Cocoa grindings: how to compare demand without mixing the baselines
4 min read · Free public research
Quarterly processing figures need matching periods and reporting coverage. Learn how year-on-year, quarter-on-quarter and combined regional changes differ.
A cocoa grindings headline is most useful when it answers three questions: what was processed, over which period, and compared with what? A percentage without that context can make a demand report look stronger or weaker than the underlying observation.
This guide is a method for reading releases. It does not publish a new regional total or make a claim about today’s demand.
Invented values with identical reporting coverage. A quarterly rebound can coexist with a year-on-year decline.
The European Cocoa Association publishes a dated archive of quarterly grindings reports. Use the underlying release rather than a headline summary to record the quarter, tonnage and comparison period. Keep the publication date separate from the quarter being measured.
A simple research record contains the publisher, region, covered quarter, current tonnes, comparison tonnes and any coverage or revision notes. Leave unavailable fields empty rather than estimating them from an incomplete headline.
Year-on-year and quarter-on-quarter
Year-on-year compares a quarter with the same quarter a year earlier. Quarter-on-quarter compares it with the immediately preceding quarter. Both use (current ÷ comparison − 1) × 100, but they answer different questions.
For illustration only, suppose processing is 90 units this quarter, 100 in the same quarter last year and 80 in the previous quarter. Year-on-year is −10%; quarter-on-quarter is +12.5%. Both statements are mathematically correct. These are invented units, not cocoa observations or forecasts.
The example shows why an improving sequential figure does not necessarily mean activity has recovered to its previous annual level. It also shows why the baseline belongs next to the percentage.
Combine tonnes, not percentages
Suppose one region moves from 100 to 110 units and another from 300 to 285. The changes are +10% and −5%. Their simple average is +2.5%, but combined activity moves from 400 to 395: −1.25%.
Again, these are illustrative units. The calculation is a checking tool, not a market estimate. A valid combined figure also requires compatible periods and coverage. If regions are missing, label the result as the covered regions rather than a global total.
From processing to a research question
A processing observation is not a direct retail-sales count. Avoid jumping from a regional change to a claim about all chocolate consumption or a guaranteed futures reaction. Ask what other evidence would support that interpretation.
Read the release alongside inventory movements, supply reports and the expectations recorded before publication. Keep forecasts distinct from confirmed observations; otherwise an apparent “surprise” can be created by changing the benchmark afterwards.
Build a repeatable demand check
Start with Cocoa Market Hub’s cocoa grindings guide, then compare the dated figures available in the daily overview. The free preview introduces selected market information. Full database membership adds the wider research workspace and historical comparisons; it does not remove the need to check original sources.
Sources: European Cocoa Association quarterly archive; ICCO statistics and Quarterly Bulletin description.
Source · icco.org ↗
Original educational analysis. Numerical examples are hypothetical, not current market data. Research is not personalised investment advice.
Additional reading context
Educational explanation added to help interpret this article. It is not a new market update or a revision to the original dated figures.
Putting the guide into practice
Choose one dated observation relevant to this guide and trace it back to the original source. Record the benchmark, unit and reporting period before comparing it with another figure. This turns the explanation into a repeatable research process: establish what is measured, check what changed and identify what the observation cannot tell you.
Use the database to check the neighbouring evidence. A price observation can be compared with physical availability and positioning; an inventory or arrivals observation can be checked against compatible history and demand reporting. Agreement between several observations can strengthen a hypothesis, while disagreement highlights the next question to investigate. Neither outcome removes the need to check timing and definitions.
Applying this topic
For demand research, compare grindings for the same region and quarter before combining them. Regional percentages should not simply be averaged: the underlying tonnage bases differ. Also separate processing volumes from retail chocolate sales, because they describe different stages of the chain. Look for compatible figures and dated company commentary before treating one observation as a verdict on global consumption.
Published research and personal commentary, not personalised investment advice. Trade ideas can fail; evaluate the assumptions and dated follow-ups. The original published post is preserved.