@sdelsad/commodity-desk-daily 1.0.48 → 1.0.49
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/covered.md +1 -0
- package/ep16.html +727 -0
- package/ep16.md +256 -0
- package/ep16.mp3 +0 -0
- package/ep16.script.txt +111 -0
- package/ep16_chart1.png +0 -0
- package/ep16_chart2.png +0 -0
- package/ep16_chart3.png +0 -0
- package/feed.xml +12 -0
- package/glossary.md +10 -0
- package/package.json +2 -2
- package/email.html +0 -99
- package/email.txt +0 -479
- package/ep15.html +0 -716
- package/ep15.md +0 -196
- package/ep15.script.txt +0 -78
- package/ep15_chart1.png +0 -0
- package/ep15_chart2.png +0 -0
- package/ep15_chart3.png +0 -0
package/email.txt
DELETED
|
@@ -1,479 +0,0 @@
|
|
|
1
|
-
SOFT COMMODITY TRADING
|
|
2
|
-
Episode 15 · Friday 4 September 2026 · 13 min 35
|
|
3
|
-
|
|
4
|
-
Cotton, Rice and Juice
|
|
5
|
-
A cotton mill that has taken delivery but not set a price is short the board
|
|
6
|
-
without owning a single contract, and the CFTC publishes exactly how much of
|
|
7
|
-
that unfixed buying is still to come.
|
|
8
|
-
|
|
9
|
-
Listen: https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/ep15.mp3
|
|
10
|
-
Read online: https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/ep15.html
|
|
11
|
-
|
|
12
|
-
MARKET PULSE
|
|
13
|
-
============
|
|
14
|
-
|
|
15
|
-
Cotton gave back two and a half cents, and it was the biggest move on the
|
|
16
|
-
board.
|
|
17
|
-
|
|
18
|
-
Market Contract Settle Change
|
|
19
|
-
----------------------------------------------
|
|
20
|
-
Cotton No. 2 Dec 26 86.45 ¢/lb −248 pts
|
|
21
|
-
Chicago wheat Dec 26 754¼ ¢/bu −19¾¢
|
|
22
|
-
Soybean oil Oct 26 69.63 ¢/lb −101 pts
|
|
23
|
-
Corn Dec 26 540¾ ¢/bu −2¾¢
|
|
24
|
-
Rough rice Nov 26 $15.70 /cwt −2½¢
|
|
25
|
-
Soybeans Nov 26 1316¼ ¢/bu +6¢
|
|
26
|
-
Soybean meal Oct 26 $348.60 /st +$5.70
|
|
27
|
-
|
|
28
|
-
Thursday was a soft session with one sharp exception. December cotton
|
|
29
|
-
settled at 86.45, down 248 points — a point being one hundredth of a cent,
|
|
30
|
-
so two and a half cents came off the board in a day. Chicago wheat lost
|
|
31
|
-
nearly twenty cents, or 2.6 percent. Beans and meal were the only things
|
|
32
|
-
higher.
|
|
33
|
-
|
|
34
|
-
Cotton had climbed through August. Late in the month the December contract
|
|
35
|
-
printed a fresh contract high near 89.45, on a US crop rated 38 percent good
|
|
36
|
-
against 55 percent a year earlier, and on world ending stocks forecast at
|
|
37
|
-
the lowest since 2011/12. Thursday's break undid roughly a third of that.
|
|
38
|
-
|
|
39
|
-
The policy read. China's state reserve has been auctioning cotton into its
|
|
40
|
-
own mills, and the auctions have cleared in full for 24 consecutive sessions
|
|
41
|
-
— about 192,497 tonnes placed by 21 August. A reserve auction that sells out
|
|
42
|
-
every day is not a price cap. It is a government discovering that domestic
|
|
43
|
-
supply is short, and a domestic bid that eventually has to be met from
|
|
44
|
-
outside. That is the transmission: reserve draw first, import demand second,
|
|
45
|
-
world price third.
