Follow the tonnes, then ask the question
A balance sheet is a set of connected quantities. Production, imports and the stock carried into a year form the available supply. Domestic use and exports account for material leaving that supply. The remainder becomes the next year’s opening position. Reading one row in isolation misses the structure.
Our data desk transcribes the September 25 federal outlook for three crop years, keeps the original thousand-tonne unit and calculates a stocks-to-use ratio. Each result is reproducible from the downloadable rows. This is an independent calculation from published data, rather than a separate crop survey.
The forecast is a dated scenario
For 2026–27, the September table records supply of 24.080 million metric tonnes, exports of 7.700 million, total domestic use of 14.401 million and closing stocks of 1.979 million. Exports plus domestic use give 22.101 million metric tonnes. Divide closing stocks by that use and the resulting ratio is 8.95%.
The practical value of the ratio is consistency: the numerator and denominator come from the same release, crop year and material boundary. It does not measure oil inventories, a farm’s available grain, a delivery opportunity or the probability of a price move. A later estimate can change both halves of the calculation.
Keep seed, products and time apart
The outlook’s canola crop year runs from August to July. Its canola export row excludes oilseed products. Seed exported, oil produced and meal produced therefore belong in distinct series; adding them together would count different materials and, in some comparisons, the same material twice.
A grower can use this desk to frame a question about destination demand. A processor can inspect the relationship between domestic use and available supply. A reporter can reproduce the arithmetic before describing a change. None of those tasks requires pretending a forecast is an observed transaction.
What will change this reading
The next useful update is a verified source revision or another observed series that answers a defined question. We keep forecast vintages so a revision remains visible, instead of silently replacing yesterday’s scenario. The data desk labels the release date separately from the date we checked the source.
Use the original source to examine its definitions and revisions. Use our comparison controls to check the selected year and units. A fresh check date does not make September’s observations October observations.