😎 The Invisible Monopoly - MAN B&W / Everllence = the heart of global trade 😎

SPOILER: pure exchange-traded exposure to this chokepoint does not exist - the company whose blueprints move ~80% of global trade sat inside the auto giant Volkswagen until 2018 / in the summer of 2026 its control stake (51%) went to the private equity fund Bain Capital for ~€7.4B.

the second player in the duopoly - under the control of the Chinese state.

• MAN B&W (legal entity - Everllence, ex-MAN Energy Solutions) - a "design house," not a factory.

It designs large low-speed two-stroke marine diesel engines and licenses the design to Korean/Chinese/Japanese yards, collecting royalties on every engine built + lifetime service.

MAN B&W holds ~80.9-84.6% of the global low-speed two-stroke market.

MAN B&W BUSINESS MODEL.

they don't build engines - they sell the blueprint and collect the rent.

MAN Energy Solutions (officially Everllence since 4 June 2025) ceased serial production of large two-strokes in Copenhagen back in 1980.

Today the R&D center in Copenhagen designs engines under the MAN B&W brand (now Everllence B&W), while they are physically built by a "family of licensees" predominantly in Asia, PAYING ROYALTIES ON EVERY ENGINE.

MARKET SHARE.

• 80.9-84.6%

by number of installations - MAN 84.6% / WinGD 8.8% / Mitsubishi-UE 3.8%
by market share - MAN 80.9% / WinGD 15.3% / UE 3.8%

• MAN B&W "holds over an 80% worldwide share of marine low-speed engines".

WinGD has been eating into the share in recent years - especially in gas carriers (LNG) and methanol container ships
in the Chinese mid-bore segment WinGD claims growth from ~4% to ~20% in a year

WinGD = China, phased buyout.

the Wärtsilä/CSSC JV began operations on 19 January 2015 (CSSC 70% / Wärtsilä 30%)
in 2016 Wärtsilä exited completely - 100% to CSSC.

MAN B&W OWNERSHIP.

• Until 2018 - part of MAN SE.
• In 2018 VW bought out the business for ~€1.7B.
• 4 June 2025 - rebranding to Everllence.
• 24 June 2026 VW announced the sale of 51% to the Bain Capital fund for ~€7.4B (LBO) / retaining 49%.

THE FUEL TRANSITION = LEVERAGE

essentially, whoever sets the standard for the engine of the future (methanol/ammonia/LNG) locks in the royalty stream from the entire new fleet 20-30 years ahead.

Why the two-stroke = monopoly

The large merchant fleet = container ships / tankers / bulkers / gas carriers - is driven by low-speed two-stroke crosshead diesels.

They are connected directly to the propeller shaft, without a gearbox, turn at 60-230 rpm, burn heavy fuel oil, serve 20-30 years and are the most efficient prime mover for ocean crossings.

Only a handful in the world know how to design such a unit (as tall as a 4-story building, with output up to ~82,440 kW) / building one is already far easier.

Hence the division of labor: the designer takes the IP rent / the yards compete for low-margin "assembly."

THE DUOPOLY AND WHAT DO THE CHINESE HAVE TO DO WITH IT?

WinGD (Winterthur Gas & Diesel / Winterthur / Switzerland) - the direct heir of Sulzer (engines since 1898).

• Wärtsilä bought Sulzer in 1997
• in 2014-2015 it created a JV with CSSC (70/30), and in 2016
• exited completely - 100% to CSSC.

China controls 1/2 of the world's developers of the fleet's "heart" + is simultaneously the planet's largest shipbuilder.

In August 2025 - January 2026 CSSC absorbed its longtime rival CSIC (deal ~115.2B yuan / ~$16B), creating a giant with assets of ~400B yuan (~$56B), ~17-21% of global orders and control over 62% of the global orderbook + 80% of orders for new container ships.

WHY IS THIS CRITICAL?

More than 80% of the volume of world merchandise trade goes by sea

Practically every large container ship / tanker / bulker / gas carrier in the world is driven by an engine of MAN B&W or WinGD design.

**an annual rent stream tied to the installed base, not to the cycle of new orders.

The logic is simple - the engine of the future = the royalties of the future.

*The winner of the ammonia-methanol race will lock in IP rent from the new fleet for decades.

😎😎😎😎

😎 A Few Thoughts on Africa 😎

DEMOGRAPHICS

~1.5B population / surpassed 1.5B in 2024 - 5x its 1960 level.

By 2050, expected to reach ~2.5B / by 2100, ~3.5B (medium variant, UN DESA WPP 2024).

Sub-Saharan Africa is the only region set to nearly triple: from 1.2B to 3.4B by the end of the century (Population Connection).

By 2100, 10 African countries will increase their populations at least fivefold:

Angola / Burundi / the DRC / Malawi / Mali / Niger / Somalia / Uganda / Tanzania / Zambia.

Age structure: children under 15 make up 41% of the population / youth aged 15-24 a further 19% - this is, in effect, the "demographic window," which is only now opening, at a time when it has already closed for China, Europe, Russia, and Japan.
In short, even the most bearish case = an increase of 1B people, which, any way you look at it, is a great many.

This means the economy will expand on the back of the rising generation.

Population = demand = a healthy market = appreciating asset values.

By this formula, Africa is the only region where the equation works "in the positive" throughout the entire 21st century.

RESOURCES

An enormous basket of resources.

Around 30% of the world's reserves of critical minerals:

~55% of global cobalt reserves
~38% of manganese
~80% of platinum
~62% of chromium
~25% of natural graphite

Concentration by country:

• Guinea ~25% of the world's bauxite
• South Africa - #1 in manganese
• Zimbabwe - top 5 in lithium
• DRC - ~70% of global cobalt production

Demand:

Lithium in 2024: +30%
Nickel / cobalt / graphite /

REEs: +8%

The energy sector accounted for 85% of the growth in demand for battery metals (IEA)
Also expected: +40% in demand for copper and REEs / +60–70% for nickel and cobalt / +90% for lithium by 2040.

In addition:

Hydrocarbons (Nigeria, Algeria, Angola, Mozambique + LNG projects across all the African players = infrastructure)
Uranium (Niger supplied a significant share of France's nuclear power plants)
Gold (Ghana, Sudan, Mali)
Vast tracts of agricultural land and fresh water in the Great Lakes and the Congo Basin.

WEATHER

The Atlantic "conveyor" that drags warm water from the tropics northward (part of the Gulf Stream).

The AMOC will weaken by 18-43% by the end of the 21st century even under a very high-emissions scenario / a full collapse this century is unlikely, per a Caltech 2025 assessment.
In the event of an AMOC collapse: Europe dries out, the temperature differential between northern and southern Europe grows by ~4°C, the African and Asian monsoons weaken, and the Southern Hemisphere ocean releases more CO₂.

