TURF WAR:
Mind the soybean gap
The tangled web woven by trade wars/truces and shutdowns continues to complicate tracking oilseed SnDs. The US soybean export rate is likely too slow for the USDA’s current forecast. But by how much?
TURF WAR: Mind the soybean gap
The tangled web woven by trade wars/truces and shutdowns continues to complicate tracking oilseed SnDs. The US soybean export rate, based on any ‘normal’ pattern, is too slow for the USDA’s current forecast (44.5mmt). The trade war and truce, though, obviously mean that season 2025 will be anything but normal. Therefore, templates can be used as a starting point, but they are much less useful than usual. Degraded templates mean that closer analysis and more guesstimation are required. And even then, the problem cannot be completely overcome; we will have to wait for the pattern to emerge. And that pattern is very much price-dependent.
Comparisons with last season are of limited use. Template degradation also applies here. And, the US has fewer soybeans to export. The USDA forecasts US soybean exports at about 44.5mmt in 2025, about 6.5mmt less than in 2024. Or, to flip the logic, export sales 6.5mmt behind last year is sufficient to meet that forecast.
There are multitudes of templates that might be used. One template export rate uses the median share of exports for that season week*. The charts below use that along with a default 55/45% split China/non-China. On this method, export sales commitments are about 9.7mmt behind the rate needed to meet the USDA’s forecast. (Split: China -15.3mmt, non-China +5.6mmt). Subtracting the USDA’s large sales reports reduces the deficit to about 5.7mmt for the most recent data week (13 Nov).
TEMPLATE CASE
However, that ‘deficit’ is largely an elaboration of the template’s assumptions, namely:
- the USDA’s export target,
- the China/non-China split, and
- the weekly sales pattern through the year.
None of those are fixed – they are all actually variables. And they are largely price-determined. Therefore, we can change these variables to explore the implications. The export target is the easiest to change, but is also the least interesting. Making different assumptions about the variable we are attempting to forecast is a sterile exercise. More useful is to vary the export sales’ China/non-China split and seasonal pattern.
Many different China/non-China splits are possible. One interesting possibility is, if China buys only 12mmt (27.5%), are non-China sales on track to meet target exports? The charts below show this case using the median shares (as of Nov 13). Under this scenario, China, including large sales, still has around 8.5mmt to meet its commitment. And non-China sales, including large sales, are about 1.5mmt behind the required rate. That non-China gap is sizeable, but not unrecoverable. Season 2025 will be an abnormal year. Therefore, there is substantial latitude for the sales pattern to deviate from our (or any other template).
LOW-CHINA CASE
US soybean prices will be important to whether that non-China gap closes. The worry, as soybean prices increased sharply, was that non-China sales would drop sharply. The data show some signs that non-China sales slowed since prices reached new highs in late October. And there will now be another four export sales reports for the higher-price period. The non-China gap may widen during that period if sales slow. Not until after Christmas will export sales data cover prices returning to pre-trade truce levels. Importantly though, prices falling well before China has finished its commitment buying is a plus. The US seemed to extend China’s window several times. Those extensions risked an extended high-price period that, in turn, extended the period of slower non-China sales.
* Specifically, the median share of export commitments for week X for seasons 2010-24, ex. 2013, 2017, 2018. Two sets of shares are generated, one for China and one for non-China sales.
