The short answer: On an economic calendar, Previous usually shows an earlier reading of the same indicator; Forecast is an estimate of what the next reading might be; and Actual is the value reported when the new data is published. Comparing Actual with Forecast can help identify a data surprise. But that comparison is only meaningful when the units, reference period, source, forecast timestamp and revisions line up.
A seemingly simple calendar row can hide several questions. Was the previous number revised? Does 0.3% mean month-over-month or year-over-year? Does “better” mean a larger number or a smaller one? And if a report beats expectations, why might the US dollar, gold or Nasdaq-100 react differently from a simplistic rule?
This guide gives you a repeatable method for answering those questions without turning an economic calendar into an unreliable BUY/SELL signal.
Important scope note: SeerTrend’s own economic calendar currently lists selected verified official release schedules. It does not supply consensus Forecast, released Actual or Previous values. The three-column examples in this article are fictional and exist only to teach the analysis method. Consult the original statistical agency for results and a separately identified, appropriately licensed forecasting source for genuine pre-release consensus.
What do Actual, Forecast and Previous mean?
| Calendar field | What it usually refers to | What it does not guarantee |
|---|---|---|
| Previous | The earlier observation for the same series, often the preceding month or quarter. | It may have been revised since its first publication. It is not automatically the market’s expectation. |
| Forecast | A pre-release estimate or consensus from a third-party forecasting process. | It is not the official agency’s reported result; it can vary by provider and timestamp. |
| Actual | A newly published result, ideally checked against the official release. | It is not necessarily final. Some statistics are revised later. |
A row labeled “CPI m/m” and a row labeled “CPI y/y” refer to different comparison periods. Even if both display percent signs, they are not interchangeable. The Bureau of Labor Statistics’ CPI FAQ explains the distinction between monthly and 12-month changes and between headline and core indexes.
Likewise, two calendar providers can show different Forecast values because their panels, cutoffs or latest estimates differ. Before calling anything a “miss” or a “beat,” identify the forecast provider and when that estimate was recorded.
Who publishes these numbers?
The official Actual value should be checked against the original publisher: for example, the BLS Employment Situation release, the BLS CPI releases, or the BEA GDP information.
An economic-calendar website may organize those facts and add third-party forecasts. It does not become the original statistical agency by displaying them. A scheduled release time tells you when to check, not what the released number will be.
How do you calculate an economic surprise?
For two observations measured in the same unit, a straightforward signed gap is:
Economic surprise = Actual − Forecast
That calculation describes the arithmetic gap. It does not automatically describe how important the gap is to markets, whether it is statistically large, or whether “higher” is good.
Consider this completely fictional CPI teaching example:
| Indicator | Previous | Forecast | Actual | Signed gap |
|---|---|---|---|---|
| Headline CPI, m/m | +0.1% | +0.2% | +0.3% | +0.1 percentage point |
The calculation is 0.3 − 0.2 = +0.1 percentage point. It is not a surprise of “+0.1 percent growth” and it is not appropriate to describe it as a “50% inflation surprise” without defining a special normalization method. Inflation changes are already percentages; the difference between two rates is conventionally expressed in percentage points.
Now suppose another fictional row says:
| Indicator | Previous | Forecast | Actual | Signed gap |
|---|---|---|---|---|
| Unemployment rate | 4.2% | 4.2% | 4.1% | −0.1 percentage point |
Here Actual is numerically below Forecast. Calling the result automatically “worse” because the subtraction is negative would be misleading. A lower unemployment rate may suggest a tighter labor market, but its interpretation depends on participation, employment, the survey details and the broader economy.
A color-coded red or green cell is a calendar site’s presentation convention, not a universal economic judgment.
The difference between a forecast surprise and a change from Previous
These answer two separate questions:
- Actual − Forecast: How far did the published reading differ from a stated pre-release expectation?
- Actual − Previous: How does the reported statistic compare with the previously published observation?
A report can beat Forecast while moving down from Previous, or miss Forecast while improving versus Previous. Both descriptions can be true without contradiction.
Worked example: stronger than Forecast, weaker than Previous
Imagine a fictional survey index with:
| Previous | Forecast | Actual |
|---|---|---|
| 53.0 | 51.5 | 52.0 |
- Surprise against Forecast: 52.0 − 51.5 = +0.5 index point.
- Change against Previous: 52.0 − 53.0 = −1.0 index point.