From this, the following follows:

Part of northern Europe becomes less comfortable - a redistribution of people and investment southward.
Part of tropical Africa undergoes a physical redistribution of precipitation - drought in some places (the Sahel), intensified rainfall in others - which redraws the map of agricultural investment attractiveness.

The UAE and Arabia - on a 30-50-year horizon - risk becoming barely habitable on account of temperature, with the added powerful factor of the U.S.

Iran dynamic (which spurs an even greater flight of capital from the region) - after which their "demographic + tax bonus" deflates like a balloon.

Africa inherits their function as a growing market.
Accordingly, we may see a "migration of capital toward the equator, to the high-altitude plateaus" (Ethiopia / Kenya / Uganda / Rwanda / Tanzania / South Africa) - many sovereign wealth funds already hold land allocations there.

😎😎😎😎

😎 Spring 2023 😎

SVB blows up in 48H, then Signature and First Republic follow.

The fiasco: the bank loaded up on Treasuries at ~1.5% / rates got hiked / the securities went underwater.
On paper everything's fine (at face value) In reality - a goose egg.
Such "paper" losses across the system by the end of 2025 - ~$306B.
Until recently the regulator promised to count it all honestly.
Now let's break down the revamped scheme for 2026.

The scheme:

eSLR - FREED UP $219B

On November 25, 2025, the federal reserve / OCC / FDICgov rewrote the leverage rule for systemically important banks.

Tier 1 capital requirements at their "subsidiaries" were cut by ~28% - roughly $219B that they NO LONGER have to hold as a cushion.

Vote at the Fed: 5 against 2 / Barr and Cook against ("overkill").
Passed anyway - took effect April 1, 2026.

**Without a single vote in the House BASEL III ENDGAME

The 2023 version raised requirements on the largest banks by nearly 30%.

The current version: under the new Vice Chair for Supervision, Bowman - CUTS capital across the system by ~$87.7B. They called this move "proper calibration," but in fact it's a rollback of the tightening.

AOCI It was precisely the failure to count underwater securities in capital that finished off SVB.
Essentially the one place where they smeared the terms across 6 years, leaving room to "maneuver":

regional banks' "don't count the losses" option (opt-out) is being eliminated, meaning the portfolio drawdown is finally being forced straight into capital.

The transition is stretched out to 2032.
In year one you exclude 100% of the loss, and only by 2032 - the goose egg.

• What's actually happening?
• Capital is being freed up NOW, while the one hole (the one that killed SVB) is being smeared out over 6 years at a leisurely pace.

NOW THE MAIN THING - WHERE'S CONGRESS, WHICH WAS SUPPOSED TO PUT THE BRAKES ON THIS?

Senator Warren tried to repeal eSLR via a resolution (S.J.Res.110), introduced March 4, 2026. After that it went to committee and quietly died (the resolution) without a single vote. 0 cosponsors + the window closed.

• So what's the trick of this whole act?
• Publicly - "we've learned the lessons of SVB, the system is sturdier."

In fact:

1) the cushion was cut by $219B immediately
2) the Basel tightening was turned into a loosening
3) the one honest item (counting the losses) was put on the brakes until 2032.

And all of this - by a decision of 3 agencies, bypassing the House + Congress didn't even get to a vote.

😎😎😎😎

😎 Why bother researching when you can just buy the Index? 😎

The thesis is pretty simple:

• why the hell do research if the money is going to be where the institutions are + why waste your time when composite instruments exist?

• $COIN50 - an index from Coinbase in partnership with MarketVector Indexes, tracking the 50 largest and most liquid digital assets by market capitalization.

Asset Selection Methodology

Universe - all assets traded spot on Coinbase Exchange with a USD/USDC/USDT pair and covered by the CCIX price index from CCData.

Then comes fundamental filtering on 3 (well, 4) criteria:

a) Tokenomics - pegs are banned (stables / wrapped assets) / exchange tokens excluded / the bulk of supply must be in free float.

b) Blockchain architecture - public blockchain with all attributes (transparency / immutability / permissionless consensus / peer-to-peer).

c) Security - no known vulnerabilities / availability of approved custodians.

d) Track record - minimum 360 days since the token's TGE.

Selection procedure: top-40 by cap make it in automatically / the remaining 10 slots are filled from positions 41-60 (priority to existing index components); if that's not enough - they fill up to 50 from any next-ranked candidates.

Rebalancing is quarterly - February / May / August / November. Between rebalances, the dominants can grow above 50% on market dynamics alone.

Current Structure (Extreme Concentration)

BTC: 51.22%
ETH: 26.71%
XRP: 8.17%
SOL: 4.63%
DOGE: 1.76%

Current weights:

Top-2 = ~78%
Top-5 = ~92.5%

THE REMAINING 45 ASSETS SHARE LESS THAN 8% OF THE WEIGHT AMONG THEMSELVES.

$COIN50 is effectively a proxy for BTC / ETH with a very thin alt tail. Long-tail ALT components have weights of 0.02–0.04% - dust used to fill the index.

Sector Breakdown

Smart Contract Platforms (22)
Infrastructure & Application (9)
Decentralized Finance (7)
Memecoins (4)
Payments (4)
Media & Entertainment (3)
Store of Value - $BTC

$COIN50 is structurally heavily correlated with $BTC and $ETH.

If you actually want to slam yourself into real alt-exposure, you probably want equal-weighted indexes. That said, the spread inside $COIN50 is enormous: from -57% to +49% - which tells us that the cap-weighted approach in crypto dilutes alpha inside the tail.

Following $COIN50 Logic, You can build your own Indexes

Some examples of how to assemble such "indexes":

1) Ex-megacap (BTC/ETH excluded) Logic: the simplest way to isolate alt-beta. Take the same universe, throw out TOP2-5, and weight whatever's left by FDV or by your own "indicators / hype / love for the project."

2) Equal-Weighted indexes Logic: every position gets the same weight (e.g. = 50 assets × 2%) / regardless of market cap, with rebalancing returning weights back to equal.

3) Sectoral indexes Logic: any sectors you want (RWA / DeFi / L1 / Infrastructure & Middleware / AI / DePIN / etc.) - betting on a narrative.

4) Smart-beta / factor indexes logic:

a) Momentum - overweight tokens with the best returns over N months (Bitwise Crypto Momentum)

b) Quality - tokens with real revenue / cash flow (Token Terminal–based indexes)

c) On-chain activity - weights by real usage metrics

You can also use ready-made instruments:

1) VANECK Crypto and Blockchain Innovators ETF ($DAPP) - equity exposure to crypto companies

2) 21shares Crypto Basket Index ETP ($HODLX)

3) Grayscale Digital Large Cap Fund ($GDLC) - top-5 / BTC-heavy

4) HASHDEX Nasdaq Crypto Index ETF ($HDX) - Brazil / EU via UCITS wrappers

5) Bitwise ($BITW) - top-10 / market-cap weighted

Moral of the Post

Any newbie / amateur / hobbyist can now assemble a portfolio however their soul desires.