|
|
46
|
-
|
|
47
|
-
[chart] Thursday's moves — Cotton and wheat carried the session. Everything
|
|
48
|
-
else was noise around them, and only the bean complex held. — CBOT
|
|
49
|
-
and ICE settlements, Thursday 3 September 2026 —
|
|
50
|
-
https://storage.googleapis.com/podcast-audio-2647223968/commodity-
|
|
51
|
-
desk-daily/ep15_chart1.png
|
|
52
|
-
|
|
53
|
-
* An unfixed on-call sale is a buy order that has not been placed yet, and
|
|
54
|
-
it carries a deadline. Read the report by futures month, not by total:
|
|
55
|
-
the total gives the mood, the months give the trade.
|
|
56
|
-
|
|
57
|
-
* Mills fix into weakness, growers fix into strength. Whoever is closer to
|
|
58
|
-
first notice day is the one who has to move.
|
|
59
|
-
|
|
60
|
-
* A big net on-call number is a spread signal, not a flat price signal. In
|
|
61
|
-
late August the pressure sat in March and July 2027, while December 2026
|
|
62
|
-
carried net latent selling.
|
|
63
|
-
|
|
64
|
-
* Notional is not size. A rice lot and a corn lot are worth about the same
|
|
65
|
-
money and are not remotely the same position.
|
|
66
|
-
|
|
67
|
-
* In a thin market, size is measured in days to liquidate, not dollars.
|
|
68
|
-
|
|
69
|
-
* Rice is thin because policy is its supply curve. Juice is thin because
|
|
70
|
-
its supply is biology, and the trees are not coming back.
|
|
71
|
-
|
|
72
|
-
Term Meaning
|
|
73
|
-
----------------------------------------------------------------------------
|
|
74
|
-
Cotton No. 2 (CT) The ICE cotton contract: 50,000 lb net weight, quoted
|
|
75
|
-
in US cents per pound. One cent is $500 a lot, one
|
|
76
|
-
point is $5
|
|
77
|
-
bale The unit every cotton statistic uses: 480 lb net in
|
|
78
|
-
the United States, so one CT lot is about 104 bales
|
|
79
|
-
on-call sale Cotton sold by a merchant to a mill at an agreed
|
|
80
|
-
differential, with the futures leg left for the buyer
|
|
81
|
-
to fix later
|
|
82
|
-
on-call purchase Cotton bought by a merchant from a grower, with the
|
|
83
|
-
futures leg left for the seller to fix later
|
|
84
|
-
unfixed A price-to-be-fixed contract whose futures leg has
|
|
85
|
-
not been set, so the exposure is still outright
|
|
86
|
-
cotton on-call report The weekly CFTC table of unfixed sales and purchases
|
|
87
|
-
by futures month — a map of forced order flow
|
|
88
|
-
first notice day The first date a short may tender delivery, and the
|
|
89
|
-
practical deadline for fixing or rolling an unfixed
|
|
90
|
-
position
|
|
91
|
-
rough rice contract The CBOT rice future: 2,000 cwt of long grain rough
|
|
92
|
-
rice, quoted in dollars per hundredweight
|
|
93
|
-
FCOJ-A The ICE orange juice contract: 15,000 lb of orange
|
|
94
|
-
solids, quoted in US cents per pound
|
|
95
|
-
thin market A market where the price you get depends materially
|
|
96
|
-
on the size you want, whatever a single lot is worth
|
|
97
|
-
market depth The quantity resting near the touch — what actually
|
|
98
|
-
sets execution cost, as opposed to headline volume
|
|
99
|
-
days to liquidate Position divided by honest daily volume: the sizing
|
|
100
|
-
measure that replaces a notional limit in a thin
|
|
101
|
-
market
|
|
102
|
-
state reserve auction A government selling from its own stockpile into its
|
|
103
|
-
domestic market, whose clearing rate reads as a
|
|
104
|
-
tightness signal
|
|
105
|
-
citrus greening Huanglongbing, the bacterial disease that permanently
|
|
106
|
-
cuts an infected orange tree's yield and cannot be
|
|
107
|
-
cured
|
|
108
|
-
|
|
109
|
-
|
|
110
|
-
CONVERSION DRILL 3 OF 12 — BUSHELS ↔ TONNES, WHEAT AND SOYBEANS
|
|
111
|
-
===============================================================
|
|
112
|
-
|
|
113
|
-
Rule: 1 tonne ≈ 36.7 bushels (a bushel of wheat or soybeans is 60 lb)
|
|
114
|
-
|
|
115
|
-
Fast method: tonnes → bushels: ×37 is close enough, and ×37 = ×40 minus ×3.