So the report is above Forecast but below Previous. The correct explanation contains both facts. Saying only “positive surprise” hides the weaker sequential trend; saying only “index fell” hides the comparison with expectations.
This is why a single traffic-light color cannot replace a sentence describing the measurement.
Why the Previous number can move after a release
Many official economic indicators are published with incomplete early source data and subsequently updated.
For US nonfarm payroll employment, BLS explains that the initial monthly estimates are revised in the next two months as additional survey responses arrive. Separate annual benchmark revisions can further change the historical series. See the BLS Employment Situation quick guide and CES revisions explanation.
BEA similarly publishes advance, second and third estimates for quarterly GDP as more comprehensive source information arrives. See the BEA GDP release information and GDP revisions page.
Fictional payroll example: why a headline beat may be incomplete
Suppose these are invented values, not a real NFP report:
| Payroll item | Thousands of jobs |
|---|---|
| Forecast for the new month | +180K |
| Newly published Actual | +200K |
| Prior month as first reported earlier | +160K |
| Prior month after revision | +120K |
The new month’s payroll beats the stated Forecast by +20K. At the same time, the prior month’s reported growth has been revised down by 40K.
These are two different pieces of information. You should not silently replace the earlier vintage and pretend analysts already knew +120K when they created the current month’s forecast. Nor should you mechanically subtract 40K from the 20K surprise and call “−20K” a scientifically valid combined surprise. That calculation would require an explicit model and a corresponding forecast of revisions.
A careful summary would say:
In this hypothetical case, the new month came in 20K above the cited pre-release forecast, while the previous month’s estimate was revised 40K lower. The broader labor-market picture is therefore more mixed than the headline comparison alone suggests.
For historical analysis, record what was known at the time of release. Today’s revised database is not necessarily the information available to market participants then.
A second trap: the same percent sign can describe different statistics
Before comparing numbers, inspect the row’s measurement label:
| Label | Typical meaning | Common error |
|---|---|---|
| m/m | Month-over-month change | Comparing it directly to a year-over-year forecast. |
| y/y | Year-over-year change | Confusing a slower yearly inflation rate with prices falling outright. |
| q/q | Quarter-over-quarter change | Mixing annualized and non-annualized GDP growth. |
| SA | Seasonally adjusted | Comparing it with an unadjusted earlier value. |
| NSA | Not seasonally adjusted | Treating seasonal movement as a new underlying trend. |
For instance, a positive inflation rate that falls from 3% to 2% normally means the measured price level is still increasing, just more slowly over that year. A negative change in the inflation rate is not the same as a decline in the overall price level.
BLS specifically distinguishes seasonally adjusted figures useful for short-term CPI comparisons from unadjusted series; seasonal adjustment factors can also be revised. See how BLS uses adjusted and unadjusted CPI.
The labels and methodology can matter more than the size or color of the number. A percent sign without a period and adjustment convention is incomplete information.
Why “better than forecast” is not a price prediction
Markets do not mechanically reward a report that looks “good” in isolation. One economic announcement can affect multiple competing channels.
For example:
- A hotter inflation reading could change expectations for interest-rate policy, bond yields or the dollar, but the market may already have expected it or focus on another component.
- Stronger employment can suggest economic resilience and also affect expectations for future monetary policy. Those implications need not pull every asset in the same direction.
- Weaker inflation might relieve rate-pressure concerns, while simultaneously raising questions about demand or growth depending on the accompanying data.
The Federal Reserve’s policy framework explicitly considers employment and price stability, with decisions informed by a range of evolving economic conditions. The Federal Reserve monetary policy statement explains those objectives. It does not promise a fixed market response to a particular news figure.
Applying the framework to gold, the dollar and Nasdaq-100
Gold (XAUUSD): Consider dollar movements, real-yield expectations, liquidity, risk sentiment and positioning. A higher-than-forecast CPI print does not prove gold must fall; an initial move can reverse.
US dollar pairs: A USD-positive interpretation of one number may be countered by risk sentiment, relative foreign rates or revisions to another series. The two currencies in a pair and the broader context still matter.
Nasdaq-100 (US100): Growth news can affect expected company earnings while inflation news can influence discount-rate expectations. Those channels can conflict, so a single calendar row cannot determine the index’s next candle.
SeerTrend does not claim to supply live market prices, consensus survey data or an empirically verified event-trading strategy. The goal is to read information correctly before assigning an interpretation.
