@sdelsad/commodity-desk-daily 1.0.6 → 1.0.7

package/covered.md

@@ -3,4 +3,4 @@
 Running log. Read before writing a new episode: avoid repeating material, and only make callbacks to episodes listed here.
 
 - **Ep 1** (Mon) — *What a Commodity Merchant Actually Does*: What a merchant does: three transformations (space/time/form); risk absorber with a balance sheet; ABCD + COFCO + Viterra/Bunge; physical vs paper, paper is the hedge not the bet; flat price killed by hedge, profit lives in differentials; asset-heavy = options + information machines; 1851 Louis-Dreyfus Alsace-Basel origin story. Vocab: flat price, basis, book, the screen, origination, execution, ABCD. Example: 66,000 t Santos->Qingdao cargo, +80 in / +175 out, freight 70, costs 10 = 15c/bu ~ $5.50/t ~ $360k, direction-neutral.
-- **Ep 2** (Tue) — *Flat Price vs Basis*: Flat price vs basis: cash price = futures + basis; quoting 'plus 80'; desk kills flat price via hedge; long the basis (physical + short futures) vs short the basis; basis moved by logistics, quality, urgency, farmer selling; basis risk as the chosen risk. Vocab: flat price, cash price, differential, plus eighty, hedged position, basis risk. Example: 66,000 t Santos cargo bought at Nov +80 — board -$1 hedged to zero (~$2.4M each way) vs +10c basis = ~$242k kept.
+- **Ep 2** (Tue) — *Flat Price vs Basis*: Flat price vs basis: cash = futures + basis; quoting 'November plus 80'; desk kills flat price via hedge; long the basis (physical + short futures) vs short the basis (sold unowned + long futures placeholder, crusher example); basis moved by freight, quality, congestion, urgency, farmer selling; basis risk as the chosen, analyzable risk. Vocab: flat price, cash price, differential, plus eighty, hedged position, long/short the basis, basis risk. Example: 66,000 t Santos cargo at Nov +80 — board - hedged to zero (~.4M each way) vs +10c basis = ~40k kept.

package/ep02.md

@@ -0,0 +1,126 @@
+# Commodity Desk Daily — Episode 2: Flat Price vs Basis
+
+*Tuesday, August 11, 2026 · ~10 min listen*
+
+## Key takeaways
+
+- Every physical price is **two numbers added together**: cash = futures + basis. The futures leg is the world price — public, violent, seen by everyone at once. The basis is the **local price of reality**: freight, quality, congestion, urgency, farmer selling.
+- Physical offers are quoted as a differential — "**November plus 80**" — not as a full price. Both sides assume the futures leg because both can hedge it in one click; the only number actually negotiated is *the plus*.
+- A desk **kills the flat price within minutes** by selling futures against every physical purchase. What remains is a basis position — the risk the desk *chose* to keep.
+- **Long the basis**: own physical, hedged with short futures — you win if the differential strengthens. **Short the basis**: sold physical you don't yet own, holding long futures as a placeholder — you win if the differential weakens before you buy.
+- The worked cargo: 66,000 t of Brazilian beans (~2.4M bu) bought FOB Santos at November +80, hedged. Chicago falls $1: beans lose ~$2.4M, the short hedge makes ~$2.4M — **net zero**. The differential moves +80 → +90: 10¢ × 2.4M bu = **$240k of real, banked P&L**.
+- Basis moves on **logistics, quality and urgency** — the things elevators, vessels and relationships see before any screen does. That's where physical information gets paid (Episode 1's "information machines").
+- **Hedged does not mean safe.** The differential can move against you: 10¢ the wrong way on that cargo is a $240k loss, hedge or no hedge. That is *basis risk* — smaller than flat-price risk, local, analyzable. You don't escape risk in this business; you choose it.