|
|
116
|
-
Bushels → tonnes: ÷37, i.e. ÷40 then add 8%.
|
|
117
|
-
|
|
118
|
-
* 60,000 t → 2,400,000 − 180,000 = 2.22 million bu
|
|
119
|
-
|
|
120
|
-
* 25,000 t → 1,000,000 − 75,000 = 925,000 bu (exact 918,000)
|
|
121
|
-
|
|
122
|
-
* 500,000 bu → 12,500 + 1,000 ≈ 13,600 t (exact 13,608)
|
|
123
|
-
|
|
124
|
-
Why it matters: the corn factor and the wheat/soybean factor differ by 7%.
|
|
125
|
-
Using the wrong one on a cargo is a six-figure error.
|
|
126
|
-
|
|
127
|
-
|
|
128
|
-
QUIZ
|
|
129
|
-
====
|
|
130
|
-
|
|
131
|
-
Q1. A spinning mill has bought 620 lots of cotton on call against March
|
|
132
|
-
2027, at March plus 780 points. March futures were at 89.93 the day the
|
|
133
|
-
contract was agreed. The mill takes the view that the market has run too
|
|
134
|
-
far, waits, and ends up fixing the whole position at 93.40 shortly before
|
|
135
|
-
first notice day.
|
|
136
|
-
|
|
137
|
-
* What is the mill's final delivered cost in cents per pound, and what is
|
|
138
|
-
the total invoice in dollars?
|
|
139
|
-
|
|
140
|
-
* What did the decision to wait cost, against fixing on the day the
|
|
141
|
-
contract was agreed?
|
|
142
|
-
|
|
143
|
-
* The treasurer says: "we should simply have bought 620 March futures on
|
|
144
|
-
day one." Would that have removed the exposure — and what would the mill
|
|
145
|
-
have been left holding at fixation?
|
|
146
|
-
|
|
147
|
-
Q2. The CFTC on-call report of 21 August 2026 showed December 2026 carrying
|
|
148
|
-
30,866 lots of unfixed sales against 32,711 lots of unfixed purchases. What
|
|
149
|
-
does that month's balance imply about order flow still to come in December
|
|
150
|
-
futures?
|
|
151
|
-
|
|
152
|
-
Q3. Brazilian hydrous ethanol parity sat at 12.60 c/lb while the No. 11
|
|
153
|
-
screen was at 17.56. A fund manager argues that a further rally in raw sugar
|
|
154
|
-
will pull more Brazilian sugar out of the mills. Why is that wrong?
|
|
155
|
-
|
|
156
|
-
Q4. September arabica settled 36.10 cents over December on the first day of
|
|
157
|
-
the spot contract's delivery notice period. What does an inverse that steep
|
|
158
|
-
say about the position of the front-month shorts?
|
|
159
|
-
|
|
160
|
-
Q5. Conversion drill. A Panamax of 44,000 t of soybeans is sold. How many
|
|
161
|
-
bushels is that, and how many Chicago lots would hedge it?