Either one-click buy the $GDLC index, or endless Twitter-research and DIY attempts to build an "inflation-beating portfolio" (you won't).

The only difference is how many hours of your life you're willing to throw at it.

*Inspired by TradFi / ETFs / composite indexes.

😎😎😎😎

😎 From bakery to feedlot 😎

-23% YEAR OVER YEAR

USDA: US wheat harvest 1.53B bu (41.7 MMT), -23% y/y.

• HRW 463M bu (12.6 MMT) -42% y/y
• HRS 434M bu (11.8 MMT) -5%
• SW 263M bu (7.2 MMT) +2%
• Durum 66.4M bu (1.8 MMT) -23%

The shortfall is concentrated exactly where bread flour gets milled.

FEED

Wheat/corn 7.54/4.59 = 1.64
**Back to where the season started, and the premium keeps widening even on the corn rally.

Gulf HRW 12 = $332/MT vs feed parity $192
Spread $140 (73%), was $102 (53%)

HARVEST

HRW: 90%+ harvested / grade firmed up to 1 HRW

• protein 12.5%
• TW 60.2 lb/bu (79.2 kg/hl)
• moisture 11.2%
• TKW 28.7g / FN 356

FLOUR:

• extraction 71.2%
• wet gluten 28.4%
• absorption 59.2%
• stability 14.9 min
• loaf 859 cc

• The south of the HRW belt is in severe to extreme drought / exceptional in places.
• Fall planting is in question - that's already a 2027/28 story.

SW: winter 68% / spring 28%

Protein 9.0% (range 6.7-14.0) / FN 292, 1 SW
Northern Idaho is doing 150-200 bu/ac - hence the low protein.

HRS: 24% harvested, G/E 51%.
SD protein 15-16% with yields above expectations, MN 50-70 bu/ac, MT holding.
Weather on the Northern Plains is turning - rains and normal temperatures.

Durum:

• MT 15%
• ND 12%
• G/E ND 47%
• MT 28%

-23% is a direct hit to pasta / couscous / semolina.

FUTURES (SEP U26)

• KCBT HRW $7.5425 (+$0.40)
• CBOT SRW $6.7475 (+$0.35)
• MIAX HRS $6.7825 (-$0.01)
• Corn $4.59 (+$0.20)

• MAY K27: HRW $7.825 / SRW $7.1175 / HRS $7.3425
JUL N27 below the Mays in HRW and SRW. Contango = still no turn

FOB / YEAR OVER YEAR

Gulf HRW 12: $332/MT (was $234, +42%)
Gulf SRW: $281 ($223, +26%)
Gulf HRS 14: $317 ($272, +17%)
PNW HRW 12: $329 ($240, +37%)
PNW SW: $254 ($241, +5%)
N. Durum Gulf: $300

Russian export duty:

Wheat 19-25.08: 721.1 RUB/MT (+120.79%)
Corn 284 (+170.48%)

Germany cut to 20.6 MMT (-11.2%)
France + the July EU cut (soft wheat 124.4 MMT / exports 29.0 / ending stocks 12.9)

PHYSICAL

US Gulf HRW 11: $329-339/MT
Romania 12.5: $270-275
France: $271-274
Argentina 12.0: $240-245
Russia 12.5: $220-225

The US is carrying $105-115/MT over the Black Sea (was $60-100).
The gap is growing, but there's physically less Black Sea material.

FREIGHT

BDI 2863 (-226 / -7.3% on the week)
Panamax earnings $20,055/day (-3%)

Gulf:

Egypt $44 (+2)
Turkey $45 (+2)
Algeria-Tunisia $45 (+2)
Morocco $44 (+2)
Nigeria $52 (+2)
China $71
Japan $73-75

SPREAD:

• Egypt $45 / $28.
• Morocco $59 / $29.
• Tunisia-Algeria $58 / $28

Hormuz has stopped fertilizer flows into East and South Africa ahead of planting.
It will show up in early-2027 yields.

CURRENCIES (y/y)

JPY 159.3 (+8.29%)
EUR 0.864 (+1.23%)
EGP 50.25 (+4.13%)
RUB 84.25 (+5.16%)
BRL 5.217 (-3.31%)
AUD 1.412 (-8.03%)

• The dollar weakened after the July retail sales miss (-0.6%)
• EGP down 4% on the year with wheat +42% in $$$ (Baladi subsidy)

BUYING

Net sales (as of 06.08): 255,931 MT
Outstanding 7.47 MMT = 35% of USDA's 21.1 MMT (was 33%)

• Export demand for Black Sea hasn't woken up yet. Once it does, basis goes up - on top of futures that have already rallied.

BOTTOM LINE

Harvest: -23%
HRW -42%.

You can't plug a hole in tonnage with quality.
Novorossiysk is down, Germany and France are cut.
Feed has gone 73% above parity, everything went into the mill.

November CPI remains the point where this shows up on the receipt.

😎😎😎😎

😎 Why Ken Griffin from CITADEL has nothing to do with it 😎

We propose looking at the other side of this event and why you can't specifically claim that CITADEL was involved.

The algorithm is fairly simple - we'll need just 6 steps to prove the opposite:

2 LEGAL ENTITIES.

On July 27, the note saying the market was underpricing the Fed's hawkish shift was published by Citadel Securities - the market maker.

Author - Frank Flight, head of macro strategies.

The fund's public portfolio was bought on July 30 by Citadel - the hedge fund.

These are 2 different companies with separate P&Ls, separate teams and information barriers that exist not out of Griffin's goodwill but by regulatory requirement.

Their presence is the subject of constant SEC and FINRA examinations.

The whole "scheme" falls apart on the very first thesis: who exactly passed what to whom across a barrier that gets audited professionally and regularly.

Neither WSJ / nor FT / nor Reuters draws a causal link between these 2 events.

They were linked by a post on X:

Shkreli - CONVICTED IN 2017 OF SECURITIES FRAUD (2 counts of securities fraud + conspiracy, tied to his hedge funds MSMB and Retrophin / 7-year sentence).

Shay Boloor - A TABLOID GUY.

MALICIOUS INTENT?

• On July 29 the FOMC left rates unchanged, 3.5-3.75%, vote 9-3.
• Meaning Citadel Securities' forecast didn't pan out.

IF THIS WERE REALLY MANIPULATION - THEN IT WAS BUILT ON A PUBLICLY FAILED THESIS - THAT'S NOT HOW ACTUAL SCHEMES GET BUILT.

+ a detail everyone misses: before the note, the swap market was already pricing in ~40% probability of a hike. Citadel Securities didn't create that fear - they commented on it.

