A qualified appetite

The most useful news for the cocoa market last week came from factories rather than farms. Figures released on October 15th showed European processors using more beans than a year earlier, interrupting a long run of contraction. North America went the other way. For a commodity whose price depends on both harvests and the willingness of manufacturers to buy them, the contrast matters. Bean use is no longer contracting in every major reported market. A general recovery in chocolate consumption is a much larger claim.

European bean usage reached 348,215 tonnes in July–September, up 3.2% from the same quarter of 2025, according to the European Cocoa Association (ECA). It was the first annual increase since the second quarter of 2024. North American grindings, published by the National Confectioners Association (NCA), fell by about 2.5% to 109,978 tonnes. Separate reports by ADM Investor Services and CocoaRadar also described a sharp increase in Asian processing. The October releases therefore challenge a uniformly gloomy account of demand, without supplying a uniformly cheerful alternative.

What the machines measure

Grinding is an important bridge between cocoa grown on a farm and products bought in a shop. Beans become cocoa mass, also called liquor, which can be used in chocolate or pressed into butter and cake. The cake is subsequently processed into powder. Those ingredients have different customers: powder goes into foods including biscuits and desserts, while butter also has uses outside confectionery. More beans entering a grinder therefore need not mean a matching increase in chocolate bars leaving supermarkets.

Nor does the timing match. A processor may grind beans to fulfil orders, rebuild stocks of ingredients or prepare for expected demand. A chocolate manufacturer can, in turn, draw down its own stocks rather than order fresh material immediately. The latest reports measure a real industrial activity, but they do not separate these explanations. Calling the increase a recovery in consumer spending would turn a useful indicator into an answer it cannot provide.

That distinction does not make the figures unimportant for prices. A processor running more beans through its factory has a use for physical cocoa, whatever the eventual destination of its products. If stronger activity persists, it can absorb material that might otherwise remain in storage. But persistence is the test. A temporary rebuilding of inventories can support purchases without establishing a lasting increase in the amount households consume.

Europe's incomplete turn

The European result is more encouraging when read by quarter than by year. Despite the latest improvement, cumulative bean usage for January–September remained 3.1% below the corresponding period of 2025. Earlier weakness had not been erased. The comparison is also with a quarter that was itself below the previous year's level. Growth from that starting point is meaningful, but it is not a return to the industry's earlier trajectory.

There is a geographical problem as well as a timing problem. The ECA measures processing in Europe, not cocoa consumed by Europeans. Chocolate makers can use ingredients processed elsewhere. ADM Investor Services points to the expansion of processing in Côte d'Ivoire as part of the explanation for Europe's longer decline. That is an analyst's interpretation, rather than something the latest European total alone can demonstrate. Nevertheless, it shows why a regional factory series should not be treated as a complete measure of demand.

European, Asian and North American reports describe different industrial footprints and reporting groups. National figures that sit inside a regional total cannot be added to it again. And a factory's location does not identify the final consumer of its output. The right question is whether more beans are being absorbed across the supply chain, not simply which region has produced the most attractive growth rate.

The reporting lists show why the industrial boundary matters. The ECA includes ingredient suppliers such as Barry Callebaut and Cargill alongside branded manufacturers including Ferrero and Nestlé, and says companies report their group bean usage. The NCA's regional definition covers the United States, Canada and Mexico. Its figure is therefore neither a US-only measure nor a measure confined to independent ingredient suppliers. A familiar chocolate brand on a respondent list does not turn factory throughput into retail sales.

A smaller North American sample

North America's decline carries a particularly important qualification. The NCA report contains submissions from 14 plants, against 16 a year earlier. It records less processing in the reported aggregate; it does not provide a comparison restricted to an unchanged set of plants. Without that information, it is impossible to assign the entire decline to weaker activity at continuing facilities. Nor can the missing plants simply be assumed to account for all of it.

The report also excludes melted cocoa liquor and butter. That boundary matters when manufacturers can use semi-finished ingredients as well as grind beans. A fall in the reported bean grind does not establish an equivalent fall in every form of cocoa used by the region's chocolate industry. The sample change and product definition both counsel against turning the North American headline into a sweeping verdict on its shoppers.

Europe has supplied evidence of a turn in factory activity, while North America's smaller sample limits the interpretation of its decline. That is a meaningful change in the information available to bean sellers. It supports a more discriminating view of processing demand; it does not yet establish that households are buying more chocolate.