The Slot, and the Stockholder
Hello Everyone, once again.
It's been a while since I sat down to write one of these properly. Life and this trade have a way of keeping you busy. But a few things have been building in the sesame market over the last months that I've been wanting to put down on paper—mostly to organize my own thinking. So, here it is.
As always, these are just my readings: from the field, from conversations with friends across the trade, and this time, also from a fair bit of data that landed in my lap. Take it for what it's worth, and let's discuss.
Let me start with something I've been chewing on for a while. I was speaking in Istanbul last year, and after my talk, a Brazilian producer asked me a simple question: Now that China has opened up to Brazil, would India still buy Brazilian sesame and could China buy it all?
My answer, I think, surprised him. I said it doesn't really matter how much Brazil can produce. What matters is which slot in China's chain the Brazilian seed fits into and same for India.
The Slot
I've come to think of China not as one market, but as six or seven markets wearing one coat. They import somewhere around 1.2 million tonnes in a normal year (last year wasn't normal, but we'll come to that), and each slice of that has its own quality need, its own price tolerance, its own favorite origin.
We all know the varieties better than I could ever explain them here, but the thing I keep coming back to is how neatly they sort by destination. China takes white natural and crushing for its mills and black for its own trade; Europe, the US, and Australia take the hulled and sortex; and we in India buy raw white for our machines and sell on the finished product. Each of those is really its own distinct market.
The point I made in Istanbul was really about the ceiling on each slot. My sense, after all these years, is that if an origin's slot in China is 300,000 tonnes, dropping your price won't make China take 500,000. The usage is what it is, the blends are set, the downstream contracts are already written. I think this is why our own exports from India have hovered in that 200,000–250,000 tonne band year after year, even when the crop was much bigger. We filled the slot. The slot didn't grow just because our crop did.
Now, plenty of people in our trade see it differently, and they've told so in our groups:
China buys on price, not loyalty, and every buyer has a replacement.
I don't disagree. Price does decide who fills a slot. Where I land slightly differently is that price doesn't seem to decide how big the slot is. In a normal year, one origin slips and another quietly takes its place, and nobody blinks. It's when everyone is short together—which is more or less where we are now—that there's no one left to step in.
There's one more thing sitting quietly under all of this. Most of Africa has always enjoyed duty-free access, and Pakistan secured preferential treatment a few years ago, which predictably drove their volumes up. The simple, underlying point is that Brazil and India carry a significant tariff disadvantage into China compared to the other major exporting countries. It feels like a quiet handicap on those two lanes that has nothing to do with how good the crop is.
Brazil is running into all of this now. Which is really the news of the season.
Brazil — I Asked Them Directly
Rather than add to the guessing, I got on a call with a few Brazilian exporters and simply asked them how it looks on the ground.
What I heard was sobering. The new crop is coming in around 200,000–220,000 tonnes. Against last year, that's close to a 60% fall. One of them said it was shocking even to them.
And it isn't the usual "area is down" story, which is what makes it worth writing about. The area did fall—growers moved to corn and to black matpe after last winter's price crash made sesame look unattractive at sowing. But the ground they gave up was the good ground, the fertile fields with the better rain. Sesame got pushed to the poorer plots and the late window. Then the rains ran late through the soybean harvest and cut off early (as they usually do around the 20th of April), so the late crop got barely forty days of water. Yield came off from about 550 kilos a hectare to around 350. Less area, and less on the area that remained—the two piling on top of each other. That's how you get to minus sixty.
A few other things they told me have stayed with me more than the headline number:
The carryover is basically gone. Maybe 20,000 tonnes, which is nothing. Whatever the new crop is, that's what the world gets; there's no old stock hiding behind it.
They're holding back. Where Brazil shipped 109,000 tonnes in a single month last September, this year they're offering little and waiting to see if their own farmers can even fill the contracts they signed. The sense I got was a trickle over four or five months rather than the flood of last year.
Domestic processing is growing. Three new hulling factories have come up inside Brazil. For the first time, a real slice of the crop will be processed at home instead of shipped raw. For those of us who hull for a living, that's the line I'd underline: Brazil is starting to climb the value chain. Slowly, but it's begun.
One detail explains why Brazilian white walked into China's slot so easily: the cargo is almost always machine-cleaned, purity up near 99.90%. African white is machine-cleaned too, but not always, and often carries a couple of percent of impurity and light seed. So I don't think Brazil won only on price. It won because the seed fit the slot cleanly. Worth remembering when someone calls it pure dumping.
