@sdelsad/commodity-desk-daily 1.0.59 → 1.0.60

package/email.txt

@@ -0,0 +1,556 @@
+SOFT COMMODITY TRADING
+Episode 18 · Thursday 10 September 2026 · 13 min 36
+
+EFP, Delivery and the Squeeze
+How a futures position and a cargo change hands together off-exchange, and
+what Chicago actually delivers when a contract expires.
+
+Listen:      https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/ep18.mp3
+Read online: https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/ep18.html
+
+MARKET PULSE
+============
+
+The record long started coming out. Corn, wheat and soybeans all fell on
+Wednesday, wheat hardest, and the liquidation began one session before the
+report the whole complex has been waiting for.
+
+Contract                 Last                Change
+---------------------------------------------------
+Dec corn (CBOT)          527.75 c/bu            −5¾
+Nov soybeans (CBOT)      1,309.50 c/bu          −6¾
+Dec Chicago SRW (CBOT)   728.75 c/bu           −18¼
+Oct soybean meal (CBOT)  $345.10/short ton    +1.80
+Oct soybean oil (CBOT)   70.08 c/lb         −14 pts
+
+Wheat did the damage, giving back Tuesday's thirteen-cent bounce and more on
+fresh reports of Russian-Ukrainian negotiations. Corn fell as speculators
+trimmed what had been their largest recorded net long, around 431,000
+contracts a few days ago. Soybeans slipped despite flash sales of 340,000 t
+to China and 100,000 t to an unknown buyer, with heavy Midwest rain slowing
+maturity.
+
+Friday's USDA supply and demand report is now the only thing on the
+calendar, and the private estimates have stopped agreeing with each other.
+USDA still carries a corn yield of 180.7 bushels an acre. Pro Farmer's tour
+came back at 173.2. A Reuters poll of the trade sits at 178.2. StoneX went
+the other way entirely and raised production to 16.207 billion bushels, 194
+million above the government. Corn was rated 56 percent good to excellent,
+down a point on the week and against 69 percent a year ago, with harvest 5
+percent done.
+
+That spread of estimates is the actual news. When the private numbers
+straddle the government's in both directions, the report is not a data
+release — it is a resolution of a disagreement, and a record long is
+standing on one side of it. Historically the September report moves corn
+about 8 cents on the day and soybeans about 17.
+
+[chart] Russia's northern detour — The Baltic outlets Russian exporters are
+        turning to can move roughly seven million tonnes in a year. The
+        southern ports they are replacing moved forty-six last season. At
+        this ratio rerouting is not a substitute, and the gap shows up as
+        freight cost long before it shows up as a wheat price. — Reported
+        rail booking requests and Baltic terminal capacity, early September
+        2026, against Azov-Black Sea throughput in the 2025/26 season —
+        https://storage.googleapis.com/podcast-audio-2647223968/commodity-
+        desk-daily/ep18_chart1.png
+
+Two mechanisms ran at once on Wednesday, and they pointed opposite ways.
+
+The first is the familiar one. Negotiation headlines deflated the war
+premium again, on the same day Ukrainian drones struck Novorossiysk. No
+loading capacity was repaired and none was destroyed at a scale that changes
+the season. What repriced was the probability the market assigns to capacity
+returning.
+
+The second is new, and it is policy rather than expectation. Latvia has
+proposed a 300 percent tariff on Russian grain. Aim that at a map and it
+lands precisely on the detour Russian exporters are currently building: shut
+out of Azov and most of Novorossiysk, they have been booking rail north,
+with requests to Baltic outlets running around 5 million tonnes against
+roughly 7 million tonnes a year of terminal capacity — replacing southern
+ports that moved 46.3 million tonnes last season. A tariff on that corridor
+destroys no grain at all. It removes the exit, and grain without an exit is
+priced at the farmgate rather than on a screen.