+
+## Vocabulary
+
+| Term | Desk meaning |
+|---|---|
+| Flat price | The full outright price level (futures + basis together) |
+| Cash price | The price of the real, physical commodity in a real place |
+| Basis / differential | The premium or discount over a named futures month ("plus 80") |
+| "November plus 80" | Quote convention: 80¢/bu over November futures |
+| Hedged position | Physical position with the futures leg sold (or bought) against it |
+| Long the basis | Own physical + short futures: profit if the differential strengthens |
+| Short the basis | Sold physical not yet owned + long futures: profit if it weakens |
+| Basis risk | The residual risk of the differential moving against a hedged position |
+
+## Market pulse (Monday Aug 10 close — eve of WASDE)
+
+Grains are marking time ahead of **Wednesday's August WASDE**, the first with survey-based yields. The trade expects corn near **182.5 bpa** (−0.5 from July — still a ~15.95bn bu crop, second-largest ever) and soybeans near **52.9 bpa**. Monday was quiet: September beans slipped about a cent, meal eased, oil firmed; demand support came from China booking **238,000 t of US beans** and ~105,000 t of corn. Wheat is the live story: Russia's harvest is only **46% complete** — the slowest pace in five years — with Russian offers around **$223/t** and August exports running below the usual pace as Azov Sea shipping is disrupted by Ukrainian strikes; Matif firmed. Weather: rain reached the northern Corn Belt, the south stayed hot and dry, and France's maize crop is rated its **worst since 1980**.
+
+---
+
+## QUIZ
+
+### Block A — Today (Ep 2: flat price vs basis)
+
+**A1 — Decompose the month.** A desk buys 30,000 t of corn (≈1.18M bu) from an interior elevator at "December futures minus 10" and sells December futures against it the same hour. A month later, December corn has rallied 50¢, and the desk sells the corn to an exporter at "December minus 2". Separate the flat-price P&L from the basis P&L, in ¢/bu and in dollars. Which number was the desk's actual trade, and what was the 50¢ rally to them?
+
+**A2 — Which offer do you lift?** You buy soybeans for a crusher. Two firm offers for the same Santos October boat arrive at 9:00 with November futures at $11.80: Exporter X offers "November plus 85"; Exporter Y offers flat $12.70. By 11:00, November has dropped 25¢ and both offers are still on the table, unchanged. Which offer is cheaper at 9:00? At 11:00? Explain which seller is carrying flat-price risk while the offers sit, and what that tells you about why the physical market quotes in basis terms.
+
+**A3 — Name the position.** (a) A merchant holds 40,000 t of unsold wheat in a port silo, fully hedged with short futures. (b) A miller has sold flour forward for Q4, owns no wheat, and holds long futures as a placeholder. Freight rates out of that port suddenly spike and export premiums jump 12¢. For each player: long or short the basis? Who gained, who lost, by how much per bushel — and why does the direction of the futures market not appear anywhere in your answer?
+
+### Block B — Episode 1 (what a merchant actually does)
+
+**B1 — The recruiter's question.** "Commodity traders bet on prices going up, right?" Give the desk-level correction in three moves: what a merchant is actually paid for (name the three transformations with one concrete example each), what paper is for, and where the profit therefore lives.
+
+**B2 — Rerun the cargo.** Episode 1's cargo: buy FOB Santos at futures +80, sell delivered Qingdao at futures +175, freight 70¢, execution 10¢, 66,000 t ≈ 2.4M bu. (a) Recompute the net margin in ¢/bu and dollars. (b) During the voyage, Chicago *rallies* 90¢ instead of falling. A colleague says the desk "left $2M on the table by hedging". What did the hedge actually cost or save, and why is the colleague's framing the wrong way to run a merchant book?
+
+*(No J-3 block: three episodes back would be Episode −1.)*
+
+---
+
+ 
+
+ 
+
+ 
+
+## ▼ SOLUTIONS (spoilers) ▼
+
+**S-A1.** Basis P&L: bought at −10, sold at −2 → the differential appreciated 8¢. On ~1.18M bu that is ≈ **$94k**. Flat-price P&L: zero by construction — the 50¢ rally lifted the physical corn by 50¢ (+$590k on the cargo) and cost the short futures exactly the same (−$590k). The desk's actual trade was *the basis*: buy the differential at −10, sell it at −2. The 50¢ rally was noise passing through a hedged book — a cash-flow event on margin (Episode 3's subject), not a P&L event. The trap: crediting the rally to the trader. A desk that "made money because corn rallied" wasn't hedged — and that's a different job.