WHO SELLS THE FUND'S POSITION?

Margin calls were issued to the fund by Bank of America / Goldman Sachs / JPMorgan - THE FUND'S PRIME BROKERS. They were also the ones selling the positions.

Leverage up to 4x - the fund's own decision.
Concentration in an overcrowded trade = the fund's own decision.

The forced seller was created by leverage and the risk management of 3 banks = KEN HAS NOTHING TO DO WITH IT.

THE SELLOFF CALENDAR.

It started ~July 1 - with the signal that Meta might resell excess compute capacity.

That day Micron -10.6% / AMD -6.9%.
The two-year thesis of an "eternal GPU shortage" came to an end on those time.

SOX peaked on June 22 and by July 28 was in a bear market = -25-28%.

In parallel, retail leverage was collapsing in Korea:

margin debt fell from a record 38.63T won (June 24) to ~32.7T by July 23 / KOSPI -33% in a month.

The note came out July 27 - 4 weeks after the collapse began.
The chronology doesn't allow you to assign it as the cause.

WHY CITADEL?

When a $10-16B block gets dumped on the market, THERE ARE PHYSICALLY ABOUT 5 BUYERS IN THE WORLD:

CITADEL / MILLENNIUM / POINT72 / BALYASNY.

BEING THE BUYER IN A TRADE LIKE THAT IS A CONSEQUENCE OF BALANCE-SHEET SIZE, NOT PROOF OF A CONSPIRACY.
*at enormous risk you take on a book that just fell 30-50%, into a falling market, with no guarantee the bottom is in.

ASYMMETRY.

A firm with $69B under management is supposed to be believed to have gone, for the sake of one trade, into coordinated market manipulation - that is, into a criminal case that would cost it its license, its business and management's freedom.

For the sake of a portfolio that would have come to it through the prime brokers anyway.

WHAT DO WE ACTUALLY HAVE?

Confirmed:

1) losses
2) letter to investors July 24
3) margin calls
4) sale of the portfolio NOT CONFIRMED BY CITADEL around July 30.

• Both sides declined to comment.
• There is no official confirmation of the trade.
• It won't surface on the regulatory side until the Q3 13F, deadline ~November 16.

A fund at 4x leverage in the most overcrowded trade of the year caught a correction, got calls from 3 banks, and sold the book to the only party that could eat it whole.

• to whom?
• only the 13F will show

😎😎😎😎

😎 Market Psychology 😎
contrast = leverage

*a couple of cascading green candles and everyone forgets that the distance is negative.

in any assessment there are 2 quantities:

• level (where we are)
• contrast (how much this differs from what was nearby)
• most people think they trade the level.
• in reality everyone trades the contrast.

***THE BRAIN DOESN'T SEE PRICE / THE BRAIN SEES DIFFERENCE.

THE EXPERIMENT
Haynes, Roth, Stadler & Heinze (2003), J Neurophysiol

148-channel MEG + EEG / 8 subjects.

targets were embedded in a field of high contrast. physically identical / only the orientation of the surrounding background changed.

• against a "matching" background the target was perceived as dimmer
• to be perceived as equal, the second one only needed 60% of the physical contrast
• 40% of the signal simply disappeared • meanwhile cortical responses to such a pair were INDISTINGUISHABLE
• and to physically identical stimuli they DIFFERED SIGNIFICANTLY.

• a standard of 0.50 matched 0.33 in perception
• a standard of 0.32 - matched 0.18.

dipole localization assigned the early components to primary visual cortex / 87-88% of variance explained.

***THE CORTEX ENCODES NOT THE SIGNAL / BUT THE SIGNAL DIVIDED BY CONTEXT.

THE LATE RESPONSE IS MORE ACCURATE THAN THE EARLY ONE

there are 2 cortical components in the study:

• early (80 ms) - already distorted by the background, but with error at high contrasts
• late (130-180 ms) - predicts perception almost perfectly

the difference comes from slow horizontal connections inside the cortex / 0.1-0.3 m/s.

they need time to drag the context across the network.

meaning the picture doesn't arrive finished.

it is BUILT UP / and it is built up already spoiled by the background.

***YOU DON'T SEE THE MARKET AND THEN INTERPRET IT / YOU ALREADY SEE THE INTERPRETATION.

economists arrived at the same formula knowing nothing about visual cortex - We recommend everyone dig into this work:

• Kahneman & Tversky (1979) - what decides is the deviation from a reference point, not the final level

• Bordalo, Gennaioli & Shleifer (2012) - the weight of an attribute in a choice equals its contrast with the background of alternatives

• Barber & Odean (2008) - retail buys whatever stood out abnormally in volume or return

one and the same computational principle on 2 floors / normalization to the surroundings.

THE CROWD ISN'T DUMB - IT'S NORMALIZED.

synchrony doesn't require the participants to be foolish:

• Bikhchandani, Hirshleifer & Welch (1992) - a cascade arises with fully rational agents
• the public signal outweighs the private one / from there the herd moves in one direction
• Shiller (2017) - a narrative spreads along an epidemic curve = not a logical one.

***THE CROWD'S UNIFORM BEHAVIOR IS NOT COLLUSION AND NOT IDIOCY = IT IS THE EFFECT OF IDENTICAL NORMALIZATION.

MEMORY ≠ SHORT
MEMORY = SKEWED MEMORY

• Greenwood & Shleifer (2014) - expectations are built on the most recent realizations, not on history
• De Bondt & Thaler (1985) - hence overreaction and reversal on the long horizon
• Malmendier & Nagel (2011) - a personally lived-through cycle weighs more than 100 years of statistics

whoever hasn't sat through -80% doesn't know what -80% is - he had no skin in the game.
he READ about it or SERVED it up to the masses / that's a different background and a different scale.

HOW THIS LOOKS ON THE MARKET

• after a month of chop, any move reads as an impulse
• after -60%, a 15% bounce reads as a reversal
• after a year of growth, a correction reads as the end of the world

the price is the same / the background is different / the decisions are opposite.


WHAT CONCLUSION CAN WE DRAW?

• the crowd doesn't see the level / the crowd sees the contrast
• contrast is set by the background / the background is set by the last few weeks
• that's why the behavior reproduces itself every cycle without change
• and that's why it's predictable

**On perception, the crowd never sees the price. only the difference.

😎😎😎😎

😎 Heirs of the Sultan of Sulu (PETRONAS) or how an 1878 concession nearly seized the assets of an oil giant? 😎

THE SETTING

Malaysia, the state of Sabah in northern Borneo - oil-rich. State oil giant Petronas.

THE INSTRUMENT

• an arbitration award built on a colonial treaty from 1878.
• in 1878 the Sultan of Sulu leased the territory (present-day Sabah) to a British trading company in exchange for annual payments.
• after independence, Malaysia kept paying the heirs a symbolic token - until it stopped in 2013.