|
|
162
|
-
|
|
163
|
-
|
|
164
|
-
============================================================================
|
|
165
|
-
SOLUTIONS BELOW — ANSWER FIRST
|
|
166
|
-
============================================================================
|
|
167
|
-
|
|
168
|
-
|
|
169
|
-
|
|
170
|
-
|
|
171
|
-
|
|
172
|
-
|
|
173
|
-
|
|
174
|
-
|
|
175
|
-
|
|
176
|
-
|
|
177
|
-
|
|
178
|
-
|
|
179
|
-
|
|
180
|
-
|
|
181
|
-
|
|
182
|
-
|
|
183
|
-
|
|
184
|
-
|
|
185
|
-
|
|
186
|
-
|
|
187
|
-
|
|
188
|
-
|
|
189
|
-
|
|
190
|
-
|
|
191
|
-
|
|
192
|
-
|
|
193
|
-
SOLUTIONS
|
|
194
|
-
=========
|
|
195
|
-
|
|
196
|
-
A1. Work it in three steps, and keep the differential separate from the
|
|
197
|
-
board throughout — that separation is the whole point of an on-call
|
|
198
|
-
contract.
|
|
199
|
-
|
|
200
|
-
Size. 620 lots × 50,000 lb = 31,000,000 lb, which at 480 lb to the bale is
|
|
201
|
-
about 64,583 bales. Note the tick value that follows from it: one cent on
|
|
202
|
-
620 lots is 620 × $500 = $310,000.
|
|
203
|
-
|
|
204
|
-
The delivered cost. The differential was agreed and never changes. Only the
|
|
205
|
-
board moves.
|
|
206
|
-
|
|
207
|
-
¢/lb
|
|
208
|
-
---------------------------------
|
|
209
|
-
March futures at fixation 93.40
|
|
210
|
-
Differential +7.80
|
|
211
|
-
Delivered cost 101.20
|
|
212
|
-
|
|
213
|
-
Invoice: 31,000,000 lb × $1.0120 = $31,372,000.
|
|
214
|
-
|
|
215
|
-
The cost of waiting. Fixing on day one at 89.93 would have given 89.93 +
|
|
216
|
-
7.80 = 97.73 ¢/lb, and an invoice of 31,000,000 × $0.9773 = $30,296,300. The
|
|
217
|
-
wait cost $1,075,700 — which is simply the 3.47¢ rally times the
|
|
218
|
-
$310,000-a-cent tick value. Per lot, $1,735.
|
|
219
|
-
|
|
220
|
-
The treasurer. Yes, buying 620 March futures at 89.93 on day one would have
|
|
221
|
-
removed the exposure exactly. That long would have gained 3.47¢ × 620 × $500
|
|
222
|
-
= $1,075,700, precisely offsetting the higher fixation. At fixation the mill
|
|
223
|
-
sells the futures out and is left holding cotton at an effective delivered
|
|
224
|
-
cost of 97.73 ¢/lb.
|
|
225
|
-
|
|
226
|
-
The trap the question is testing: the mill was never "unhedged and waiting".
|
|
227
|
-
An unfixed on-call purchase is a position — the mill is short the board, one
|
|
228
|
-
for one, from the moment it takes the cotton. Doing nothing was not
|
|
229
|
-
neutrality, it was a 620-lot short held by a company that does not consider
|
|
230
|
-
itself a speculator. And the mill was 620 of the 19,481 lots of unfixed
|
|
231
|
-
March sales in the market, about 3 percent of a queue of buyers all waiting
|
|
232
|
-
for the same dip.
|
|
233
|
-
|
|
234
|
-
A2. December carried net latent selling, not buying — 32,711 − 30,866 =
|
|
235
|
-
1,845 lots more unfixed purchases than unfixed sales.
|
|
236
|
-
|
|
237
|
-
The direction is the part people get backwards. An unfixed sale is a mill
|
|
238
|
-
that has taken cotton and not priced it: its cost rises with the board, so
|
|
239
|
-
it must eventually buy futures. That is latent buying. An unfixed purchase
|
|
240
|
-
is a merchant who has bought from a grower with the grower holding the right
|
|
241
|
-
to fix: that fixation is a sale. That is latent selling.
