blackout period makes pre-meeting positioning a reading of what officials last said.

The vote, the minutes and the facilities underneath

The statement is the committee’s negotiated voice, and the vote is where the negotiation leaves a trace. The policy decision is taken by a fixed-size committee whose voting membership rotates, and the published vote records not just the number of dissenters but who they are and in which direction they dissented — a dissent for a larger increase is a different statement from a dissent for no change, and a dissent from a regional reserve bank president is a different signal from one by a governor, because the governors sit at every meeting and the presidents rotate. The count is usually reported as a headline number, and the useful reading is the composition: a lone dissenter in the direction of more tightening is a pocket of hawkishness, while three dissents in opposite directions is a committee that cannot agree on the risk, which is a statement about the level of uncertainty rather than about the likely path. Since the vote is published with the statement and the projections arrive at the same moment, the vote is one of the few same-day signals that is not already priced. The **minutes** are the second document and they are structurally different from the statement in a way worth understanding. They are published three weeks later, they are written by staff rather than approved sentence by sentence, and their purpose is to record the discussion — so they contain the alternatives that were considered and rejected, the staff’s own economic assessment, and the committee’s view of the risks around the forecast. That makes them the best available evidence of the reaction function, and it also makes them stale: a market that moves three weeks after the meeting because the minutes revealed “some participants” were worried about something is reacting to a description of a meeting that has already been superseded by subsequent data. The professional use is not to trade the minutes but to check a thesis: if the minutes show the committee debated a specific risk in detail, that risk is inside the framework, and a later data print that speaks to it will be interpreted through that framework. Underneath both documents sits the **operational framework**, which is where the plumbing decisions live and which is announced as often and as consequentially as the policy stance. The modern operating system targets the policy rate through the price the central bank pays on reserve balances, with ample reserves in the system, so the rate is managed by administrative prices rather than by active daily open-market operations of the traditional kind — and that means decisions about the balance sheet, about the pace at which holdings run off, and about the standing facilities that backstop short-term funding are policy. When a standing repo facility is created or its terms are widened, the central bank is changing the price of the last-resort liquidity that underpins the whole money market; that shows up in the spread between overnight rates and the policy rate before it shows up anywhere else. The practical reading order for the day is therefore: the decision is priced, the statement and the projections carry the forward guidance, the press conference carries the criteria, and the vote and the operational notices carry the direction of travel — in that order of immediacy, and from the shortest horizon to the longest. • Read the vote by direction and by who dissented, not by the headline count. • The minutes record the debate and arrive three weeks late — they check a thesis rather than trade a print. • The operating system uses administered prices on ample reserves, so balance-sheet and facility decisions are policy. • A change at a standing facility moves short-term funding spreads before it moves the policy rate. • Order of immediacy: decision, statement, projections, press conference, vote, operational notices. The discipline that ties all of it together: most of what is published on the day is priced within minutes, so the value of reading the documents is not the same-day move. It is knowing what the committee will do with the next data point, which is a question only the language and the framework can answer.

Sources

• FOMC statements, the Summary of Economic Projections and the dot plotBoard of Governors of the Federal Reserve System, published meeting materials
• The information content of central-bank language and communicationStandard literature on central-bank communication and the reaction of asset prices around announcements
• Path versus term-premium decomposition of long yieldsStandard term-structure decomposition; MR4 on the components of a yield