How We Calculate Event Reaction Data
The numbers on every event page in this tool are not estimates or a backtest. Every release shown there is measured directly from raw intraday price data around the exact moment the report hit the wire, using the same three-step pipeline for every event this tool tracks. Here is exactly how it works, so you can judge the data on its own terms rather than take it on faith.
Step 1: Pulling intraday candles around the release
For each release date, we request 5-minute intraday candles for the event's primary currency pair from EODHD's market data API, covering a window that starts 30 minutes before the scheduled release time and extends 65 minutes after it. That window is wide enough to capture the pre-release baseline price and the full first hour of the reaction, without pulling in an entire day of unrelated price action that would dilute the signal. The release timestamp itself comes from EODHD's economic-events calendar, which timestamps each report to the minute in UTC. We match that timestamp against the closest available candle to establish the baseline price the market was trading at the instant the numbers were published.
Step 2: Measuring the pip move at three checkpoints
From that baseline candle, we look forward to three fixed checkpoints (5, 15 and 60 minutes after the release) and find the closest available candle to each one. The pip move at each checkpoint is the difference between that candle's closing price and the baseline, converted into pips using the pair's actual pip size. Tracking three checkpoints instead of just one shows whether a release produced an immediate spike that faded, a move that kept building through the hour, or a reversal once the initial reaction was digested. Those are three different patterns that a single 60-minute figure would flatten into one number.
Step 3: Classifying direction and averaging across releases
The 60-minute pip move also decides the direction badge shown in the table: a move beyond +1 pip is classified 'up', beyond -1 pip is classified 'down', and anything inside that band is classified 'flat', a small buffer that keeps ordinary noise from being misread as a directional reaction. The average-move figure quoted near the top of the page is the mean of the absolute 60-minute pip moves across the last 10 tracked releases, recalculated as new releases roll in and the oldest one drops out of the window. Every step above runs identically for every event on this tool; only the pair, release type and pip size change.
This means the release calendar dates and the description of what each event measures come from the outside sources linked on each event page, but every pip figure, chart and direction badge on those pages is calculated in-house from the raw price data described above, not sourced from a third party. If a release ever looks off to you, the fastest way to check it is to pull up the event's primary currency pair on your own chart around the timestamp in the Date column and compare it against the checkpoints in the table.
Events measured with this methodology
Historical price reactions are shown for educational purposes only. Past reactions do not guarantee future outcomes and are not financial advice.