+
+Ukraine shows the same shape from the other side. Rail arrivals into Greater
+Odesa fell 94.9 percent in August to 68,500 t, while Danube arrivals nearly
+tripled to 248,800 t, and Danube freight to Italy and Spain rose $20–25/t in
+a single week. Demand is rerouting too: Pakistan tendered for 750,000 t
+after Saudi Arabia cancelled an order of 535,000 t.
+
+That is the transmission worth holding on to. The wheat exists. It is
+harvested, it is in store, and it cannot reach the place a contract requires
+it to be. Every dollar of that gap lands on freight first, on the arb
+second, and on the flat price last — which is why a week of severe
+disruption can end with Chicago wheat eighteen cents lower.
+
+  * An EFP is not a clever structure. It is the ordinary way a physical
+    trade gets into futures and out of them, and its real value is that the
+    paper leg and the cargo leg move in the same instant.
+
+  * The futures price inside an EFP is negotiated, not traded. It cannot
+    change what the wheat costs, but it moves where the profit sits — which
+    is why exchanges police it.
+
+  * In Chicago grain, delivery hands the long a shipping certificate, not
+    grain. It is an obligation to be loaded out, with a storage meter
+    running against it.
+
+  * Convergence is not a courtesy the market extends. It is the cost of
+    holding paper that yields nothing, and it works only while that cost
+    exceeds what the carry pays.
+
+  * A squeeze is never about world supply. It is about the tonnes that can
+    physically reach a delivery point before first notice day, and its
+    ceiling is the cost of getting them there.
+
+  * Full carry has a term set by an exchange rule rather than by the market.
+    Percent of full carry is therefore a feedback loop, not a thermometer.
+
+Term                         What it means
+----------------------------------------------------------------------------
+EFP (exchange for physical)  A privately negotiated trade in which a futures
+                             position and a physical position change hands
+                             together, reported to the exchange but not
+                             executed on it
+AA (against actuals)         The softs market's name for the same
+                             transaction
+EFS (exchange for swap)      The same mechanism where the paper leg is an
+                             OTC swap rather than a future
+Legging risk                 The exposure created when the two halves of a
+                             trade are executed minutes apart instead of
+                             simultaneously
+Shipping certificate         What Chicago wheat and corn actually deliver:
+                             paper obliging a regular warehouse to load the
+                             holder out on demand, against a daily storage
+                             charge
+Regular warehouse            A facility approved by the exchange to issue
+                             deliverable certificates at a named delivery
+                             point
+Registered stocks            The quantity currently certificated and
+                             therefore available to settle a delivery
+Load-out rate                The minimum tonnage per day the issuer of a
+                             certificate is contractually obliged to ship
+Variable storage rate (VSR)  The CBOT rule that resets the daily storage
+                             charge on a wheat certificate according to
+                             where the nearby calendar spread sits as a
+                             percentage of full carry
+Squeeze                      A front month bid far above the cash value of
+                             its deliverable because open interest exceeds
+                             what can be delivered in the time available
+Corner                       Control of enough of the deliverable supply
+                             that the shorts have no economic alternative to
+                             paying the holder's price
+Non-convergence              The opposite failure: a future that will not
+                             fall to cash at expiry, because holding the
+                             certificate is cheaper than the carry the
+                             market is paying
+
+
+CONVERSION DRILL 6 OF 12 — CELSIUS ↔ FAHRENHEIT
+===============================================
+
+Rule: °F = °C × 1.8 + 32
+
+Fast method: C → F: double it and add 30 (rough, and about 2° high in the
+mid range). F → C: subtract 30 and halve.
+
+  * 30°C → 60 + 30 = 90°F (exact 86)
+
+  * 35°C → 70 + 30 = 100°F (exact 95)
+
+  * 95°F → 65 ÷ 2 ≈ 32°C (exact 35)
+
+Why it matters: US crop weather is reported in Fahrenheit; the rest of the
+world thinks in Celsius. Pollination stress in corn starts around 32-35°C —
+know what that is on a US forecast.
+
+
+QUIZ
+====
+
+Q1. A merchant is long 40,000 t of soft red winter wheat in a Toledo house.