WHO HOLDS IT?

8 Filipino heirs of the last Sultan of Sulu.
behind them - London litigation fund Therium, which sank tens of millions into the claim.

THE MECHANICS

the heirs launched an arbitration, and in February 2022 Spanish arbitrator Gonzalo Stampa issued a $14.92B award - the second-largest arbitration award in history (the heirs originally demanded $32.2B, and the process itself cost over $20M - lawyers and researchers across 8 jurisdictions).

under the UN New York Convention, the award became enforceable worldwide - even while Malaysia was contesting it at home.

the heirs went after Petronas assets abroad - because Petronas, as a state company, was treated as an "extension" of the state.

WHO STUMBLED?
Petronas / Malaysia.

at the peak of the threat, bailiffs in Luxembourg were walking into banks with account-freeze orders - a 19th-century colonial paper was genuinely putting the assets of a modern oil giant at risk.

HOW DID THEY GET OUT?
through the courts of the seat of arbitration.

• the Paris Court of Appeal annulled the award (2023, final - December 9, 2025: the arbitration clause "inapplicable," the reference to the British Consul-General of Borneo vanished back in 1963).

• the Hague Court of Appeal refused enforcement (upheld by the Dutch Supreme Court on September 6, 2024). Stampa was found criminally liable in Spain in December 2023.

• as of 2026 the heirs have essentially lost, and Petronas is litigating against the Therium fund in Jersey.

WHAT'S THE TAKEAWAY FROM THIS CASE?

in essence - this case shows the limit of the lever.
an arbitration award born from a colonial concession is one of the most powerful instruments out there (global enforceability under the UN convention), surviving 144 years and a change of sovereignty.

• but the lever only worked as long as it held up procedurally. once the courts of the seat knocked out the foundation - the gate collapsed.

unlike Gertler's royalties or the Franco-Nevada stream (which we'll cover in the next episode), which sit on the title of the asset itself, an arbitration lives exactly until its first successful annulment (fragility).

😎😎😎😎

😎Layering - how a fake buyer wall gets painted in the Bitcoin order book 😎

Guys, today we're breaking down one of the most cynical manipulation tactics in the exchange order book. It doesn't require huge capital, doesn't leave obvious traces, and works even against experienced algorithms.

It's called Layering.

What it is

You open the BTC order book and see a ladder of buy orders.
Not one, but an entire layer.

Say, 800 contracts at 80,000, 1000 at 79,900,
1200 at 79,800, 1500 at 79,700.

• Buy-side depth is three times greater than sell-side depth.

The brain automatically concludes:
there's a big player sitting below, the market is protected, there's nowhere to fall.

The problem is that this wall of liquidity can be a phantom.

Layering is when one participant places multiple orders at different levels to create an illusion of strength. It's not just one large order that vanishes on touch. It's an entire system designed to make you believe there's real demand below the price. In reality, there isn't.

Why it works

The market sees an imbalance. Sellers pull their orders back, unwilling to fight the "buyer." Buyers get more aggressive, thinking they're running out of time. Algorithms register the growing depth and recalibrate their models. Everyone starts acting as if there's a mega-fund sitting under the market. And then the orders disappear one by one. First the 800, then the 1000, then the 1200. Price plunges, stop-losses trigger, and you hand your money over to those very orders that no longer exist.

This isn't price manipulation - it's manipulation of your perception of the market.
Nobody tells you "buy."
You're shown a picture in which buying seems like the only reasonable decision.

How to spot it

There are several signs worth watching so you don't walk into the trap.

1) the lifespan of orders.

A real large order usually sits for quite a while, because there's genuine intent to execute behind it. Phantom orders live for seconds or milliseconds, especially as price approaches.

2) behavior on touch.

If an order gets partially filled and immediately refills, that's a sign of real demand. But if it vanishes before price even reaches it, or gets pulled at the first serious touch, that's a reason to be suspicious.

4) synchronicity.

Phantom levels often appear and disappear simultaneously, as if controlled by a single bot. If you see several orders at different levels get pulled at the same moment, that's not a coincidence.

4) repetition.

The same structure can appear in the same section of the order book several times a day. A real buyer isn't that predictable. A slick bot - absolutely.

But there's a nuance almost nobody talks about. Even if you've learned to recognize phantom liquidity, that doesn't mean you know what to do next. Because Layering is often used not on its own, but as part of a more complex construction. The same wall can be a trap for longs, or it can be cover for a short position.

What exactly does the MM do after pulling the orders?
How do you tell simple manipulation from preparation for a real move?
And why is it sometimes better not to trade at all when you see Layering in the book?

That's for the next post.

In the meantime, think about it: how many times have you seen a perfect buyer wall
that then vanished without a trace?
And what happened to the price a minute later?

😎😎😎😎

😎 Ferguson's Limit 😎

THE PATIENT

USA, 2026

debt ≈ $38–40T (~123% of GDP)
pressing against the $41.1T ceiling.

FY2025 deficit - $1.83T.
interest on the debt - $970B a year, breaking through a trillion in FY2026.

already the third-largest budget line - bigger than defense, bigger than Medicare.

THE EXORBITANT PRIVILEGE

not an arbitration, not a concession - the dollar's reserve status.

a 1944 paper (Bretton Woods) + the world's habit of parking savings in Treasuries = the right to borrow endlessly and cheaply.

the lever is 80+ years old: survived the gold decoupling, 2008 and covid. more durable than the Sulu concession of 1878, and monetized far harder.

THE CAPTIVE JURY

formally the lever belongs to the US Treasury.
in practice - to everyone obligated to buy: foreign central banks, pension funds, commercial banks. and behind them stands the Fed - the litigation fund of last resort, ready to top up liquidity in any proceeding, at any stage, against any counterparty.

3 HANDS AND ONLY 1 POCKET

in 2024 the US crossed "Ferguson's Limit":

interest payments exceeded defense spending - first time since the 1930s.
the rule is simple: a great power that pays its creditors more than its soldiers stops being a great power.

and at exactly that moment, two more players line up for the same capital and the same energy:

rearmament:

WORLD - $2,887B a year (11th consecutive year of growth)
NATO - 5% of GDP by 2035
EU mobilizing €800B
Germany broke its constitutional debt brake to make it happen.
AI CAPEX: the "big four" - ~$725B planned for 2026 (+77% YoY)
hyperscaler bonds >$100B
private credit for data centers up fivefold.

in June 2026 the BIS gave the construction its proper name:
the "fiscal-financial nexus" - sovereign debt, AI credit and shadow banking stitched into one fabric. one seam tears - it pulls the rest.