+
+**S-A2.** At 9:00, X works out to 11.80 + 0.85 = $12.65 against Y's flat $12.70 — **X is 5¢ cheaper**. At 11:00, X's offer has fallen with the board to 11.55 + 0.85 = $12.40, while Y is still $12.70 — **X is now 30¢ cheaper**. While the offers sit, Y is the one carrying flat-price risk: a flat offer is an implicit bet that the board won't fall before someone lifts it (in a falling market it becomes more and more expensive relative to replacement, and nobody lifts it; in a rallying market it gets lifted instantly — adverse selection both ways). X's exposure is only the basis component. That asymmetry is exactly why the physical market quotes "plus 85" and not $12.65: it lets an offer stay firm for hours while the world price does whatever it wants.
+
+**S-A3.** (a) Physical + short futures = **long the basis**. (b) Sold product forward, long futures placeholder, still needs to buy physical = **short the basis**. Export premiums jump 12¢: the merchant's differential appreciated → **gains 12¢/bu** on the tonnage (≈1.47M bu on 40,000 t of wheat → ≈ $176k). The miller must now pay 12¢ more over futures to get real wheat → **loses 12¢/bu** on what remains to buy. Futures never enter the answer because both players hedged the flat price away on day one — what was left in both books was pure differential, and the differential is what moved. One event, two mirror-image P&Ls: that is basis as a market of its own.
+
+**S-B1.** (1) Merchants are paid to *transform*, not predict: **space** (move Mato Grosso beans to a Shandong crusher — Santos → Qingdao), **time** (buy at harvest glut, store, sell into spring scarcity — paid via carry), **form** (crush beans into meal + oil; blend two off-spec wheats into one on-spec cargo). (2) Paper (futures/options) is the *hedge*: it cancels the price risk of physical positions rather than expressing views — the flat price is killed within minutes. (3) So profit lives in the **differentials** — the margins on each transformation, like Episode 1's 15¢/bu Santos→Qingdao cargo ≈ $360k, earned with zero opinion on direction.
+
+**S-B2.** (a) Gross 175 − 80 = 95¢; net 95 − 70 − 10 = **15¢/bu** ≈ $5.50/t ≈ **$360k** on 2.4M bu. (b) The hedge "cost" ~$2.16M on the futures leg (90¢ × 2.4M bu) — and the physical beans *gained* the same ~$2.16M. Net effect on the book: zero; the $360k came through untouched. The colleague is comparing the hedged book to a naked long — but a naked long is a flat-price bet the desk never had a mandate (or edge) to run, and the same logic in a falling market means ruin: −90¢ unhedged is −$2.16M against a 360k margin. Merchant P&L must be repeatable and direction-neutral; "what if we hadn't hedged" is a casino counterfactual, not attribution. (Attribution done properly gets its own episode.)
+
+---
+
+## The episode, in writing
+
+### The number on the screen is not your price
+
+On Monday, soybeans in Chicago drifted lower — and a desk sitting on 66,000 tonnes of soybeans didn't care. Not out of recklessness: the number on the screen simply is not the price of their beans, and never was. Understanding why is the mental model at the heart of the physical trading job.
+
+Every physical price in this business is two numbers added together:
+
+> **cash = futures + basis**
+
+The futures leg is the world price. It carries the big story — crop sizes, weather, funds, war. It is violent, public, and everyone on earth sees it at the same instant. The basis is the **local price of reality**: it prices what the screen cannot see — freight, quality, port congestion, how badly a buyer needs beans in October rather than January, whether farmers are selling or sulking.
+
+### "November plus 80"
+
+Listen to how soybeans are actually offered in Santos. Nobody says "$12.60". They say **"November plus 80"** — 80 cents a bushel over the November CBOT contract. The full offer moves all day as Chicago moves; the *plus 80* barely moves at all. That differential is the **basis**.
+
+A real conversation goes: *"Where are Santos beans for October?" — "Plus eighty."* Not a full price — just the basis. Both sides assume the futures leg, because both sides can hedge it in one click. The only number actually being negotiated is the plus.
+
+### Killing the flat price
+
+Here is what a desk does with that split: it kills the futures leg, within minutes. Buy a cargo of physical beans and, before the coffee goes cold, sell futures against it. The flat-price risk is handed to the screen, where thousands of speculators are happy to hold it.
+
+What remains is the part the desk *chose* to keep — and it has a direction, like any trade:
+
+- **Long the basis** — own physical, hedged with short futures. You want plus 80 to become plus 90: a bet that real beans, in that place, at that time, get scarcer relative to paper.