+He bought it at December minus 12 and hedged it short December Chicago at
+738.00. A miller agrees to take the whole parcel at December plus 30, and
+has been carrying 300 December longs at an average of 744.00 while the
+negotiation ran. They cross it as an EFP with the futures leg struck at
+747.00.
+
+Work out the size in bushels and in lots, the invoice, and each side's
+profit and loss on the futures that changed hands. Then reconcile the
+merchant's total on the trade against his basis margin to the dollar, and
+say what the miller is left holding the moment the EFP is booked.
+
+Q2. A long stands for delivery in Chicago wheat and receives shipping
+certificates. The nearby calendar spread is paying him less carry than the
+daily storage charge those certificates cost him to hold. What does that gap
+push him to do?
+
+Q3. In grain options, out-of-the-money calls usually carry higher implied
+volatility than equidistant puts. Who is paying for that skew?
+
+Q4. A December rough rice lot and a December corn lot carry roughly the same
+notional value. Why does a risk limit written in dollars of notional fail to
+control the rice position?
+
+Q5 — conversion drill. A US forecast puts next week's high across western
+Kansas at 101°F. What is that in Celsius?
+
+----------------------------------------------------------------------------
+
+----------------------------------------------------------------------------
+
+----------------------------------------------------------------------------
+
+
+============================================================================
+                       SOLUTIONS BELOW — ANSWER FIRST                       
+============================================================================
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+SOLUTIONS
+=========
+
+A1. Take it in five steps, and keep the differential separate from the board
+throughout.
+
+Size. 40,000 t × 36.744 = 1,469,760 bushels. At 5,000 bushels a lot that is
+293.95, so the hedge is 294 lots — 1,470,000 bushels, a shade more wheat
+than he owns. One cent on 294 lots is $14,700.
+
+The invoice. The differential was agreed and never moves. Only the board
+does.
+
+                               ¢/bu
+-----------------------------------
+December futures at the EFP  747.00
+Differential                 +30.00
+Delivered price              777.00
+
+1,469,760 bu × $7.77 = $11,420,035.20.
+
+The futures that changed hands. The miller's 294 longs go to the merchant,
+who uses them to extinguish his short.
+
+                      Entry    Exit     Move        P&L
+-------------------------------------------------------
+Merchant, short 294  738.00  747.00  −9 c/bu  −$132,300
+Miller, long 294     744.00  747.00  +3 c/bu   +$44,100
+
+The merchant's total. He bought the physical at 738.00 − 12 = 726.00 and
+sold it at 777.00, a gain of 51 cents on 1,469,760 bushels, or $749,577.60.
+Against the futures loss of $132,300 he nets $617,277.60.
+
+The reconciliation. His basis margin was 42 cents — bought 12 under, sold 30
+over — which on 1,469,760 bushels is $617,299.20. The two differ by exactly
+$21.60, and that is not a rounding artefact. He hedged 1,470,000 bushels
+against 1,469,760 of wheat, so he was short an extra 240 bushels into a
+9-cent rally: 240 × $0.09 = $21.60. The basis margin is the trade.
+Everything else is the rounding on a lot size.
+
+The miller. The EFP can only transfer the futures that correspond to the
+physical, which is 294 lots. He bought 300. He is left long 6 lots — 30,000
+bushels — outright at 747.00, with no wheat behind them and $300 a cent
+riding on it. That is the trap: a placeholder position sized by eye rather
+than by the conversion becomes a naked speculative position the moment the
+real trade is booked, and it is exactly the sort of residual that shows up
+on a position sheet a week later with nobody claiming it.
+
+A2. Load out, or sell the certificate to somebody who wants the wheat.
+
+A certificate pays no yield. It costs storage every day. If the market is
+paying less to carry grain from this month to the next than the issuer is
+charging to hold the paper, then holding it is a slow, certain loss, and the
+holder's cheapest response is to take the wheat and stop the meter — or to
+sell the paper to a miller who has a use for it.