And on the slot argument, they confirmed it without my asking. The big advance commitments went to China; but India, they said, is now paying better than China on spot. Which I read simply: China locked its slot early on forward contracts, and India, arriving late, is bidding up what's left. Price choosing who fills the slot, not making it bigger.
The Rest of the World Looks Thin Too
The wider supply picture is about as tight as I can remember, and India is where I'd start, because I think it's widely misread.
On paper, the government sowing looks alarming, but I've stopped reading it straight. Take a look at this last year IOPEPC data compilation comparing government estimates against remote sensing:
Acreage, Yield & Production Estimates
| States | Govt. Acreage 2024 (Ha) | Govt. Acreage 2025 (Ha) | Remote Sensing Acreage 2024 (Ha) | Remote Sensing Acreage 2025 (Ha) | Yield 2025 (kg/Ha) | Production 2024 (Tonnes) | Production 2025 (Tonnes) |
| Uttar Pradesh | 3,53,328 | 4,39,011 | 2,91,824 | 1,97,857 | 85 | 41,914 | 17,806 |
| Madhya Pradesh | 3,28,800 | 2,27,200 | 2,59,628 | 1,06,032 | 74 | 35,723 | 7,893 |
| Rajasthan | 2,09,804 | 1,92,808 | 1,94,805 | 1,21,260 | 118 | 41,999 | 14,435 |
| Gujarat | 49,400 | 40,100 | 47,847 | 32,199 | 209 | 9,340 | 7,902 |
| Sub Total | 9,41,332 | 8,99,119 | 7,94,103 | 4,57,348 | 1,28,976 | 48,035 | |
| Other States | 1,97,968 | 1,51,881 | 65,464 | 43,931 | 15,502 | 4,614 | |
| Country Total | 1,139,300 | 1,051,000 | 8,59,567 | 5,01,279 | 105 | 1,44,478 | 52,649 |
For your understanding, the national yield of 105 kg/ha recorded here was calculated simply by dividing the total estimated crop by the total area. This means a massive chunk of acreage that got completely washed away is still sitting in that denominator, dragging the math down to an absurdly low average.
This year, assuming the total area remains the same or maybe a little more due to El Niño, and assuming a return to somewhat normal yields, we will at max get around 150,000–160,000 tonnes even if we double the yield and assume that most farms give an output. (Though, of course, that is just an assumption for now).
But the crop isn't really the story this year. The difference from last year is what's around it. Last year, India had a huge import influx from Brazil and Nigeria. That isn't happening this year. The summer crop was revised by IOPEPC to about 125,000 tonnes, arrivals have slowed, and yes, there's stock, but a lot of it has already gone into holding.
Here's the piece I think matters most and gets overlooked. For the last three or four years, there was a reliable release valve on hulled prices. Every time hulled crossed a certain level—usually into the October–December domestic peak—two cheap things landed at once: imports arriving in volume, and the Bengal double-skin natural (the crushing material) coming in cheap. The moment hulled ran up, people simply switched to Bengal double skin for hulling and pulled the price back down.
This year, I don't see that valve. The Bengal crop is barely 80,000–90,000 tonnes and it's expensive. It won't do its usual job of capping the hulling market. And the flood of cheap imports that used to break the rallies isn't lined up either.
So when we come into October with our domestic demand, we'll have a trickle of summer-crop stock plus whatever the new crop gives us. Harvest is running late this year because the monsoon came late, so I don't see real pressure before mid-to-late October. That's where it gets interesting. Early October on, if the material simply isn't showing up, people start to get a little nervous, and that's usually where the game is played.
One honest caveat on our own competitiveness: right now, India isn't competitive globally because freights are very high. But freights are the only thing holding it—on the seed itself, we're still very competitive. The bulk market just isn't here yet. Korea isn't aggressive, Europe is quiet, and the Middle East is running mostly back-to-back because of the war.
But that demand doesn't disappear; it defers. Ramadan falls early next year, around the 8th of February. Goods have to be in position by early January—colliding with our own peak demand and a thin supply side. Ramadan demand is mostly hulled, and that's India, Pakistan, or Africa. A lot of things arriving at the exact same window.
