Back
Oct 23, 2025
By ASA Chief Economist Scott Gerlt
The world’s largest soybean buyer needs soy to produce high-protein feed to sustain its massive hog population, which represents almost 60% of the global sow herd, and feed a large population hungry for protein. Simply put, China needs soy, and the United States soy industry has depended on access to this market for decades.
Despite this, China has not purchased any of the U.S. soy crop being harvested now. The new crop marketing year started on Sept. 1. Due to the government shutdown, the most recent data on soybean sales is from Sept. 18. At that point, for the first time in the USDA dataset starting in 1999, China has not purchased any of the new crop U.S. soy. Last year they had purchased 6.5 million metric tonnes (MMT) by that time.
Additionally, export bookings to unknown destinations are down 1.8 MMT year over year. Some of China’s orders originally appear in this category, so the change reflects the larger lack of China purchases. Exports booked to known destinations besides China are up 1.5 MMT year over year, but that is a fraction of the missing export bookings. Last marketing year, 32% of annual U.S. soybean exports had been ordered by mid-September.
China’s absence in the U.S. soy market leaves a hole for both countries. China needs soy, but does it need U.S. soy right now? South America starts its soy harvest in late January, and the crop will start arriving in earnest in China by March. This article examines whether China might be able to meet domestic crush needs until this time without U.S. soy.
Brazil is the world’s largest producer and exporter of soybeans. The country continually expands soybean area and produced a record soybean crop yet again in 2025 of 169 MMT according to USDA. Being in the Southern Hemisphere, Brazil’s crop season is counter-cyclical to the United States with planting occurring in the fall and harvest in the spring. This results in its exports spiking in March when the new crop becomes available.
This seasonal pattern continued in 2025, but the tail off in exports to China has been much slower than usual (Figure 1). Brazil’s large harvest coupled with likely stockpiling by China to buffer the trade war has led to the longer export season. The projections in Figure 1 are based on projections by Brazil exporters and are for total soy exports. Given that Brazil’s soybeans are selling at a premium to U.S. beans, it is quite likely that China is the overwhelming buyer as Brazilian beans are still cheaper for China than U.S. soy when the Chinese retaliatory tariffs are considered.
Like Brazil, Argentina’s exports to China remain well above average in 2025 (Figure 2). While Argentina is the third largest soybean producer in the world, the country typically exports soybean meal and oil as opposed to whole beans. This is encouraged by Argentina’s export tariffs of 26% on whole beans and 24.5% on meal and oil.
Argentina suspended the export tariffs effective Sept. 23 until Oct. 31 or $7 billion of exempt ag products had been sold. The financial limit was reached on Sept. 24, only one day later. According to the Argentina’s Ministry of Economy export license applications, there were 103 soybean export commitments submitted for a total of 5.1 MMT during the two days of eligibility. As with Brazil, it was assumed for this analysis that export commitments from Argentina are bound for China given its need for soy.
While South America has been exporting to China at a higher than normal pace this year, their ability to continue to do so is an important factor in determining China’s unmet soy needs. To analyze this, several sources of data were considered in addition to the export numbers above. USDA’s Oct. 1 stock numbers were used as the starting point of current supply in each country. USDA’s marketing year crush number was converted to monthly crush by adjusting based on the share of marketing year crush per month in historical Fastmarkets data.
Figure 3 contains the results for Brazil. USDA estimates the country started October with 36 MMT on hand. Given the monthly export projections and estimated monthly crush, Brazil would enter its new crop harvest around late January/early February with 3.8 MMT on hand, which is less than one month of crush needs. For comparison, USDA forecasts that Brazil will have 5.4 MMT of soy stocks on hand at the beginning of January and CONAB (Brazil’s agricultural data provider) estimates 10.7 MMT at that time. The results of this analysis at 7.4 MMT is squarely in the middle of that range. This indicates that Brazil has little ability to export more soy beyond what is projected through the end of 2025.
It is possible Brazil could reduce its domestic crush rate to satisfy export orders. Through late October, Brazil’s crush margins stood at -$5/MT, suggesting export purchases were likely to command Brazilian soybean supplies instead of domestic crush. While this analysis does not account for any reduction in Brazilian crush volumes, it is an option which can be pursued by the Brazilian soy industry.
While Brazil imports almost no soy, Argentina does import a significant volume, so historical monthly import rates were added to the estimated soy availability for the country. Given Brazil has few soybeans left to sell, the historical Brazilian sales were not included, but these have been small the past two years.
Imports from the United States were also not included, but they have been almost zero recently. Sales beyond what is assumed could obviously provide Argentina with more exportable supply. In fact, during the last U.S.-China trade war, Argentina did buy much higher volumes of soybeans from the United States. An additional export tariff holiday by Argentina would allow the country to capture economic rents from being able to buy U.S. soy for domestic needs while exporting its own beans to China without the retaliatory tariffs. Brazil could do the same. We have seen them import from the U.S. when they have needed supplies.
Figure 4 shows the results assuming Argentinian soy import patterns from the past two years. USDA estimated that Argentina started October with 24 MMT tons of soy on hand. Argentina harvests soy later than Brazil, often starting around April. The results of this analysis indicate that Argentina will have 4.7 MMT of soy at that point, which is about one to two months of crush needs. USDA projects 6.8 MMT of ending stocks at the beginning of April, which is almost identical to USDA reported stocks the prior two years. The numbers point to limited ability by Argentina to book more exports without also increasing imports.
Brazil and the United States accounted for 92% of China’s soy imports in 2024. Outside of the United States, Brazil and Argentina options are limited for sourcing soybeans. China has been sourcing increased levels of soy from some other countries such as Uruguay, Russia and Canada, but the amounts tend to be small.
USDA estimates China’s soybean crush will be 108 MMT in the 2025/26 marketing year. Figure 5 accounts for the seasonality of crush by adjusting with historical Fastmarkets data. Data indicate negative crush margins in China at the present time, which may slow actual crush rates below assumed amounts.
The Brazil and Argentina data was converted to Chinese imports by assuming 1.5 months between South American export and unloading in China. Under these assumptions, China can supply its soy needs by imports through November. The reduction in pace of booked exports later in the year allows crush to exceed imports in Dember 2025 and January 2026, although not by large amounts.
Constructing balance sheets for China is difficult given the opaque market data. The country’s soybean stocks are not known with any amount of certainty. For this reason, Figure 6 instead compares China’s projected Brazil and Argentina imports against crush needs until the new South American crop begins to arrive in April. China’s domestic soy harvest is almost entirely destined for its food market as opposed to the feed market, which makes it largely irrelevant in this context.
As previously mentioned, China’s soy needs largely appear filled through the end of the year. Normally, U.S. supplies start to land in the country at about this time and taper off after March. Without these supplies, China is projected to be 19.2 MMT short for its domestic crush needs before the new South American crop arrives. In the last complete marketing year with data (2024/2025), China reported soy imports from the United States of 26.1 MMT – slightly above the 19.2 MMT deficit.
USDA estimates China had 43.5 MMT of soybean stocks on hand on Oct.1. If this amount is correct, China could cover its soy needs with old crop stocks and state reserves until the South American harvest arrives next spring. Yet the uncertainty around China’s stock levels and policy preferences to tap into state reserves allows only speculative conclusions.
Other options exist to fill the gaps including reducing crush rates, reducing soy inclusion rates in livestock rations, finding more soy from non-U.S. sources, importing soybean meal or importing alternative feedstocks. These carry their own set of issues such as ensuring adequate feed supply, increased costs and idling domestic facilities. As has been shown in many economics classes, free trade would allow cheaper and more efficient outcomes for both the buyer and seller.