+
+That flow is the delivery mechanism working exactly as designed. It converts
+an unwanted certificate into physical demand at the delivery point, and it
+pulls the expiring future down toward the cash value of the wheat.
+Convergence is not politeness on the market's part. It is the arithmetic of
+a storage charge against a spread, and when the storage charge is set too
+low — as Chicago found after 2008 — convergence simply stops happening.
+
+A3. The consumers and the shorts. Not the farmer.
+
+A crop can fail; it cannot over-succeed by the same magnitude. Supply shocks
+in agriculture push the price up, so the fat tail is a rally and the
+expensive wing is the call. The people bidding for it are the ones a rally
+hurts: the feeder who still has to buy his ration, the exporter who has sold
+cargo he has not yet bought, the fund short into a weather market, the
+importer with a tender to cover.
+
+The farmer's fear runs the other way — he is long the crop and afraid of a
+lower price — so he is a natural seller of calls and a buyer of puts, which
+is precisely why the put side stays comparatively cheap. Read that way, skew
+is not a pricing curiosity. It tells you the direction of the fear in the
+market, not merely its size.
+
+A4. Because notional measures how much money one lot represents, and it says
+nothing about whether anyone will trade it with you.
+
+The two lots are similar sizes of money and completely different positions.
+What separates them is depth: the quantity resting near the touch, and the
+honest daily volume behind it. In corn, a few hundred lots move without
+anyone noticing. In rough rice, the price you get depends on the size you
+want, because roughly a tenth of world production is traded at all and
+policy — not the market — is the supply curve.
+
+The measure that works in a thin market is days to liquidate: the position
+divided by realistic daily volume. A limit expressed that way tells a risk
+manager what he actually needs to know, which is how long it would take to
+be flat if he decided this morning that he wanted to be. A dollar limit
+tells him what the position is worth on a screen that would not honour it.
+
+A5 — conversion drill. Subtract 30 and halve: 101 − 30 = 71, then 71 ÷ 2 ≈
+35.5°C. Exact: (101 − 32) ÷ 1.8 = 38.3°C.
+
+Note where the shortcut breaks. It runs about three degrees cold at the top
+of the range, and three degrees is the whole question here: 35.5 sits at the
+edge of the 32–35°C band where corn pollination stress begins, while 38.3 is
+comfortably past it. In the mid range the quick method is fine. On a heat-
+dome forecast, do the real arithmetic.
+
+
+THE EPISODE, IN WRITING
+=======================
+
+
+
+A merchant is long 25,000 t of soft red winter wheat in a Toledo house. That
+is 918,600 bushels, or 184 Chicago lots, and he is short 184 December
+futures against it. The flat price is dead. What he owns is the basis:
+bought at December minus 10, and he intends to sell at December plus 25.
+
+A miller wants the wheat. While the two were negotiating, the miller did
+what buyers do — he bought 184 December futures as a placeholder, so that a
+rally during the conversation would not cost him the trade.
+
+Look at what those two positions are. The merchant is short 184 December.
+The miller is long 184 December. The same contract, the same month, facing
+each other.
+
+So they do not go anywhere near the exchange. The miller hands his longs to
+the merchant, the merchant hands over the wheat, and the two futures
+positions extinguish each other. The trade is reported to the exchange for
+clearing, but it was never executed on it.
+
+That is an exchange for physical — an EFP. In softs the same thing is called
+against actuals, or AA. When the paper leg is a swap rather than a future it
+is an EFS. The names differ; the mechanism does not.
+
+Three things about it are worth knowing, and the third is the one people
+miss.
+
+It moves no price. One hundred and eighty-four lots pushed into a screen
+would. An EFP prints as a transfer, not as a trade, and the market learns
+about it afterwards if at all.
+
+Both legs happen in the same instant. This is the reason the mechanism
+exists. Do it the other way — lift the hedge on the screen at ten o'clock,
+price the physical at twenty past — and for those twenty minutes the
+merchant is naked on 918,600 bushels of wheat.
