A Fed statement, its projections and the Chair’s answers arrive as separate pieces of information. A hypothetical EUR/USD move of 30 pips followed by an 80-pip reversal illustrates why reading only the first headline can miss the rest of the event. This article explains how to compare policy, projections and communication in their actual sequence.

EFFR: weighted median, not mean

New York Fed defines EFFR as a volume-weighted median of overnight transactions. Our simple dataset shows why three aggregation methods can give three rates.

Scroll sideways to see the full diagram

EFFR: weighted median, not meanThree observations at 3%, 3.5% and 4% have volumes 10, 10 and 80. Sorted cumulative shares are 10%, 20%, 100%. Half the volume falls within the 4% group, so the weighted median is 4%. Ordinary median is 3.5%; weighted mean 3.85%.Weighted median: volume mattersThree fictional overnight transactionsRateVolumeCumulative3.0%1010%3.5%1020%4.0%80100%Volume sorted by rate0%50%100%Weighted median: 4.00%Ordinary median: 3.50%Weighted mean: 3.85%
Three fictional overnight observations with annual rates 3%, 3.5%, 4% and volumes 10, 10, 80 arbitrary units of the same currency. Weights total 100; sorted cumulative shares are 10%, 20%, 100%. Half-volume lies inside the last group, so no tie arises. This illustrates a volume-weighted median, not full EFFR reconstruction from FR 2420 or current Fed data. Actual rate methodology: New York Fed — EFFR.
The same transaction list yields different measures under different aggregation methods.

Ordinary median weights each observation equally and selects the middle 3.5%. Weighted mean is (3 × 10 + 3.5 × 10 + 4 × 80) / 100 = 3.85%. Weighted median locates half the volume rather than averaging rates. It is not an FOMC target-range decision.

What the Fed actually publishes

The FOMC normally holds eight scheduled meetings a year and can also act between meetings. It sets the target range for the federal funds rate as part of monetary policy. Check the official calendar rather than assuming every policy action follows a regular scheduled meeting.

The target range is distinct from EFFR, the New York Fed’s volume-weighted median of specified overnight transactions. Neither is the contractual financing rate on a CFD position. The diagram compares aggregation methods; it does not reconstruct the full FR 2420 dataset.

Read the decision, statement, any projections and the later conference separately. Four regular meetings normally include SEP materials, as identified in the official calendar. There is no permanent ranking of which field is least interesting or determines the currency move:

  1. The statement: record the target range, economic assessment, risks, votes and conditions, comparing whole passages with the previous version.
  2. SEP and dot plot: participants, including nonvoting Reserve Bank presidents, submit individual assessments of appropriate policy under their assumptions. The median is not an official commitment, vote, market forecast or guaranteed path.
  3. The conference: retain later answers and timestamps. A conditional answer about the outlook does not fix the next decision or guarantee a reversal.

The timing, step by step

Regular FOMC statements are scheduled at 14:00 US Eastern Time, followed by the Chair’s press conference around 14:30 Eastern. Convert both events for the meeting date. Eastern Time alternates between EST and EDT; your local zone or broker server may follow different clock-change rules.

A scheduled statement and press conference provide separate information events. The interval before or between them need not be calm: news, positioning and liquidity can change at any time. Observe the actual sequence instead of assuming quiet first, then a larger conference move.

Expected outcome and expected average differ

CME FedWatch derives implied probabilities from 30-Day Federal Funds futures prices under assumptions about rate increments and the relation of EFFR to the target. These are model outputs with a timestamp, not guaranteed decisions. The contracts refer to average daily EFFR during the month, not simply the rate after a meeting; position within the month matters.

In our invented two-outcome model, the old target range is 4.25–4.50%, midpoint 4.375%. Assign 70% to a 25-basis-point cut to 4.00–4.25%, midpoint 4.125%, and 30% to a hold. Expected midpoint is 0.70 × 4.125 + 0.30 × 4.375 = 4.200%. If the cut occurs, change versus old is −25 basis points, difference from the most likely outcome is zero, and difference from the weighted expectation is −7.5. These are different reference points, not actual FedWatch readings or complete measures of policy surprise.

Hypothetical monthly contract · same price, different end rate
Price 96.00 implies monthly-average EFFR 4.00% under 100 minus price30-day model month
15 days at old 4.25%, then 15 days at xx = 3.75%
25 days at old 4.25%, then 5 days at xx = 2.75%
Each weighted monthly average remains 4.00%not a full FedWatch probability calculation

Policy interpretation and the other currency

Read the hawkish and dovish terminology as relative interpretations, alongside conditions and prior information. These possible situations do not provide a stable ranking of surprise size or market response:

  • A hawkish surprise can support USD if it raises expected relative yields, but currency direction and magnitude also depend on risk premia, simultaneous news and the information already priced.
  • A dovish surprise can weigh on USD through revised expected relative yields; EUR/USD need not rise in every case.
  • An unchanged policy rate can accompany unexpectedly restrictive communication. Its currency reaction need not be smaller than the reaction to a rate change.
  • Even an expected decision can accompany new projections, guidance or simultaneous news. No verified frequency ranking or near-zero reaction is assigned to this scenario.

Interest-rate parity relates spot and forward prices to comparable interest rates under stated assumptions; it does not prove that higher policy rates cause a currency to appreciate. Revisions to expected policy and yields can affect spot alongside risk premia and other news.

Separate later information from a cross-rate model

Read the substantive reasons and conditions with the later answers; no communication stage is automatically the largest source of volatility. Later minutes cannot be used as knowledge available at statement time. If the hypothetical decision instead holds the midpoint at 4.375%, its change is zero but difference from the model expectation 4.200% is +17.5 basis points. That is an absent expected cut, not a hike, and it supplies no pip forecast.

The second currency can change a cross-rate conclusion. Fictional consistent reference quotes EUR/PLN 4.40 and EUR/USD 1.10 imply USD/PLN 4.00. If EUR/USD falls 2% to 1.078 while EUR/PLN holds, USD/PLN is about 4.0816. If EUR/PLN also falls 2% to 4.312, USD/PLN remains 4.00. This identity excludes spreads, costs and executable offers; it is not a forecast of a Fed effect. Apply the same dated reading discipline to Fed chair public remarks.

What to do at the next meeting

  1. Record the expectation. Save its source, timestamp, model assumptions and outcomes; a modal scenario and weighted average need not coincide.
  2. Plan exposure. Define a budget and actual execution terms for positions, pending orders and automated strategies. Account age or conference end cannot eliminate risk.
  3. Preserve the event sequence. Save statement, projections and later answers separately, with publication times and policy effective dates.
  4. Keep a reproducible observation. Choose price type and windows beforehand, include the other currency and costs, and retain contrary cases. A few meetings cannot establish reliable reversal prediction.