|
|
242
|
-
|
|
243
|
-
So the December column says that the flow still to come in that contract
|
|
244
|
-
leans, mildly, to the sell side — even though the market-wide total of
|
|
245
|
-
79,167 sales against 67,696 purchases leans to the buy side by 11,471 lots.
|
|
246
|
-
The overhang was real, but it was not in December. It was in March (+12,519)
|
|
247
|
-
and July 2027 (+12,651), with December 2027 running hard the other way
|
|
248
|
-
(−18,583) because that is where growers have sold forward and not yet fixed.
|
|
249
|
-
Reading the total and buying the front month gets the right thesis in the
|
|
250
|
-
wrong contract.
|
|
251
|
-
|
|
252
|
-
A3. Because the switch is already spent. Ethanol parity is the level at
|
|
253
|
-
which a mill earns the same per unit of ATR from sugar as from ethanol. At
|
|
254
|
-
17.56 against a parity of 12.60, mills are already about five cents above
|
|
255
|
-
the point where the decision flips, so every mill that can make sugar is
|
|
256
|
-
already making the maximum its crystallisation capacity allows. The mix is
|
|
257
|
-
set at the start of the season by hardware, not by the screen.
|
|
258
|
-
|
|
259
|
-
Above parity, a rally therefore buys no extra Brazilian tonnes. It has only
|
|
260
|
-
one remaining job, which is to ration demand. The price sensitivity of
|
|
261
|
-
Brazilian supply lives below parity, not above it: it is a fall towards
|
|
262
|
-
12.60 that changes behaviour, by making ethanol the better home for the
|
|
263
|
-
cane. A trader who models Brazilian supply as a smooth function of the sugar
|
|
264
|
-
price will keep expecting an elasticity that has already been used up.
|
|
265
|
-
|
|
266
|
-
A4. It says the front-month shorts cannot deliver. A 36.10-cent inverse is
|
|
267
|
-
the market charging an enormous price for coffee now versus coffee in three
|
|
268
|
-
months, on the first day of the notice period — which is precisely the
|
|
269
|
-
moment a short has to choose between tendering certified coffee and buying
|
|
270
|
-
its position back. When the certified float is small enough (226,242 bags,
|
|
271
|
-
under half a day of world consumption), most shorts do not have the coffee,
|
|
272
|
-
and everyone knows it. The inverse is the price of that fact.
|
|
273
|
-
|
|
274
|
-
Two things follow. A carry market pays you to store and an inverse punishes
|
|
275
|
-
you for it, so nobody with coffee is holding it back. And an inverse this
|
|
276
|
-
steep is a statement about deliverable supply, not about world supply — the
|
|
277
|
-
crop can be a record and the front month can still be squeezed, because only
|
|
278
|
-
certified stock settles a contract.
|
|
279
|
-
|
|
280
|
-
A5. A tonne of soybeans is 36.744 bushels (60 lb to the bushel).
|
|
281
|
-
|
|
282
|
-
44,000 t × 36.744 = 1,616,736 bu
|
|
283
|
-
|
|
284
|
-
In your head: ×37 is ×40 minus ×3, so 1,760,000 − 132,000 ≈ 1,628,000 bu —
|
|
285
|
-
close enough to quote.
|
|
286
|
-
|
|
287
|
-
At 5,000 bu to a Chicago lot, 1,616,736 bu is 323 lots. The trap is the corn
|
|
288
|
-
factor: using 39.368 would give 1,732,192 bu and 346 lots, and you would be
|
|
289
|
-
over-hedged by 23 lots.
|
|
290
|
-
|
|
291
|
-
|
|
292
|
-
THE EPISODE, IN WRITING
|
|
293
|
-
=======================
|
|
294
|
-
|
|
295
|
-
|
|
296
|
-
|
|
297
|
-
The contract, and the unit under it
|
|
298
|
-
-----------------------------------
|
|
299
|
-
|
|
300
|
-
ICE Cotton No. 2 is 50,000 lb of net weight, quoted in US cents per pound.
