Economic Calendar
The economic calendar puts the US data schedule inside your replay session. You see when CPI, payrolls, GDP and twelve more releases are due, you watch the countdown run down as the candles stream, and the number appears only once the playhead has passed it — never earlier, and never on the boundary itself. It turns “I replayed an NFP Friday” into “I traded into the print without knowing the print.”
What it is
TestMax stores every US release exactly as it was first published — the original print, from the St. Louis Fed’s ALFRED archive of historical vintages, not the number as later revised. During a replay the platform delivers the releases that fall inside a rolling forward window of up to 48 hours from the playhead, so a session always knows what is coming next without knowing what happened after.
The window is bounded by the session itself: a session only ever loads the releases inside its own bar range, so the schedule never reaches past the session’s last bar. On a one-day session the last hours of the replay show fewer and fewer releases ahead, and near the end the panel is empty.
Coverage is 15 US government series, the ones that actually move ES, NQ, CL, GC and EURUSD.
Where it appears
Four surfaces, all driven by the same data:
- Floating calendar panel — A draggable, pinnable schedule listing each release with its date, time in your chosen timezone, impact stripe, reporting period, and the three value columns. Drag it by its header, collapse it, filter to High impact only, and it stays where you put it — the position, the filter and the collapsed state persist across sessions. Turn it on from the Visible Panels menu on the dashboard (the gear in the panel header) by ticking Economic Calendar (desktop). The panel is desktop-only — it needs the width to hold three value columns next to the chart, so it does not render on narrow screens even when the tick box is on. Clicking any upcoming row jumps the replay to one bar before that release, so you arrive positioned in front of the print rather than after it.
- Chart event rail (Pro) — A vertical marker on the price chart at the exact moment of each release, with a pill label naming it. Dashed while the release is still ahead of the playhead, solid once it has fired. Releases landing inside a candle are placed at their true position within that candle, not snapped to its edge.
- Scrubber ticks — Thin marks on the progress bar above the playback controls, one per release instant, showing at a glance where in the session the news lands. Brand-coloured for high impact, muted for medium; dimmed while upcoming, full strength once passed.
- Countdown HUD — A pill floating above the playback bar showing the next release and the time remaining, in replay time. It runs grey while the release is far off, amber inside 15 minutes, and glows red inside 5 minutes, then flips to the printed value (on Pro — free accounts see the reveal with a PRO chip in place of the number) when the release fires and dwells there for ten seconds before returning to the countdown.
Multiple releases at the same instant are collapsed into a single entry. On an NFP Friday payrolls, unemployment and average hourly earnings all print at 08:30 — the rail, the tick and the HUD show one marker headlined “Nonfarm Payrolls +2”, and the panel lists all three rows under one time.
The no-lookahead guarantee
This is the whole point of the feature, so it is worth stating precisely.
A release value appears only once the playhead has strictly passed it. The playhead is the exclusive end of the last emitted bar. A release whose timestamp equals the current playhead has not happened yet: it is still upcoming, still counting down, and its value is still withheld. It reveals on the next bar, which is the first bar that could possibly contain the reaction.
That distinction matters more than it sounds. Almost every modern US release lands exactly on a minute boundary, so “equal to the playhead” is the ordinary case, not an edge case. Treating it as already-fired would hand you the number one bar before the market got it — on every single release.
The same rule governs the countdown (“due” means due, never released), the rail (dashed until strictly passed), and the panel row (dashes, not numbers, until then). Jumping to a release from the panel lands the playhead exactly on the release with the value still hidden.
Reading the three columns
The panel shows three value columns for each release — labelled Actual, Prior and Chg — and they are only filled in once the playhead has passed it. What the third column holds depends on which kind of series it is.
Movement series — payrolls, average hourly earnings, CPI, core CPI, PPI, PCE, core PCE, retail sales, GDP, durable goods and industrial production, eleven of the fifteen — are printed the way a calendar prints them, as the move itself:
| Column | What it is |
|---|---|
| Actual | The change since the prior period, signed, in the form a trader reads it — +73K for payrolls, +0.2% for CPI |
| Prior | The previous release’s printed change — +0.2% for last month’s CPI. An em-dash when that release falls outside your session: the raw level the change was computed from measures something else and does not belong in this column |
| Change | An em-dash. There is nothing to put here: the actual column already is the change against the prior print, and restating it would just repeat the same number |
Level series — Unemployment Rate, Initial Jobless Claims, Housing Starts and Michigan Consumer Sentiment, the other four — are readings rather than moves, so they carry no sign and the third column does the arithmetic:
| Column | What it is |
|---|---|
| Actual | The reading as originally published — 4.1% for the unemployment rate, 71.8 for Michigan sentiment, 1,427K for housing starts (the archive counts homes in thousands, so that is 1.427M) |
| Prior | The previous reading of the same series |
| Change | The move from that prior reading to this one, signed |
Where a change is shown it is change versus the prior print — not surprise versus a forecast. TestMax has no consensus column (see Limits below), so nothing here tells you whether the number beat expectations. It tells you whether the data got better or worse than last time.