+
+
+-----------------------------------------------
+One cent on 184 lots                     $9,200
+A four-cent drift while legging         $36,800
+Basis margin on the trade (35 c/bu)    $321,510
+Legging slip as a share of the margin     11.4%
+
+A four-cent wobble is nothing. It is an ordinary twenty minutes in Chicago.
+And it has just taken an eighth of the margin on a trade the merchant spent
+three weeks originating.
+
+The price of the futures leg is negotiated. This is the part that surprises
+people, and it is not a loophole — it is intrinsic. The EFP has two prices,
+the futures level and the differential, and only their sum is fixed by what
+the wheat is worth.
+
+Where do you want the futures?
+------------------------------
+
+  | MILLER: Twenty-five thousand, December plus twenty-five.
+
+  | MERCHANT: Fine. Where do you want the futures?
+
+  | MILLER: Seven forty-seven. The settle.
+
+  | MERCHANT: Seven forty-seven, and I take your one eighty-four longs
+  | against it.
+
+  | MILLER: Done. Clear it as an EFP.
+
+Where do you want the futures is a real question. Strike the futures leg
+seven cents lower and the differential has to be seven cents higher for the
+wheat to cost the same. The miller pays an identical price either way — but
+each side's futures P&L moves by 7 × $9,200 = $64,400, and it moves in
+opposite directions.
+
+Same wheat, same money, different books. Which is why every exchange
+requires both parties to an EFP to hold a genuine, related physical
+position, and why the futures level has to be commercially defensible rather
+than merely agreed.
+
+Now the other end of the contract.
+
+Most people assume nobody ever delivers, and mostly nobody does. But
+somebody always can, and that possibility is the only reason a future is
+worth the cash grain it is supposed to track.
+
+Here is what almost nobody outside the business knows. In Chicago wheat and
+corn, what gets delivered is not grain.
+
+It is a shipping certificate: paper issued by a regular warehouse at a named
+delivery point, obliging that warehouse to load the holder out, on demand,
+at a contractual daily rate. And every day the holder keeps it, he pays the
+issuer storage.
+
+So the long who stands for delivery does not receive wheat on a quay. He
+receives a queue ticket with a meter running.
+
+That single fact governs how the whole front of the curve behaves. If the
+nearby calendar spread pays less carry than the certificate costs to hold,
+the holder is bleeding, and he stops the bleeding by loading out or by
+selling the paper to someone who wants wheat. That flow is what drags the
+expiring contract down onto the cash market.
+
+Convergence, in other words, is not a courtesy. It is a cost — and it works
+only for as long as the cost is real.
+
+Run it the other way. Suppose the pile is too small.
+
+It is first notice day in an expiring wheat contract. There are still 1,200
+lots of open interest. Registered certificates at the delivery points come
+to 3,100,000 bushels, which is 620 lots.
+
+[chart] The shorts who cannot deliver — Twelve hundred lots are open at
+        first notice day and only six hundred and twenty lots of
+        certificates exist. The other five hundred and eighty have to buy
+        their way out, at a level the longs get to name. — Worked example,
+        episode 18 — https://storage.googleapis.com/podcast-
+        audio-2647223968/commodity-desk-daily/ep18_chart2.png
+
+Five hundred and eighty lots of shorts have nothing to deliver. They have
+three ways out, and a desk prices all three before breakfast.
+
+Exit      What it costs                                              Cost
+----------------------------------------------------------------------------
+Deliver   Freight and handling to a regular point, 18c, plus         22 c/bu
+          registration and grade risk, 4c
+Roll      Buy back the front, sell the next month, pay the inverse   34 c/bu
+Buy back  Pay the squeezed front, above the cash value of the wheat  41 c/bu
+
+[chart] Three ways out of a short — Delivering is the cheapest exit at
+        twenty-two cents, which is why an inverse cannot stay far above that
+        level — provided there are enough days left to move the wheat. A
+        squeeze lives entirely inside the days there are not. — Worked
+        example, episode 18 — https://storage.googleapis.com/podcast-
+        audio-2647223968/commodity-desk-daily/ep18_chart3.png
+
+Twenty-two cents is the cheapest, so the inverse should not be able to stay
+above twenty-two. At twenty-three, every short with wheat in a country
+elevator starts loading trucks, the new certificates appear, and the squeeze
+dies of its own success.