|
|
301
|
-
One cent is $500 a lot. One point — a hundredth of a cent — is $5.
|
|
302
|
-
|
|
303
|
-
Underneath it sits the bale, which is what every cotton statistic in the
|
|
304
|
-
world uses: production, stocks, exports, mill use. A US bale is 480 lb net.
|
|
305
|
-
So one futures lot is about 104 bales. The number is not elegant, and there
|
|
306
|
-
is no reason for it beyond history: 480 lb is what a compressed bale weighed
|
|
307
|
-
when the standard was written, and the trade never revisited it. The
|
|
308
|
-
practical consequence is that a cotton desk converts between bales and
|
|
309
|
-
pounds continuously, because the fundamentals arrive in one unit and the
|
|
310
|
-
hedge is denominated in the other.
|
|
311
|
-
|
|
312
|
-
On call: the report nobody else gets
|
|
313
|
-
------------------------------------
|
|
314
|
-
|
|
315
|
-
A mill does not usually buy cotton at a price. It buys cotton on call: the
|
|
316
|
-
differential is agreed today, the futures leg is fixed later, at a moment
|
|
317
|
-
one side gets to choose. Structurally this is coffee's price-to-be-fixed,
|
|
318
|
-
and the mechanics of fixation are the same.
|
|
319
|
-
|
|
320
|
-
What cotton has that coffee does not is a public tally. Every week the CFTC
|
|
321
|
-
publishes how much cotton has been sold on call and not yet fixed, and how
|
|
322
|
-
much has been bought on call and not yet fixed, broken down by futures
|
|
323
|
-
month. No other soft market publishes its unfixed book.
|
|
324
|
-
|
|
325
|
-
The direction is where people go wrong, so it is worth stating slowly.
|
|
326
|
-
|
|
327
|
-
Position Who holds it What fixing requires Latent flow
|
|
328
|
-
----------------------------------------------------------------------------
|
|
329
|
-
Unfixed on-call A merchant has sold The mill's cost Buying
|
|
330
|
-
sale to a mill; the mill rises with the
|
|
331
|
-
fixes board, so it must
|
|
332
|
-
buy futures
|
|
333
|
-
Unfixed on-call A merchant has The grower's revenue Selling
|
|
334
|
-
purchase bought from a rises with the
|
|
335
|
-
grower; the grower board, so fixing is
|
|
336
|
-
fixes a sale
|
|
337
|
-
|
|
338
|
-
A mill sitting on unfixed purchases is short the board without owning a
|
|
339
|
-
single contract. Its cost moves one for one with December or March, and the
|
|
340
|
-
only way out is to buy — either as a hedge now, or as the act of fixation
|
|
341
|
-
later. Fixation is not an administrative step. It is a purchase.
|
|
342
|
-
|
|
343
|
-
Which means every unfixed on-call sale in that table is a buy order that has
|
|
344
|
-
not been placed yet, and unlike a speculator's order it has a deadline
|
|
345
|
-
attached: first notice day.
|
|
346
|
-
|
|
347
|
-
Here is what it sounds like in late January:
|
|
348
|
-
|
|
349
|
-
| MILL: What's March showing?
|
|
350
|
-
|
|
351
|
-
| MERCHANT: Ninety twenty.
|
|
352
|
-
|
|
353
|
-
| MILL: I'll wait.
|
|
354
|
-
|
|
355
|
-
| MERCHANT: You've four hundred lots to fix and eleven sessions to first
|
|
356
|
-
| notice.
|
|
357
|
-
|
|
358
|
-
| MILL: Then I'll wait ten of them.
|
|
359
|
-
|
|
360
|
-
Nobody argued about the differential. It was agreed months ago. The entire
|
|
361
|
-
negotiation was about a calendar.
