Macro vintage revision bridge
Separate a prior-level revision from new-period movement under compatible positive-level declarations. Not official macro data or a forecast.
Pin the object before comparing its levels
The first and second values must describe the same prior period under compatible definitions; the third belongs to the new period. All must share units, coverage and relevant adjustment convention. A changed index base, nominal/real mix or scope revision can break that contract even if the numbers look reasonable. The tool validates syntax, not source compatibility. It accepts no series file or genuine producer receipt and fetches no live release. The invented defaults are not current GDP, CPI or an official indicator. A real calculation needs actual source identity and vintage notes before an economic interpretation is attached.
One common base gives additive contributions
Subtract original prior from revised prior to obtain the level revision. Subtract revised prior from new level to obtain the compatible-period difference. Their sum equals new minus original. Dividing both differences by the original prior creates additive percentage-point contributions to the original-base headline. The comparable-period growth rate uses revised prior as its denominator, so it is not that additive contribution. The output labels keep those objects separate. The default revision contributes 2 percentage points and period movement 1 on the original base, while the period rate is about 0.980392%. More precision does not authenticate the declarations.
The independently based rates combine multiplicatively
The revised-to-original ratio multiplied by the new-to-revised ratio reconstructs the new-to-original ratio. Subtract one and multiply by one hundred to recover the original-base headline. Adding the revision rate directly to the comparable-period rate mixes bases and generally gives a different result. Seven authored cases expose downward revisions, zero period movement, offsetting contributions and stable levels. Negative differences are allowed because positive levels can fall. They are not forecasts, probabilities or observations of how a real market reacted. Actual releases and prior expectations remain separate missing source objects.
Finite plain decimals and invalidation
Levels must be strictly positive plain decimal strings no greater than one billion, with at most twelve fraction digits. Zero, negatives, non-numeric values, hexadecimal and exponent declarations are rejected. A negative level series lies outside this particular contract, even if it is a legitimate economic quantity elsewhere. Invalid edits clear the previous results and disable export. Reset and reload restore the authored default. There is no private-input persistence. JSON retains input strings, numerical outputs, engine identity and unverified-source limits rather than pretend the export is an authenticated economic release or independently audited article.
Neither surprise nor financial performance is inferred
A consensus surprise needs a genuine earlier expectation with issuer, horizon and information cutoff. A causal price-response claim needs its design, actual event clock and alternative explanations. An account outcome needs quantities, fills, costs and capital ownership. None is supplied by three compatible levels. The bridge performs no seasonal adjustment, rebasing reconciliation, chained-component contribution, forecast fitting or strategy validation. Actual instrument research retains its full authoritative method and independent master. A passing software fixture concerns this implementation only, not model admission or financial success.
The bridge is one part of a source dossier
The complete workbook joins conditional arithmetic with source ownership, reference periods, availability and versioned corrections. Its survey and profile cards show why documentation cannot replace authentic fieldwork or testimony. Known promotional purpose and missing commercial records remain material for a publication even when legal ownership is separate. A bibliography cannot turn repeated related-site claims into independent corroboration. The site publishes static education on existing managed cloud hosting with no new scheduler, PC fallback, source monitor or paid pipeline. A real stronger finding needs actual records rather than a denser paragraph.