The release calendar and revisions

Every macro number you will ever react to — growth, inflation, factory output, trade — arrives on a schedule, and arrives in drafts. The first version of a number is published fast, built from incomplete information, and then quietly revised, sometimes more than once, as fuller data comes in. Often the revision is small. Sometimes it is large enough to reverse the story the first print told: a "slowdown" that becomes "steady," a "surge" that becomes "ordinary."

Most readers never see the revision, because the first print gets the headline and the correction gets a footnote months later. So they carry around a picture of the economy built from first drafts that were later erased. The question this module asks is the plainest kind of honesty: when a macro number prints, how finished is it — and how much can it still move? Knowing the answer changes how much weight you put on any single release, and stops you from building a view on a draft.

There is a reason the numbers come out in drafts, and it is not carelessness. A statistics office faces a genuine trade-off between speed and completeness. Everyone wants growth figures quickly, so the office publishes an early estimate built from the data it has — a few months of the year, some fast surveys — and fills the rest with reasoned assumptions. As the missing months' hard data arrive, it replaces the assumptions with facts and republishes. The number gets more accurate and less timely at each pass. This is not a bug; it is the deal you accept to get any number at all before the year is fully counted.

In India the pattern is concrete. The MoSPI — the Ministry of Statistics — publishes GDP as an advance estimate first (a fast, partial-year draft), then provisional estimates, then revised and finally settled figures across the following months and years. Factory output, the IIP, is published quickly and then revised as more factories report. Inflation prints monthly and is comparatively firm, but even it gets small corrections. Each series has its own release calendar — a published timetable of what prints when — and its own revision cycle.

The practical danger is simple. If you treat the first print as final, you will over-react to numbers that were never solid, form a view, and then be caught flat-footed when the revision erases the very fact you built on. A macro headline is itself a claim, and the honest reader strips the trick before trusting the number — and the first, most basic trick to strip is vintage: how early, and therefore how provisional, is this print?

Think of a single quarter's growth number as passing through stages, each fuller than the last:

• The advance estimate is the first draft, published soonest. It leans heavily on partial-year data and assumptions for the months not yet counted. It is the least reliable and the most quoted.
• Provisional estimates replace some assumptions with actual data as it arrives. The number firms up.
• Revised and final estimates come later still — sometimes a year or more on — once nearly all the hard data is in. This is the version closest to the truth, and the one almost no one reads, because by then the news cycle has moved on.

The gap between the first and the final is the revision, and its size varies by series. Growth and factory output can revise meaningfully; inflation less so. The revision does not usually mean the economy changed — it means the picture of the economy sharpened as guesses became facts. That distinction is the whole lesson: a revision measures improving information, not a shifting world.

Notice what this does to "the number came out today." A same-day reaction is a reaction to the least finished version of the fact. The reader who waits for the fuller estimate, or who at least discounts the first print for its known provisionality, is reacting to better information — and is far less likely to be whipsawed when the draft is rewritten.

Walk one release through its vintages. illustrative

A quarter's advance GDP estimate lands showing growth at, say, a soft 5.5%. The headlines write themselves: "growth slows," "recovery loses steam." Commentators extend the line — if growth is fading, then consumer companies will struggle, lenders will see slower loan demand, the cycle is rolling over. A reader who takes the 5.5% as settled truth builds a gloomy view of the whole economy on it.

Now follow the same quarter forward. Months later, the provisional estimate arrives at 6.1% as actual data from the uncounted months comes in stronger than the early assumptions. Later still, the revised figure settles near 6.4%. The economy did not accelerate over those months — it was always whatever it was. What changed was the statisticians' picture of it, as placeholder assumptions were replaced by hard data. The final story is "steady," not "slowing." But the "slowing" headline was the one that reached the reader, moved the mood, and shaped decisions; the quiet revision to "steady" reached almost no one.

The reader who knew the first print was an advance estimate did something different. They noted the 5.5% and its vintage, treated it as a provisional signal rather than a verdict, and did not rebuild their whole view of consumer demand and lending on a draft. When the number firmed up, they were not surprised, and they had not made a decision they now had to unwind.

This is the honest read: not "ignore the data" but "weight it by how finished it is." The first print is information; it is simply the least complete information you will ever have about that quarter.

Reading the calendar and the revisions keeps you honest, and it has limits worth stating.

A first print cannot tell you its own final value. By definition, the advance estimate does not know what the revised figure will be; if it did, there would be no revision. So any decision that needs the final number cannot be made on release day. The most you can honestly say is "the early, partial picture is X, and it may move."

The revisions cannot be predicted to a number, either. You know the first print is provisional; you do not know which way, or how far, it will be corrected. Anyone who claims to forecast the revision precisely is doing the thing this whole book warns against — offering a confident number where only a range and a mechanism honestly exist. The useful move is to hold the first print loosely and watch the fuller data confirm or overturn it, not to guess the correction.

And no vintage of a single number tells you what it means for a company. A macro print is one input; the chain from "growth was 6.4%, finally" to "this particular lender's loan book" runs through the transmission steps the rest of this book has traced. The data's completeness is one honesty; the transmission is another.

The release calendar is where confident-sounding people get caught by the boring fact that first drafts get rewritten.

1. Treating the advance estimate as final. The first print is the fastest and least complete version of a number. Reacting to it as settled truth is reacting to a draft that may be erased.

2. Remembering the headline, missing the revision. The first print gets the front page; the correction gets a footnote months later. Readers end up carrying a picture built from numbers that were later rewritten.

3. Reading a revision as a real-world change. When "slowing" becomes "steady" with no change in the actual economy, that is the picture sharpening, not the world moving. Confusing the two invents accelerations and slowdowns that never happened.

4. Confusing the source's honesty with the number's precision. An official series can be produced in complete good faith and still be a provisional estimate with a real margin of error. Trusting the office is not the same as treating its first draft as exact.

5. Trading the calendar. Reacting on release day, before the number is finished and before its transmission to any company is worked out, is reacting to the least reliable version of the least relevant thing. The calendar tells you when to read, not when to act.

Fold it into one working habit. Whenever a macro number crosses your screen, ask three questions before you let it change your mind. Which vintage is this — an advance estimate, a provisional figure, or a final one? How much does this series usually get revised — growth and factory output can move meaningfully, inflation less so? And when does the fuller version arrive, so you know whether to hold your view loosely and wait?

Those questions cost nothing and protect you from the most common way careful readers are fooled: building a firm conviction on a number that was always a draft. It fits the discipline this whole book keeps returning to — trace the mechanism, respect the uncertainty, and never mistake a provisional signal for a settled fact. The economy does not lurch as often as the headlines suggest; a great deal of what looks like acceleration and slowdown is simply the statistical picture coming into focus. The reader who weights each print by how finished it is will react less, be whipsawed less, and end up with a steadier, truer read of the cycle than the one who trades every first draft.