COLLATERAL DAMAGE

the welfare state.
as always - the one with no lobbyists in the queue for capital.

global development aid: -23.1% in a year - the largest cut on record.
Medicaid/SNAP: -$1.1T, up to 16M people losing coverage.
Britain cutting aid from 0.5% to 0.3% of GNI - stated openly, "to pay for defense."

• welfare - warfare.
• routinely, no referendums, just a line item in the budget.

THE QUIET DISCOVERY

no single court can void this lever. so the challenge runs as discovery, not as trial - and the world's reserve managers are filing exhibits:

the dollar's share of global reserves - 56.9%, lowest since 1995.
central banks bought 863 tonnes of gold in a year, 57% of purchases anonymous.
GOLD - its best year since 1979. Moody's, meanwhile, stripped the US of its last AAA.

every tonne of gold in a central bank vault is, in essence, a motion for annulment.

CLOSING ARGUMENTS

the Sulu award lived exactly until its first successful annulment - Paris knocked out the foundation, and the gate collapsed.

a reserve currency is the same genre of lever, with one difference: there is no "court of the seat" to void it with a single ruling. it gets annulled not by decision, but by the tonne, quarter by quarter.

so there are 3 endings:

• FORTRESS - the debt gets quietly monetized, platforms fuse with the defense complex, resilience becomes a mandated asset class

• REPRESSION - the most likely and the most boring: rates below inflation, savings forcibly earmarked for defense and AI, paid for by a silent middle class

• STORM - the Sulu case in reverse: the bailiffs show up not at Luxembourg banks, but at Treasury auctions.

and in all 3, the same position wins:

sell the state its fortress and sell capital the way out of it.

😎😎😎😎

😎 Spoofing and how one fat order deceives the entire bitcoin order book 😎

Guys, last time we covered Layering, when the interested party paints a whole wall of buyers. Today let's talk about its older brother - Spoofing.

This is when a single order pretends to be your best friend, but is actually luring you into a trap.

What is it?

You open the BTC order book.
At the best bid price there are 200 contracts, below that another 300.

Suddenly an order for 2,000 appears.
Below the market there seems to be a huge buyer.

Everyone thinks: that's it, now we're flying up.

Sellers pull their orders, buyers get more aggressive, algorithms register the increase in depth. The price starts creeping toward that order and when it almost touches it, the order disappears.

That's Spoofing.

Formally everything is legal: the order was real, it could have been filled. But its purpose was not to buy, but to create a false impression. It's like installing a dummy camera, only in reverse: not to scare off, but to attract.

An important point: not every order cancellation is spoofing.
An ordinary market maker constantly repositions quotes, because price and risk change. A high cancellation rate by itself proves nothing. The question is the economic function of the order. If its only meaning was to change the behavior of others,
then you're looking at manipulation.

Why does it work?

The market reacts not only to real trades, but also to the expectation of trades. One large buy order creates the illusion of demand.

Participants readjust their prices: someone cancels sells, someone decides to buy earlier, algorithms change their models.
In the end the price moves not because someone actually bought a lot, but because everyone believed a big buy was coming.

To control the market, you don't have to execute trades. It's enough to control the information about trades that supposedly might happen.

How to spot it?

1) lifetime.

A real large order usually sits longer, because there is intent behind it.
Spoofing order doesn't live long, especially as the price approaches.

2) behavior on contact.

If the order is partially filled and replenished, that's a sign of genuine demand.
If it's pulled before contact or right after the first serious touch, that's a warning sign.

3) synchronicity.

Spoofing order often appears and disappears in tandem with other actions on the opposite side of the book.
For example, simultaneously with the disappearance of a large bid, aggressive sellers may appear.

4) repeatability.

If the same large order appears at the same level several times in a short span and disappears each time, that's no longer coincidence.

But there's a nuance nobody talks about.

Even if you've learned to recognize spoofing, that doesn't mean you know how to make money on it. Because spoofing is often used not for petty manipulation, but as cover for a larger position. The same phantom order can be a trap for longs,
or it can be a smokescreen behind which the interested party prepares a reversal.

Spoofing is one large order that creates the illusion of demand or supply and disappears before contact. Layering is several orders at different levels that together create the appearance of a deep wall of liquidity.

Spoofing hits one point, layering hits the whole range. Spoofing is easier to notice because of the anomalous size, layering looks like ordinary depth. Both tools often conceal a real position, so the main thing is not to try to outplay the manipulator, but to exit the market.

😎😎😎😎

😎 From bakery to feedlot 😎

ONCE AGAIN, FREIGHT MATTERS MORE THAN THE LAB

SIEVE

HRW closed: 1 HRW (580 samples / 550 expected)

• protein 12.3% (2025: 12.1 / 5Y: 12.3)
• TW 60.4 lb/bu (79.4 kg/hl)
• TKW 29.3g vs 30.7 on the 5Y (smaller kernel)
• FN 355 (2025: 370 / 5Y: 362)
• SKCS 60.4, large fraction 66%

flour (58 composites):

• extraction 71.2% = MLU 202 mill ≠ crop
• absorption 58.9%
• dev 6.1 min / stability 13.7 min
• loaf 865 cc (target 850)

**Good dough quality, but the kernel is small.

HRS: harvested ND 86 / MT 75
SD-MN-PNW effectively closed

260 samples = 58% of the sampling target:

• protein 14.7% / FN 416
• DHV 73% = 1 NS
• TW 60.3 vs 61.5 on the 5Y
• TKW 30.0 vs 31.5

2025 final was 14.4% / DHV 62.

***Sample count went from 175 to 260 in a week.
**Drought gave the concentrate and took the weight.

SW: harvested ~99%

• 386 samples / 390 expected
• protein 9.0%, range 6.4-14.8
• 48% below 9.0 / 34% in 9.0-10.0 / 18% above 10.0
• FN 288 (2025: 317 / 5Y: 332)
• TW 60.9 = 1 SW

SRW final:

2 SRW, protein 10.0 (2025: 9.3 / 5Y: 9.5)

• FN 305
• TW 59.9
• defects 1.5 vs 2.3

Durum: 79% ND / 80% MT

44 samples = 34% of the sampling target:

• 2 HAD (5Y held 1 HAD)
• protein 15.2 (2025: 14.2 / 5Y: 14.2)
• TKW 35.4 vs 42.9 a year ago
• TW 59.9 vs 61.7
• HVAC 79 vs 84 / 86 - rain hitting color and vitreousness
• FN 405, dockage 1.7

**Best protein in years. Everything else is shit quality.

BOARD

• Dec SRW 7.2525 (-0.0875)
• Dec KCBT HRW 7.9850 (-0.0375)
• Dec MIAX HRS 7.4500 (0.0000)
• Corn Dec 5.3025 (-0.0650) / Beans Nov 12.97 (-0.13)

KCBT curve: Dec 7.985 - May 8.165, then Jul 8.06 / Sep 8.05

MGEX: Dec 7.45 - May 7.745 - Sep 7.5825

**Carry only lives out to May for now.

WASDE: world production higher, trade lower.
RU+Ukraine exports cut by a combined 4.0 MMT, offset by Argentina / Australia / Canada.