A small rev badge means an earlier figure was revised after it was first published — the revision itself was news, and this flags it. On a level series it sits beside the prior reading, which is the figure that was revised. On a movement series it sits beside the actual, because the revision is to the level that this release’s change was measured against; the prior column there holds the previous release’s own print, which nothing rewrites.
Covered releases
All 15 series, with the impact TestMax assigns and the standing publication time. There are only two impact levels: high and medium.
| Release | FRED series | Impact | Release time (ET) |
|---|---|---|---|
| Nonfarm Payrolls | PAYEMS | High | 08:30 |
| Unemployment Rate | UNRATE | High | 08:30 |
| Average Hourly Earnings | CES0500000003 | High | 08:30 |
| CPI | CPIAUCSL | High | 08:30 |
| Core CPI | CPILFESL | High | 08:30 |
| PCE Price Index | PCEPI | High | 08:30 |
| Core PCE Price Index | PCEPILFE | High | 08:30 |
| Retail Sales (Advance) | RSAFS | High | 08:30 |
| Real GDP | GDPC1 | High | 08:30 |
| PPI (Final Demand) | PPIFIS | Medium | 08:30 |
| Initial Jobless Claims | ICSA | Medium | 08:30 |
| Durable Goods Orders | DGORDER | Medium | 08:30 |
| Housing Starts | HOUST | Medium | 08:30 |
| Industrial Production | INDPRO | Medium | 09:15 |
| Michigan Consumer Sentiment | UMCSENT | Medium | 10:00 |
Times are shown in your account timezone throughout the interface; the table above lists the publishing agency’s own schedule in US Eastern time.
Free vs Pro
The schedule is free. The numbers are Pro.
| Free | Pro | |
|---|---|---|
| Calendar panel with names, times, periods and impact | Yes | Yes |
| Scrubber ticks | Yes | Yes |
| Countdown HUD | Yes | Yes |
| Jump to just before a release | Yes | Yes |
| Actual, prior and change values | Locked | Yes |
| Chart event rail | — | Yes |
On Free, a passed release shows a Pro pill where its numbers would be, in both the panel and the HUD reveal; clicking it opens the upgrade prompt. Nothing is hidden on an upcoming release for anyone — the number does not exist yet for any account, so both plans see dashes.
Blind mode
Blind mode hides the instrument and date so you cannot recognise the session you are trading. Because a release schedule identifies a day instantly — “CPI plus Jobless Claims at 08:30” is a date — blind mode suppresses all four surfaces: no panel, no rail, no ticks, no countdown. They all return when you turn blind mode off.
Timeframes and the countdown
The countdown HUD renders whenever a bar is 30 minutes or shorter — the test is the bar’s own length, so a 7m chart keeps the countdown and a 90m one does not. Weekly, monthly and yearly bars never show it. Above the threshold a countdown would be silently stale: on a 1h chart a release reveals on average 29 minutes after it happened, because that is when the containing bar closes, and “red inside 5 minutes” is unreachable when a bar is an hour wide.
On longer bars everything else stays put. Every account keeps the calendar panel and the scrubber ticks, so you can still see where in the session the news lands and jump to just before it. On Pro, the release also keeps its labelled marker on the chart rail, and its values become readable once its containing bar closes; on Free the rail is absent at every timeframe and the values stay behind the Pro pill. Switch to a 30m-or-shorter bar for the live countdown.
Limits
1. There is no consensus forecast. Consensus numbers come from proprietary surveys of economists with no free, redistributable source, so TestMax does not publish one. This means every figure on the panel is movement versus the prior print, not surprise versus forecast — and market reaction is driven by surprise. A number that looks strong against last month may still have missed what the street expected.
2. Dates are real, clock times are assumed. The release date comes from the ALFRED archive and is exact. ALFRED records no clock time, so each series is placed at its agency’s standing release time in America/New_York — 08:30 for most, 09:15 for Industrial Production, 10:00 for Michigan Consumer Sentiment. A release that historically moved (a weather closure, a government shutdown) will sit on the wrong bar.
3. ISM and PMI are absent. Neither has a free redistributable source, so neither is covered. Coverage is the 15 US government series listed above — no ISM Manufacturing, no ISM Services, no S&P Global PMIs, and no non-US releases.
4. Fills inside a release bar are indicative. Orders are evaluated one bar at a time, against the bars of your chart timeframe, and nothing records the order in which the high and the low were reached inside a bar. A stop or a limit sitting inside that range fills on an assumed path rather than on the real one, and the wider the timeframe, the more of the move that assumption covers. That is true of every bar, and a release candle is where it matters most: the whole move can land inside the 08:30 bar. Treat fills inside a release candle as indicative.
5. A period appears once, on the day it was first published. Where an agency publishes the same reporting period more than once, only the first publication is a calendar event. Real GDP is the case that matters: the quarter’s advance estimate is on the calendar, and the second and third estimates that follow it a month and two months later are not. Revisions to an already-published period still reach you, but as the rev badge on the next release of that series rather than as an event of their own.