+
+Should not. Unless there are fewer days left than a truck needs.
+
+That is the whole anatomy. A squeeze is a race between a price and a
+logistics chain, and the price is faster. The ceiling on it is the cost of
+making more grain deliverable, less however many days you do not have — and
+in the last week of a delivery period, that subtraction is the entire trade.
+
+On 580 lots, 41 cents is 2,900,000 bushels × $0.41 = $1,189,000, or $2,050 a
+lot. Nobody on that side was wrong about wheat. They were wrong about the
+calendar.
+
+In July 2010 a fund took delivery of 240,100 tonnes of cocoa in London —
+about seven percent of a year's world crop, valued around £658 million.
+Cocoa reached its highest price since 1977.
+
+Then came the part that never makes the story. He had to sell it.
+
+A corner is only half a trade. Acquiring the deliverable is the easy leg;
+the hard leg is exiting into a market that now knows precisely who owns
+everything, and that will not bid for it. The cocoa went back out later the
+same year. This is the structural reason corners are rarer than folklore
+suggests: the conditions that make one possible — a narrow deliverable
+grade, few delivery points, slow load-out, concentrated stocks — are also
+the conditions that make the exit brutal.
+
+But delivery fails in the other direction too, and that failure is the
+stranger and more instructive one.
+
+Chicago wheat, from 2008 onward, would not converge. Certificates traded
+persistently above the cash value of the wheat behind them, and the expiring
+future stayed stubbornly above the market it was supposed to track. The
+cause was not manipulation. It was that storage on a certificate had been
+set too cheap. If holding paper costs less than the market pays you to carry
+grain, nobody loads out, nothing turns into physical demand, and the
+mechanism that pulls the future onto cash simply never fires.
+
+The exchange's answer, in 2010, was the variable storage rate. The daily
+storage charge a certificate holder pays the issuer stopped being a
+constant. It is now reset against a nearby calendar spread, measured as a
+percentage of full carry:
+
+Spread as % of full carry  What happens to the storage rate
+-------------------------------------------------------------------------
+80% or more                Steps up by 0.10¢/bu per day, about 3¢ a month
+Between 50% and 80%        Unchanged
+50% or less                Steps down by the same amount
+
+There is a floor — roughly 0.165¢/bu a day, about 5 cents a bushel a month,
+for Chicago and KC wheat — and no technical ceiling. It has run near 20
+cents a month in heavy-supply years.
+
+Now sit with what that implies, because it quietly rewrites something this
+show has used twice already.
+
+Full carry has three terms: interest, which the market sets; storage, which
+an exchange rule sets; and the spread you are measuring against them.
+Episode 16 read a Chicago December–March spread at 46 percent of full carry.
+Forty-six is a number below fifty — which is a number that, at the next
+reset, changes one of the inputs that produced it.
+
+Percent of full carry is not a thermometer. It is a feedback loop,
+deliberately built as one, and it is the closest thing in these markets to a
+mechanism that reads its own output and adjusts. Knowing where the
+thresholds sit is worth more than knowing where the spread is, because the
+thresholds tell you which way the ground under the spread is about to move.
+
+
+----------------------------------------------------------------------------
+Soft Commodity Trading — a daily briefing on physical commodity trading.
+
+GLOSSARY
+Every unit and expression the show has introduced lives on the episode page:
+https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/ep18.html#glossary
+
+All episodes: https://storage.googleapis.com/podcast-audio-2647223968/index.html
+RSS: https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/feed.xml

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sdelsad/commodity-desk-daily",
-  "version": "1.0.59",
+  "version": "1.0.60",
   "description": "Soft Commodity Trading - Ep 18: EFP, Delivery and the Squeeze",
   "license": "CC-BY-4.0",
   "keywords": [