Pakistan: The trade talk and the Chinese buyers put acreage down 15–20%, yet the official Punjab numbers I've seen have the province more or less flat. Both can't be fully right. My hunch is the mood is more bearish than the acreage. What isn't in doubt is the carryover—it is significantly low compared to a few months ago.
Africa: Thin too. Sudan well down and only a fraction leaving for export. West Africa had people puzzled back in February—peak arrival season and the cargo had simply vanished. Tanzania and Mozambique both short of last year. Carry overs nothing significant in either too. Nigeria/Togo/Burkina/Mali and even Somalia , not one origin is stated to say they are sitting on any significant stock.
Put it together, and there's no comfortable cushion anywhere. In a picture like that, the old rule I keep coming back to holds:
One origin short is worth maybe 10% on price; more than one at once, call it 25%.
The Part I Keep Turning Over
Here's what I can't stop thinking about:
If every origin is empty, why is the biggest buyer sitting on more stock than it has ever held?
I've had three years of weekly Qingdao port numbers to look at, and what they show is genuinely striking. Around 132,000 tonnes at the 2023 peak, 229,000 in 2024, 315,000 in 2025, and this June nearly 421,000—the highest I've ever seen it. Then it fell back toward 370,000 within a week or so. (That swing, by the way, is why I've never fully trusted any single port figure. I read it as a range and a direction, not a fixed point.)
The mix has flipped too. Three years ago, Pakistani and East African seed filled the port; today it's mostly West African white, with Pakistan and Brazil already drawing down. And the origins doing the filling are, largely, Mozambique and Tanzania .
So why does China hold so much? For as long as I've been in this trade, carryover lived at the origin. India held a buffer, Sudan held a buffer. When China imported 800,000–900,000 tonnes a year, they held maybe 180,000–200,000 at port and no more. That's changed. Last year China imported a record 1.46 million against consumption of maybe 1.2. That surplus didn't disappear. It's the pile at Qingdao.
The thought I keep landing on is that China has quietly become the carryover holder for the whole world. The buffer that used to sit at origin now sits at destination, financed cheap. When I look at that 400,000-tonne number, I'm not sure I'm looking at hungry demand—I think I might be looking at the world's carryover, just moved to one place.
But here is where the suspense really sets in for this storyline: What if the replacement supplies this year to China simply do not add up to their baseline consumption, and that port stock starts depleting? What really happens when China's port stock falls below 300,000 tonnes and continues a downward spiral, with no origin sitting on huge, sudden supplies ready to replenish that pile? Does panic set in with high-price buying, or does China sit back and stubbornly let its ultimate stock advantage wither away? It adds a layer of genuine suspense to how this season will play out.
Where I Think Prices Sit
We're at or near the median level that's held for the better part of two decades—the kind of level where, in my experience, this market has usually found its floor and turned. Those of you who've read these reports over the years know how much weight I put on these benchmarks; they've rarely let me down.
The base has already lifted, but I think the upside is capped, and the cap is that port stock. While 400,000 tonnes sits at Qingdao, sharp China-led rallies keep getting sold into. The most experienced Chinese voice I heard put the 2026 average at $1,200–$1,400 even with production cuts, because the stock is a shock absorber: below $1,000 China's appetite turns near-bottomless; above $1,400 the stock comes out to cap it. That feels about right to me as a band.
The real triggers from here are narrow: a genuine shortfall in Tanzania or Mozambique, the Indian winter crop running into weather, more Brazilian yield stress or a shocker from Pakistan, or the day China decides to rebuild rather than run its port stock down.
But if I step back from the individual pieces, here's where I land. I don't think we're oversupplied this year. The mechanisms that used to pull this market down just aren't lined up. Put that next to a thin supply side almost everywhere, a late Indian harvest, and the Ramadan and domestic peaks colliding around the turn of the year, and it feels less like a comfortable market than a quiet one waiting on a cue. The current lull could hold, or it could flip the other way fairly quickly once people realize the material simply isn't showing up.
Looking at the supply side we've walked through here, I don't think this is a market that's short of seed so much as one that's short of clarity on who needs it, and when. So that's the game I'd play.
That's all from me for now. As always, these are just one man's readings—from the field, from good friends across the trade, and this time from the data too. I've been wrong before and said so, and I may be wrong again—though I've been right more often than not, and the old blog is all still there to confirm or counter that. Weigh it against your own view, and let's keep talking.
Stay well, and good luck with your covers.
— Mukul Gupta