RU duty 16-22.09:
wheat 1146.3 rub (-2.80%) at an indicative $227.5
corn 617.5 rub (-31.24%)
barley 0.

SovEcon is penciling in -3.2 MMT off RU wheat exports, down to 41.4 MMT. Total grain 49.4 vs 53.8 a month earlier.
Reason: attacks on vessels in the Azov-Black Sea corridor.

HOLD

Black Sea, Russia / Ukraine:

• Egypt 90 / 28
• Morocco 105 / 29
• Tunisia-Algeria 105 / 28
• China 95 / 56
• Europe 46 / 28
• Turkey 85 / 30

**Risk premium 3-4x. Russian 210-215 FOB has stopped being cheap at discharge.

US Gulf:

Egypt 45 (+2)
Turkey 46 (+2)
Japan 77 (+1)
China 76 (+1)

PNW China 36 / Japan 38

BDI 3507 (-121)
Capesize 6080 (-347)
Panamax 2407 (-41)
Supramax (+44)
Handisize (+40)

USW grain freight index 519.7 vs 486.9 a week earlier.

**Capesize is falling, but the grain sizes aren't.

CASH

FOB Oct-Dec:

• HRW 11 pro Gulf 349-352
• SRW Gulf 305-315
• France 285/288 spot
• Romania 12.5 280-285
• Argentina 12.0 255-260
• Ukraine 11.5 220-225
• Russia 12.5 210-215

US/RU spread ~140

Nearby $/bu:

Gulf HRW Ord 9.39 (+0.31)
Gulf HRS 13.0 9.00 (+0.30)
Gulf SRW 8.30 (+0.06)
PNW SW 7.55 (-0.20)

Northern durum Texas Gulf Oct 8.57 ($315/MT)

Dollar firmed on oil + treasury yields and Fed expectations - market pricing ~70% on +25 bp.

EUR 0.862
JPY 153.7
RUB 84.12
EGP 51.33
BRL 5.122
CAD 1.387

😎😎😎😎

😎 From bakery to feedlot 😎

FUNDS EXIT INTO THE DOLLAR / RUSSIAN DUTY KEEPS FALLING

SIEVE

HRW final: 1 HRW (580 samples / plan 550)

• protein 12.3% (2025: 12.1 / 5Y: 12.3)
• TW 60.4 lb/bu (79.4 kg/hl)
• TKW 29.3g vs 30.7 5Y
• FN 355 (2025: 370 / 5Y: 362)

**Kernels smaller than normal, everything else flat
*Flour - waiting for the Crop Quality Report

HRS: harvested ND 92 / MT 88
SD 100 / MN 98

374 samples = 83% of the sample set:

• protein 14.7% / FN 417
• DHV 75% = 1 DNS (a week earlier 1 NS)
• TW 60.2 vs 61.5 5Y
• TKW 29.9 vs 31.5

2025 final was 14.4% / DHV 62

***Samples up from 260 to 374 in a week.
**Grade improved on vitreousness, test weight below 5Y

SW final: 1 SW

• 386 samples / plan 390
• protein 9.1% (2025: 10.0 / 5Y: 10.2)
• FN 288 (2025: 317 / 5Y: 332)
• TW 60.9

SRW final: 2 SRW, protein 10.0 (2025: 9.3 / 5Y: 9.5)

• FN 305 / TW 59.9
• defects 1.5 vs 2.3

Durum: 82% ND / 90% MT

63 samples = 47% of the sample set:

• 1 HAD (a week earlier 2 HAD)
• protein 14.9 (2025: 14.2 / 5Y: 14.2)
• TKW 35.9 vs 42.9 a year earlier
• TW 60.3 vs 61.7
• HVAC 78 vs 84 / 86 - rains
• FN 400, dockage 1.5

**Grade pulled up on test weight.
*Protein still the best in years, kernel weight and HVAC weak.

BOARD

• Dec SRW 7.1425 (-0.1100)
• Dec KCBT HRW 7.8375 (-0.1475)
• Dec MIAX HRS 7.4125 (-0.0375)
• Corn Dec 5.2750 (-0.0275) / Beans Nov 13.04 (+0.07)

KCBT curve: Dec 7.8375 - May 8.02, then Jul 7.95 / Sep 7.965

MGEX: Dec 7.4125 - Jul 7.7575 - Sep 7.6125

**Funds cutting longs ahead of US-China. HRW hit hardest, HRS holding on Canada.

US exports: 325.9KT for the week
Commitments 9.1 MMT = 43% of USDA forecast 21.1

Canada: wheat -11% y/y, spring 26.5 MMT, durum 6.4 MMT

Russian duty 23-29.09:

wheat 725.7 RUB (-36.69%) at indicative $225.9
corn 158.4 RUB (-74.35%)
barley 0.

**Minus 420 RUB = ~$5/T of room for Russian exporters.

HOLD

Black Sea, Russia / Ukraine:

• Egypt 90 / 28
• Morocco 105 / 29
• Tunisia-Algeria 105 / 28
• China 95 / 56
• Europe 46 / 28
• Turkey 85 / 30

**Risk premium holding. Iran shifting Russian imports to the Caspian.

US Gulf (steady):

Egypt 45
Turkey 46
Japan 77
China 76

PNW China 36 / Japan 38

BDI 3370 (-137)
Capesize 5768 (-312)
Panamax 2251 (-156)
Supramax / Handysize +48

**Grain panamax -6.5% for the week, hasn't reached US rates yet.

CASH

FOB Oct-Dec:

• HRW 11 pro Gulf 343-346
• SRW Gulf 299-310
• France 291/294 spot
• Romania 12.5 285-290
• Argentina 12.0 255-260
• Ukraine 11.5 220-225
• Russia 12.5 210-215

US/RU spread ~130

Nearby $/bu:

Gulf HRW Ord 9.24 (-0.15)
Gulf HRS 13.0 9.06 (+0.06)
Gulf SRW 8.14 (-0.16)
PNW SW 7.40 (-0.15)

Northern durum Texas Gulf Oct 8.57 ($315/MT)

Dollar strengthened on the Fed rate hike, DXY at a 6-week high (+1.4%).

EUR 0.871
JPY 156.9
RUB 84.19
EGP 52.10
BRL 5.141
CAD 1.398

😎😎😎😎

😎 THE ECONOMY OF AUTONOMY is out 😎

In 2025-2026 the world is going through neither a cyclical downturn nor the "geopolitical turbulence" that is so convenient to cite in quarterly investor letters.