+- **Short the basis** — you've sold a cargo you don't yet own and hold long futures as a placeholder. You want the differential to weaken before you buy the physical. A crusher who has sold meal forward but hasn't bought beans is short the basis every day of the week.
+
+Notice what is missing from both phrases: any opinion about whether the market goes up.
+
+### One cargo, two P&Ls
+
+Take yesterday's cargo: 66,000 t of Brazilian beans — about **2.4 million bushels** — bought FOB Santos at November +80, hedged with short November futures. Over three weeks, Chicago falls a full dollar and the Santos differential moves from +80 to +90.
+
+| Leg | Move | P&L |
+|---|---|---|
+| Physical beans | board −$1.00 | −$2.4M |
+| Short futures hedge | board −$1.00 | +$2.4M |
+| **Flat price, net** | | **$0** |
+| Basis: +80 → +90 | +10¢ × 2.4M bu | **+$240k** |
+
+A dollar of flat price came and went and the book barely noticed. A quiet ten-cent move in the differential was the entire profit — real money, banked. That is the anatomy of a physical trade.
+
+Why did the basis move? Maybe freight tightened. Maybe Chinese crushers turned urgent. Maybe Brazilian farmers stopped selling because prices in reais looked ugly. All local, all physical, all invisible on the screen — and all things that Episode 1's "information machines" (elevators, vessels, relationships) see before any index prints. Nobody has an edge on the flat price, the most public number on earth. On the basis for beans, in Santos, for October? A desk absolutely can.
+
+### Hedged does not mean safe
+
+The differential can move against you: ten cents the wrong way on that cargo is a $240k loss, hedge or no hedge. That risk has a name — **basis risk** — and it is the risk the desk keeps *on purpose*. The point of the hedge is not to remove risk; it is to swap a huge risk you cannot analyze for a small one you can. You don't escape risk in this business. You choose it.
+
+### Market pulse recap
+
+See the pulse section above: WASDE Wednesday (corn ~182.5 bpa expected, beans ~52.9), China booking US beans and corn, Russia's slowest harvest pace in five years with Azov shipping disrupted, and France's worst maize rating since 1980.
+
+*Tomorrow — Episode 3: Futures, desk edition. Not pricing theory — plumbing. Which contracts, which months, how many lots hedge a real cargo, and what a margin call does to your morning.*

package/feed.xml

@@ -18,6 +18,14 @@
       <title>Commodity Desk Daily</title>
       <link>https://www.npmjs.com/package/@sdelsad/commodity-desk-daily</link>
     </image>
+    <item>
+      <title>Ep 2 — Flat Price vs Basis</title>
+      <description>Why a physical desk kills the flat price within minutes, and what remains: the basis. A Santos cargo where a one-dollar board move nets to zero and a quiet ten-cent differential move is the entire profit.</description>
+      <enclosure url="https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/ep02.mp3" length="6639596" type="audio/mpeg"/>
+      <guid>https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/ep02.mp3</guid>
+      <pubDate>Tue, 11 Aug 2026 05:00:00 GMT</pubDate>
+      <itunes:duration>553</itunes:duration>
+    </item>
     <item>
       <title>Ep 1 — What a Commodity Merchant Actually Does</title>
       <description>Merchants are not paid to predict prices. Space, time and form — the three transformations — and one Santos-to-Qingdao cargo that makes $360k with no opinion on price direction.</description>
@@ -26,13 +34,5 @@
       <pubDate>Mon, 10 Aug 2026 05:00:00 GMT</pubDate>
       <itunes:duration>626</itunes:duration>
     </item>
-    <item>
-      <title>Ep 2 — Flat Price vs Basis</title>
-      <description>Why a physical desk kills the flat price within minutes, and what remains: the basis. Long the basis, short the basis, and a Santos cargo where the screen bleeds $2.4M while the book makes $242k.</description>
-      <enclosure url="https://cdn.jsdelivr.net/npm/@sdelsad/commodity-desk-daily@1.0.5/ep02.mp3" length="7273197" type="audio/mpeg"/>
-      <guid>https://cdn.jsdelivr.net/npm/@sdelsad/commodity-desk-daily@1.0.5/ep02.mp3</guid>
-      <pubDate>Tue, 11 Aug 2026 05:00:00 GMT</pubDate>
-      <itunes:duration>606</itunes:duration>
-    </item>
   </channel>
 </rss>

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@sdelsad/commodity-desk-daily",
-  "version": "1.0.6",
-  "description": "Commodity Desk Daily - Ep 1: What a Commodity Merchant Actually Does",
+  "version": "1.0.7",
+  "description": "Commodity Desk Daily - Ep 2: Flat Price vs Basis",
   "license": "CC-BY-4.0",
   "keywords": [
     "podcast",

package/ep01.md

@@ -1,116 +0,0 @@
-# Commodity Desk Daily — Episode 1: What a Commodity Merchant Actually Does
-
-*Monday, August 10, 2026 · ~10 min listen*
-
-## Key takeaways
-
-- A merchant is **not** paid to predict prices. The job is transforming commodities across three dimensions: **space** (move it to where it's worth more), **time** (store it from surplus to scarcity, paid via carry), and **form** (crush, blend, refine it into what customers actually buy).