A six-step checklist you can reuse for every important release
Use this checklist on CPI, employment data, GDP, retail sales or other macro releases. It is intentionally about information quality, not whether you should open a position.
- Verify the event — official publisher, indicator name, reference month/quarter, publication timezone and source URL.
- Match the unit — m/m, y/y, annualized q/q, count of jobs, percentage points, index level and SA/NSA.
- Freeze the Forecast — write down its provider and pre-release timestamp, if you have an authorized consensus source. Leave it blank if you don’t.
- Capture the Actual from the original release — note whether the figure is preliminary and whether the source has updated or corrected it.
- Read revisions and secondary lines — previous-vintage changes, labor-market detail, headline versus core and changes in methodology.
- Write a two-sided interpretation — what the reading appears to support, and what remains unknown or could contradict the first impression.
A compact note might look like this (the numbers below are deliberately fictional):
CPI m/m: Actual +0.3%, Forecast +0.2%, Previous +0.1%. The reported month is +0.1 percentage point above the cited forecast. Before discussing rates or prices, check the core measure, revisions, seasonal adjustment, the forecast’s timestamp and the source release. No directional market conclusion yet.
That single paragraph is more informative than “red news bullish” or “green news bearish.”
What if the Forecast or Actual column is missing?
Do not fill it from memory. Official scheduling pages may contain the event name and time but no pre-release economist consensus. Actual results do not exist until the agency publishes the data.
If you cannot verify a forecast, state “consensus not assessed.” If the event has not been released, state “Actual not yet published.” If the source only verifies a day, state “exact time unverified” and do not start a countdown.
This is precisely why SeerTrend currently shows a limited official US economic release schedule with source links and verification status. It does not invent CPI/NFP dates when its BLS schedule source is unavailable, or invent Actual/Forecast values to make a table look full.
Use the SeerTrend economic calendar to check covered official release dates, and open the linked agency page to validate details. For a worked case involving two BEA announcements at the same hour, see our GDP and PCE same-time analysis guide.
Frequently asked questions
Does Actual above Forecast always mean good news?
No. “Above” is an arithmetic comparison, not an assessment of welfare, inflation risk or asset prices. For an unemployment rate, a lower reading may indicate a tighter labor market; for inflation, higher may be concerning depending on circumstances. Verify what the indicator measures.
Can Previous change when Actual comes out?
Yes. Some series, including US payroll employment, publish revisions to previous periods. A calendar provider may display the newly revised Previous value, while an old screenshot shows what was originally reported. Both may refer to different release vintages.
Is Actual minus Forecast a percentage surprise?
Not necessarily. If the series already reports a rate such as 0.3%, subtracting 0.2% gives 0.1 percentage point. A percentage relative error requires a defined denominator and can be unstable when the forecast is near zero or negative. Always state your unit.
Why does the price sometimes move against a “positive surprise”?
A single result can affect growth, inflation, policy expectations and positioning at once. Other simultaneous data and previously priced expectations matter. A forecast beat, by itself, does not establish the cause of a particular move.
Does SeerTrend have Forecast and Actual figures?
No, not currently. The SeerTrend calendar focuses on selected official release schedules, source provenance, local-time conversion and freshness. This article teaches how to interpret Forecast/Actual/Previous when those fields are available from verified external sources.
Primary references and editorial method
This is an evergreen educational guide, reviewed on October 10, 2026. All numerical case studies above were invented for teaching; none is a real published report, economic forecast or price backtest. Review the methodology and source links periodically rather than changing the URL or title for each new release.
- BLS — Consumer Price Index FAQs: month-over-month, 12-month, headline/core CPI and reporting conventions.
- BLS — Using seasonally adjusted and unadjusted CPI: comparability, seasonal treatment and revisions.
- BLS — Employment Situation methods quick guide: sample-based payroll reporting and revisions.
- BLS — Nonfarm payroll revisions: initial, later and benchmark estimate vintages.
- BEA — GDP Release: Additional Information: preliminary quarterly GDP vintages and revisions.
- BEA — GDP Revision Information: comparison of advance, second and third estimates.
- Federal Reserve — Monetary Policy Report statement on longer-run goals: multiple policy considerations, not a mechanical market-price rule.
Educational disclosure: SeerTrend is an independent information and synthetic-learning project. The worked examples are not live data, individualized investment advice, a backtested strategy, or a recommendation to trade leveraged instruments.