What we are seeing is a structural rebalancing of resources / institutions / infrastructure:

away from the "peace dividend" that fed the Western middle class for 30 years, toward security / sovereignty / physical resilience.

And it happened INSTITUTIONALLY:

constitutional amendments + multi-year funds + mobilization laws + building codes + central bank balance sheets.
No single politician can reverse it within one business cycle.

What's inside:

• 4 records of the year: $2.9T military spending / $348T global debt / gold at $5,589 / US effective tariff at 17–18%
• why US interest payments ($970B) now cost more than defense ($917B)
• the grid, not the model, as the real bottleneck of AI
• war-risk insurance as the most honest barometer of geopolitics
• 32 segments of the economy of autonomy, scored for a small team
• "who to be friends with": 11 regime types and how to find the gatekeeper of rent

One formula to take away:

PROPERTY = ASSET * ACCESS * JURISDICTION
If any multiplier goes to zero, everything goes to zero.

Read here: https://chasetrust.substack.com/p/the-economy-of-autonomy

😎😎😎😎

X: https://x.com/chasetrustorg
Dash: https://chasetrust.org
Substack: https://chasetrust.substack.com
IG: https://www.instagram.com/chasetrustorg

😎 From bakery to feedlot and a summit without bread 😎

GRANARY

HRS: harvested ND 95 / MT 95

374 samples = 83% of the survey:

protein 14.7% / FN 417
DHV 75% = 1 DNS (2025: 1 NS / DHV 62)
TW 60.2 vs 61.5 5Y avg
TKW 29.9 vs 31.5

***60 samples still in the lab, final numbers in a week
**Class went up, but the weight didn't come back

Durum: ND 87 / MT 95

86 samples = 64% of the survey:

1 HAD (2025: 2 HAD)

• protein 14.7 (14.9 a week earlier / 5Y 14.2)
• TKW 35.7 vs 42.9 a year earlier
• TW 60.4 vs 61.3 5Y avg
• HVAC 80 vs 86 5Y avg
• FN 402, dockage 1.7

**The grade held up. Kernels are small / color is rain-damaged

Winter wheat finals:

HRW: 1 HRW, protein 12.3, FN 355, TKW 29.3
SRW: 2 SRW, protein 10.0, FN 305
SW: 1 SW, protein 9.1 / FN 288 vs 332 5Y avg

**SW FN is below 300, read the spec before you sign

New season: 2027/28 winter wheat 20% planted, 5% emerged

KS / OK / TX dry / +5-10°F above normal

TICKER

• Dec SRW 7.0325 (-0.1100)
• Dec KCBT HRW 7.6200 (-0.2175)
• Dec MIAX HRS 7.1350 (-0.2775)Corn Dec 5.2825 (+0.0075) / Beans Nov 13.19 (+0.15)

• KCBT curve: Dec 7.62 - May 7.7825, then Jul 7.745 / Sep 7.7725
• MIAX: Dec 7.135 - Jul 7.495 - Sep 7.4125
• CBOT SRW: Dec 7.0325 - Sep 7.3075

**KC over Minneapolis by ~49¢. HRW priced above HRS
**Only SRW holds a carry across the whole curve

Trump-Xi summit: ag deals were expected, nothing delivered so far. They'll probably start hyping it today.

US exports: 267.5 KT for the week ending 17.09
Commitments 9.37 MMT = 44% of the USDA forecast of 21.1

DAMPER

Russia zeroed the export duty on wheat / barley / corn through 31.12.2026, retroactive from 1.09

***Reports are still quoting the 30.09-06.10 calculation: wheat 640.3 rub / corn 220.6. There's nothing to collect.

Delivered Egypt (FOB + freight):

• Russia 12.5: 210 + 90 = ~300
• Ukraine 11.5: 220 + 28 = ~248
• Romania 12.5: 285 + 23 = ~308
• France: 278 + 35 = ~313

• US SRW: 293 + 45 = ~338
• US HRW 11: 339 + 45 = ~384

**The zero duty saves $8-14/t. The risk premium in freight is $62.
*The math doesn't add up at all.

Russia and Turkey are discussing a corridor, while ships are being hit by both sides.
France is loading its first vessel to Yemen since 2022 / the EU is picking up Egypt and Sudan.

**Demand has moved to whoever isn't being shot at

FAIRWAY

US Gulf:

• Egypt 45
• Turkey 46
• Japan 76 (-1)
• China 75 (-1)
• PNW: China 36 / Japan 38

Duluth: Europe 63 (+5) / Morocco 69 (+5)

• BDI 3426 (+56)
• Capesize 5784 (+16)
• Panamax 2407 (+156)
• FBX container 3367 (-79)
• Panamax ~$21.7k/day, +7% w/w.

USW grain freight index 520.5, third week on a plateau (early September 486.9).

**The Lakes get pricier toward the end of the shipping season.
*Bulkers are firming, containers are running into the shallows.

BERTH

FOB Oct-Dec:

• HRW 11 pro Gulf 339-341
• SRW Gulf 293-304
• France 278/281 spot
• Romania 12.5 285-290
• Argentina 12.0 260 to 250
• Ukraine 11.5 220-225
• Russia 12.5 210-215

• US/RU spread ~129

Nearby $/bu:

• Gulf HRW Ord 9.12 (-0.12)
• Gulf HRS 13.0 8.84 (-0.23)
• Gulf SRW 7.98 (-0.16)
• PNW SW 7.30 (-0.10)

Basis: Gulf HRW / HRS firmer

SRW weaker
PNW softer

**Gulf HRW +$109/t vs last year (+47%)

The dollar pulled back on cheap oil, but it's up for a second week: the market is pricing in more Fed hikes.

• EUR 0.877
• JPY 157.2
• RUB 84.15
• EGP 51.75
• BRL 5.185
• CAD 1.414

TROUGH

Soybean meal +$14-20/t w/w: crushers are short of spot beans.

meal NOLA 460-455 / Brazil 425-410
soybeans NOLA 525-535
corn NOLA 245-250 / Brazil 235-245 / Ukraine 225-235

Fishmeal: southern Peru has landed 4.5 KT of its 251K quota

Next center/north quota ~500 KT
Water stays record-warm through March 2027.
Peru FOB 65% 2600/2650

Meal was the only thing that rose this week.

😎😎😎😎

😎 Beneath the Index 😎

Over the past six months, equities and bitcoin have led the basket while gold has lagged.

The latter is easy enough to explain: with 10-year real (TIPS) yields close to 2.8%, gold has gone back to trading largely as a function of real rates.

• What we find harder to reconcile is credit.

The S&P 500 is still within roughly 1% of its august high, yet high-yield has already moved lower, and around three-quarters of the index's constituents declined in September.

Historically, divergences of this sort have not lasted long, and the way they have tended to close is worth keeping in mind.

😎😎😎😎