-- The merchant is a **risk absorber with a balance sheet**: the farmer doesn't want to carry price risk for six months, the crusher needs exact tonnage on exact dates — the margin pays for absorbing everything they don't want (logistics, timing, quality, price risk).
-- The historic big four of grain trading are the **ABCD**: ADM, Bunge, Cargill, (Louis) Dreyfus — joined today by COFCO, China's state trader, and Viterra, which merged with Bunge in 2025. The business runs on massive volumes and razor-thin margins: 1–2% net in a good year.
-- **Physical vs paper**: physical is real cargoes with quality certificates and vessels; paper is futures and options — "the screen". Merchants trade huge volumes of paper, but to *hedge* physical positions, not to speculate. Paper cancels risk; it doesn't take it.
-- Because the flat price is hedged from day one, the profit lives entirely in the **differentials**. The worked cargo: buy FOB Santos at futures +80¢/bu, sell delivered Qingdao at futures +175¢ → 95¢ gross − 70¢ freight − 10¢ execution = **15¢/bu ≈ $5.50/t ≈ $360k on a 66,000 t cargo** — with zero opinion on price direction.
-- The space transformation is the industry's oldest: in **1851**, seventeen-year-old Léopold Louis-Dreyfus carted Alsace wheat across the border to Basel. Today the cart is a 66,000-tonne vessel and the road is Santos → Qingdao. Same trade.
-- **Asset-heavy beats asset-light** in two ways: assets are *options* (your terminal loads your cargo at cost exactly when capacity is scarcest) and *information machines* (elevators and vessels see the flows before the screens do).
-
-## Vocabulary
-
-| Term | Desk meaning |
-|---|---|
-| Merchant / trading house | Firm that buys, moves, stores, transforms and sells physical commodities |
-| ABCD | ADM, Bunge, Cargill, Louis Dreyfus — the historic big four of grain trading |
-| Physical | The real commodity: cargoes, silos, quality specs, vessels |
-| Paper | Futures & options — standardized exchange contracts |
-| The screen | Desk shorthand for the futures market and its visible prices |
-| Flat price | The outright price level (e.g. the CBOT futures price) |
-| Basis | The local premium/discount over futures for real goods in a real place (Episode 2's subject) |
-| Origination | Buying from the producer end: farmers, co-ops, country elevators |
-| Execution | Everything after the trade: vessels, documents, surveyors, discharge |
-| The book | A desk's full set of positions, physical and paper together |
-| Asset-light / asset-heavy | Renting the supply chain vs owning elevators, ports, plants, vessels |
-
-## Market pulse (as of Friday Aug 7 close)
-
-Wheat led the complex into the weekend: KC September HRW +14¼¢ to $7.14, Chicago September SRW near $6.40, on Black Sea tension and firmer energy. Corn was pinned — Sep $4.39, Dec $4.62 — with the market waiting for **Wednesday's August WASDE**, where analysts expect a corn yield near 182.4 bpa. Soybeans drifted to ~$11.59 (Sep); China bought ~8.7M bu of beans and Mexico ~11.3M bu of corn. In softs, arabica sits near $3.15/lb with ICE-certified stocks at 2½-year lows; raw sugar trades around 16.5¢/lb.