|
|
362
|
-
|
|
363
|
-
Why it becomes a squeeze
|
|
364
|
-
------------------------
|
|
365
|
-
|
|
366
|
-
Take a mill holding 400 lots on call against March — 20 million pounds,
|
|
367
|
-
roughly 41,700 bales. Every cent the board rises before it fixes costs it
|
|
368
|
-
400 × $500 = $200,000. Three cents is $600,000, on cotton already sitting in
|
|
369
|
-
its own warehouse.
|
|
370
|
-
|
|
371
|
-
That is a cost, not a squeeze. The squeeze is what happens when every mill
|
|
372
|
-
reasons the same way at once.
|
|
373
|
-
|
|
374
|
-
Each of them waits because it believes the market will come back. So the
|
|
375
|
-
buying does not arrive smoothly across the quarter. It arrives compressed
|
|
376
|
-
into the last sessions before the notice period, from buyers who by then
|
|
377
|
-
have no view left — only a deadline. A rally that began with a dry Texas
|
|
378
|
-
summer ends with people buying because the calendar told them to, and the
|
|
379
|
-
last stretch of it has nothing to do with cotton fundamentals at all.
|
|
380
|
-
|
|
381
|
-
The month is the trade
|
|
382
|
-
----------------------
|
|
383
|
-
|
|
384
|
-
The 21 August report showed 79,167 lots of unfixed sales against 67,696 of
|
|
385
|
-
unfixed purchases — net about 11,471 lots of latent buying. The obvious
|
|
386
|
-
conclusion is to be long cotton. The obvious conclusion gets the contract
|
|
387
|
-
wrong.
|
|
388
|
-
|
|
389
|
-
[chart] Net unfixed on-call, by month — The overhang is not in December. It
|
|
390
|
-
is in March and July 2027 — and December 2027 leans the other way,
|
|
391
|
-
because that is where growers have sold forward and not yet priced.
|
|
392
|
-
— CFTC Cotton On-Call report, week ended 21 August 2026 —
|
|
393
|
-
https://storage.googleapis.com/podcast-audio-2647223968/commodity-
|
|
394
|
-
desk-daily/ep15_chart2.png
|
|
395
|
-
|
|
396
|
-
December 2026 is net short of buying: 30,866 sales against 32,711 purchases.
|
|
397
|
-
The pressure sits in March 2027 (+12,519 net to buy) and July 2027 (+12,651,
|
|
398
|
-
against almost nothing on the other side). December 2027 runs hard the other
|
|
399
|
-
way, at −18,583, because that is the month growers use to sell a crop
|
|
400
|
-
forward before they are willing to price it.
|
|
401
|
-
|
|
402
|
-
So the report is not a flat price signal. It is a spread signal. What it
|
|
403
|
-
argues for is long March against December, not long cotton.
|
|
404
|
-
|
|
405
|
-
And there is a layer beneath even that. The purchase side of the table is
|
|
406
|
-
the grower. Growers fix into strength; mills fix into weakness. A rally
|
|
407
|
-
therefore summons mill buying and grower selling simultaneously, and which
|
|
408
|
-
one dominates is a question of who is nearer a deadline — not of who is more
|
|
409
|
-
numerous. That is the read, and it changes week to week as the calendar
|
|
410
|
-
advances.
|
|
411
|
-
|
|
412
|
-
Thin is not small
|
|
413
|
-
-----------------
|
|
414
|
-
|
|
415
|
-
Rice and orange juice teach the second lesson, and it starts by killing an
|
|
416
|
-
intuition.
|
|
417
|
-
|
|
418
|
-
[chart] One lot, four markets — Within twenty thousand dollars of each other
|
|
419
|
-
in notional — and orders of magnitude apart in depth. Contract size
|
|
420
|
-
tells you almost nothing about how hard a position is to leave. —
|
|
421
|
-
Settlements of 3 September 2026 and exchange contract specifications
|
|
422
|
-
— https://storage.googleapis.com/podcast-audio-2647223968/commodity-
|
|
423
|
-
desk-daily/ep15_chart3.png
|
|
424
|
-
|
|
425
|
-
A rough rice lot is 2,000 cwt — 200,000 lb — and at $15.70 that is $31,400.
|
|
426
|
-
A December corn lot at 540¾ is $27,038. Wheat is $37,713, cotton $43,225. By
|
|
427
|
-
notional these are the same instrument.
|
|
428
|
-
|
|
429
|
-
They are nothing like the same instrument, because thin does not mean small.
|
|
430
|
-
Thin means the price you get depends on how much you want. In corn, 370 lots
|
|
431
|
-
is invisible. In rough rice, 319 lots — almost identical money — is
|
|
432
|
-
somebody's week.
|
|
433
|
-
|
|
434
|
-
The practical consequence is that a notional risk limit is the wrong control
|
|
435
|
-
in a thin market. The right measure is days to liquidate: take the position,
|
|
436
|
-
divide by honest daily volume, and ask how many sessions it takes to get out
|
|
437
|
-
without being the market yourself. If the answer is more than two or three,
|
|
438
|
-
the desk does not have a position. It has a commitment.
|
|
439
|
-
|
|
440
|
-
Two different kinds of thin
|
|
441
|
-
---------------------------
|
|
442
|
-
|
|
443
|
-
Rice and juice arrive at thinness from opposite directions, and the
|
|
444
|
-
distinction matters because it tells you what kind of shock to expect.
|
|
445
|
-
|
|
446
|
-
Rice is thin because most of it never moves. It is grown and eaten in the
|
|
447
|
-
same countries; only around a tenth of world production crosses a border at
|
|
448
|
-
all. And that traded tenth is dominated by governments — export duties,
|
|
449
|
-
minimum export prices, licence regimes. One notification out of Delhi resets
|
|
450
|
-
the world price faster than any monsoon.
|
|
451
|
-
|
|
452
|
-
When policy is the supply curve, information does not accumulate. There is
|
|
453
|
-
nothing, and then there is everything. That is why in rice the futures
|
|
454
|
-
matter less than the announcements, and why a position that looked liquid on
|
|
455
|
-
Friday can be untradeable on Monday morning.
|
|
456
|
-
|
|
457
|
-
Juice is thin for the opposite reason. Its supply constraint is not policy
|
|
458
|
-
but biology, and it is one-directional. Citrus greening has been working
|
|
459
|
-
through São Paulo for two decades; around 40 percent of a recent Brazilian
|
|
460
|
-
crop was affected. A tree that has it does not recover. So this is not a
|
|
461
|
-
weather premium that builds ahead of a window and decays on the calendar —
|
|
462
|
-
it is a slow, permanent reduction in the number of producing trees. FCOJ-A
|
|
463
|
-
on ICE is 15,000 lb of orange solids. A market that small does not absorb a
|
|
464
|
-
fund. It gets moved by one.
|
|
465
|
-
|
|
466
|
-
Cotton, then, sits between the two. It is liquid enough to trade properly,
|
|
467
|
-
and it publishes the one thing the thin markets never tell you: exactly how
|
|
468
|
-
much forced buying is still to come, and in which month.
|
|
469
|
-
|
|
470
|
-
|
|
471
|
-
----------------------------------------------------------------------------
|
|
472
|
-
Soft Commodity Trading — a daily briefing on physical commodity trading.
|
|
473
|
-
|
|
474
|
-
GLOSSARY
|
|
475
|
-
Every unit and expression the show has introduced lives on the episode page:
|
|
476
|
-
https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/ep15.html#glossary
|
|
477
|
-
|
|
478
|
-
All episodes: https://storage.googleapis.com/podcast-audio-2647223968/index.html
|
|
479
|
-
RSS: https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/feed.xml
|