-
----
-
-## QUIZ — Episode 1 (today). No J-1 / J-3 blocks: this is Episode 1.
-
-**Q1 — The confident analyst.** A desk's research team becomes convinced — for good, well-documented reasons — that soybeans will rally $1 over the next quarter. A junior proposes: "Simple: buy futures and wait." Why is that *not* what a merchant does, and what would a physical desk actually do with that same view? Name at least two concrete expressions of the view that stay inside the merchant business model.
-
-**Q2 — Price the cargo.** A desk can buy soybeans FOB Paranaguá at November futures +65¢/bu and sell them delivered to a crusher in Vietnam at November futures +170¢. Ocean freight on that route costs the equivalent of 82¢/bu; port and execution costs 11¢. (a) Compute the net margin per bushel, per tonne (≈36.7 bu/t), and for a 66,000 t cargo. (b) The desk hedges on day one; during the voyage CBOT falls 80¢. What happens to that margin, and why?
-
-**Q3 — The rented edge.** An asset-light startup pitches: "We can do everything the big houses do — we'll rent elevator capacity, charter vessels voyage by voyage, and buy market data." Based on today's episode: name the two advantages of owned assets that renting cannot fully replicate, and — to be fair — one real advantage the asset-light firm genuinely has.
-
----
-
- 
-
- 
-
- 
-
-## ▼ SOLUTIONS (spoilers) ▼
-
-**S1.** Buying futures outright is a flat-price bet — the one game where a merchant has no structural edge: it's the most crowded, most liquid, most analyzed number on earth, and betting it puts the firm in competition with funds built for exactly that. It also isn't what the margin machine is for: merchant P&L comes from transformations, hedged. Legitimate expressions of a bullish view inside the model include: (1) originate more aggressively now — buy more physical at today's differentials (hedged as always), so the book is positioned for the demand that a rally implies; (2) time the *hedge placement and structure* within risk limits (e.g. which month to sell, when to roll) rather than running naked length; (3) buy storage/carry positions or secure logistics capacity that becomes more valuable if the market tightens the way research expects. The trap: "bullish" for a merchant should change *which transformations you do*, not turn the firm into a fund.
-
-**S2.** (a) Gross: 170 − 65 = 105¢. Net: 105 − 82 − 11 = **12¢/bu**. Per tonne: 12¢ × 36.7 ≈ **$4.40/t**. Cargo: ≈ $4.40 × 66,000 ≈ **$291k**. (b) Essentially nothing happens to it. Both legs are priced *against futures*; the 80¢ fall hits the physical purchase and the short futures hedge equally and oppositely (≈ $1.9M each way on the cargo) and washes out. The margin was locked in the differentials on day one. What could still erode it: the costs and differentials themselves moving before being locked — freight before the vessel is fixed, the sale premium before the sale is done. (That residual risk is Episode 2's subject.)
-
-**S3.** The two un-rentable advantages: **optionality** — owned capacity serves your own cargo at cost exactly when everyone needs it and rented capacity is scarce and expensive; the rented slot exists at boom prices precisely because someone else owns it — and **information** — elevators see farmer selling, terminals see lineups and congestion, vessels see delays, all before any screen or data vendor publishes it; a data subscription is by definition what everyone else can also see. The asset-light firm's genuine advantage: a tiny fixed-cost base — in bust years it simply walks away from rented capacity, while the asset owner still pays for staff, maintenance and capital on quiet terminals. Owning assets is buying a permanent option plus a private data feed, and paying for it in bad-year fixed costs.
-
----
-
-## The episode, in writing
-
-### Not paid to predict
-
-Picture a commodity trader, and you probably imagine someone glued to screens, betting that wheat goes up. That picture is wrong — not slightly wrong, structurally wrong — and understanding why is the foundation for everything else in this series.
-
-A merchant does not get paid for predicting prices. A merchant gets paid for **transforming commodities** — in space, in time, and in form.
-
-**Space** is the oldest transformation, and one of the industry's founding stories illustrates it perfectly: in 1851, a seventeen-year-old named Léopold Louis-Dreyfus began buying wheat from farmers in Alsace and carting it across the border to Basel, where it was worth more. Buy where it's cheap, move it to where it's dear, capture the difference. A hundred and seventy-five years later the cart is a 66,000-tonne vessel and the Alsace–Basel road is Santos–Qingdao. Same trade.
-
-**Time** is the second. At harvest, grain floods the market and prices sag; by spring the flood is over, but the world still eats every day. A merchant buys at harvest, stores, and sells forward — not as a bet that prices will rise, but because the forward market usually pays a known spread for storage. That spread is called carry, and it gets its own episode this week.
-
-**Form** is the third. Nobody eats a raw soybean: crush it and you get meal for animal feed plus oil for cooking — products with actual customers. Blend cheap low-protein wheat with expensive high-protein wheat and you hit exactly the specification a miller in Algeria will pay for. Same atoms, new form, new value.
-
-Who pays for all this? Think of a farmer in Mato Grosso: he grows soybeans brilliantly but has no vessel, no buyer in China, and no desire to carry price risk for six months. Think of a crusher in Shandong: she needs 66,000 tonnes, on spec, arriving the second week of October — not "whenever". The merchant sits between them and absorbs everything they don't want: the logistics, the timing, the quality risk, the price risk. That service is what the margin pays for. A merchant is, at bottom, a risk absorber with a balance sheet.
-
-### The players
-
-The historic big four of grain go by four letters — **ABCD**: ADM, Bunge, Cargill, and Dreyfus, the house that grew out of that Alsace wheat cart. Add the newer giants: COFCO, China's state trading house, and Viterra, which merged with Bunge in 2025. Between them, these firms handle most of the grain that crosses an ocean.
-
-The shape of the business is worth internalizing early: massive volumes, razor-thin margins. A net margin of 1–2% of revenue is a good year. The game is won on repetition and reliability, not home runs.
-
-### Physical vs paper
-
-Physical is the real thing: actual soybeans in an actual silo, with quality certificates and a vessel waiting at berth. Paper is futures and options — standardized contracts on exchanges, what desks simply call **the screen**.
-
-What surprises most newcomers is that merchants trade enormous volumes of paper, yet almost none of it is speculation. Paper exists to *cancel* the price risk of physical positions — hedging. Buy a real cargo of beans and immediately sell futures against it: if the whole market drops a dollar, the cargo loses and the futures win, netting out to roughly flat.
-
-The outright price level — the number on the screen — is called the **flat price**, and the hedge kills it. What's left is the local part of the price: the premium for real beans, in a real port, on a real date. Desks call it the **basis**, and tomorrow's entire episode is built on it.
-
-### The math of one cargo
-
-| Item | ¢/bu |
-|---|---|
-| Buy FOB Santos | futures + 80 |
-| Sell delivered Qingdao | futures + 175 |
-| **Gross margin** | **95** |
-| Ocean freight | −70 |
-| Port & execution | −10 |
-| **Net margin** | **15** |
-
-Fifteen cents a bushel sounds tiny — until you scale it. A tonne of soybeans is about 36.7 bushels, so 15¢/bu ≈ **$5.50/tonne**, and on a 66,000-tonne cargo that is roughly **$360,000** — earned with *no opinion whatsoever* about whether soybeans go up or down. Both legs were quoted as futures-plus-something; the flat price was hedged on day one, and Chicago can rally or crash a dollar during the voyage without touching the result. The money lives entirely in the plus.
-
-### Three words heard daily
-
-**Origination**: buying from the producer end — farmers, cooperatives, country elevators; the desks closest to the crop. **Execution**: everything after the trade is done — vessels, documents, surveyors, discharge — where a good trade can still die of a thousand cuts. **The book**: a desk's full set of positions, physical and paper together; managing it is the actual day job.
-
-### Asset-light vs asset-heavy
-
-Some trading shops own almost nothing — a desk, screens, and credit lines — and rent the rest. Asset-light is nimble but fragile: anyone can copy a trade. The large houses sit on the heavy side: elevators, port terminals, crush plants, chartered fleets.
-
-Why own steel and concrete? Because **assets are options**: when export demand surges, your terminal loads your cargo at cost exactly when capacity is scarcest, while rivals queue and pay up. And because assets are **information machines**: elevators see what farmers are selling, vessels see which ports are jammed. You see the flows before they ever reach a screen — and that information gets paid in a very specific place.
-
-*Tomorrow — Episode 2: Flat price vs basis. The screen says one number; a cargo is worth another. The gap between them is where a physical desk actually lives.*