Cody

Money. Logic. Long-term thinking.
Breakdowns of markets and personal finance.

[Cody @cody0xx]

https://t.co/cBBPirGIaZ

[Cody @cody0xx]

Every trader wants the impossible.
A signal that is fast and certain.
A Belgian mathematician showed why you cannot have both.
Her name is Ingrid Daubechies.
And the wall she helped draw sits under every chart you trade.

The idea is the time-frequency limit.
The more precisely you know when something happened, the less precisely you know what frequency it carried.
The more you smooth the signal to understand what it was, the later you see it.
That is not a weakness of your indicator.
That is the law.
A fast signal reacts on time.
It also screams at noise.
A smooth signal ignores the noise.
It also arrives late.
No moving average escapes this.
No filter escapes this.
No model escapes this.

Every indicator you tune sits somewhere inside the same box:
fast and noisy,
or smooth and late.
Pick your poison.
Most traders blow up because they refuse to accept the tradeoff.
They keep changing settings.
Shorter window.
Longer window.
Different filter.
Cleaner trigger.
Faster confirmation.
But they are not optimizing out of the problem.

They are just moving around inside it.
This is Heisenberg uncertainty for markets.
Not particles.
Signals.
You cannot perfectly localize both time and structure.
The market gives you a choice:
react early with less confidence,
or react later with more confirmation.
Daubechies wavelets became powerful because they live close to the edge of that limit.
They do not break the wall.
They respect it better.
That is the lesson.
The best signal is not the one that is magically fast and clean.

It is the one that knows exactly what it is sacrificing.
The crowd keeps tuning indicators like the wall is optional.
Daubechies showed the wall is real.
And every trader is trapped inside it.

[Rossst.03 @Rossst_03]

https://t.co/iZLuBLz082

[Cody @cody0xx]

The most expensive skill in trading is not finding a winning streak.
It is proving the streak was not luck.
That is the math YuanYuan Xu teaches on a chalkboard.
Strip away the P&L screenshots and every strategy comes down to one question:
after N bets, how far can results drift from the truth by pure chance?
Win 10 trades in a row.
Looks amazing.
Proves almost nothing.
Run 10,000 trades with the same edge.
Now the noise has less room to hide.
The tool behind this is called a concentration inequality.
It puts a hard bound on randomness.
How much can luck flatter you?
How much can it punish you?
How far can your observed win rate move away from your real win rate before the result becomes suspicious?
That is the entire quant problem in one idea.
A trader says:
“I’m winning.”
A quant asks:
“Would this still happen if I had no edge?”
That question is the difference.
Because markets create fake proof all day.
A good month.
A clean backtest.
A beautiful equity curve.
A 70% win rate over a tiny sample.
Most of it is noise with good lighting.
Concentration inequalities force the curve to answer:
are you signal, or are you just randomness being generous?
That is why firms pay heavily for people who understand this math.
They are not paying for someone to admire a streak.
They are paying for someone to kill the fake ones.
YuanYuan Xu builds the idea from zero, in public, for free.
No guru language.
No trading mysticism.
Just the proof that separates skill from luck.
The edge is not winning.
The edge is knowing when winning actually means something.

[Cody @cody0xx]

A $120,000 salary can still leave you broke. A financial planner once published a client’s numbers: $120,000 a year. Top 10% income. Eight years of work. Net worth: $14,000. The internet called it fake. It wasn’t. The math was simple: six-figure salary, zero-percent savings rate. Every raise became a bigger apartment, a newer car, a better gym, nicer restaurants, and a lifestyle that expanded just fast enough to keep him stuck. The scary part is not that he was broke. The scary part is that he thought he was doing fine. That is where psychology enters. Dunning and Kruger showed that people who perform the worst often overestimate themselves the most. Not because they are stupid. Because the skill required to see the mistake is often the same skill they are missing. The six-figure earner with nothing saved has the same problem. If he never calculates the savings rate, the illusion never breaks. High income feels like progress. Nice apartment feels like progress. A better car feels like progress. But the number that matters is brutally simple: what percentage did you keep? If the answer is zero, you are not building wealth. You are just renting a richer identity. The lecture goes even deeper with the marshmallow test. For years, people sold it as a story about discipline. Kids who waited won. Kids who grabbed the marshmallow lost. Then researchers changed one thing: some kids were lied to before the test. They were promised something better, then the promise was broken. Those kids ate the marshmallow faster. Not because they had less willpower. Because their environment taught them waiting does not pay. That is the part most money advice ignores. Your savings rate is not only discipline. It is trust. Trust that the future is stable enough to plan for. Trust that money kept today will matter tomorrow. Trust that waiting will actually reward you. If your life has trained you that promises break, bills appear, jobs vanish, and money disappears, your brain will treat every dollar like it needs to be used now. So the answer is not just “try harder.” The answer is build a system your brain can trust. Automatic saving. Lower fixed costs. Clear rules. No lifestyle upgrades for every raise. A savings rate you can see every month. Because the illusion only survives while the number stays hidden. A high salary can make you look rich. Your savings rate tells you if you actually are.

[Cody @cody0xx]

The raise will not feel like a raise for long.

That is the trap.

Your brain adapts to new income the same way it adapts to a new apartment.

At first, it feels different.

Then it becomes normal.

This is why people are so bad at predicting what will make them happier.

Lottery winners often drift back toward ordinary happiness after the shock fades.

People who suffer devastating losses often recover emotionally far more than outsiders expect.

The brain is not measuring life objectively.

It is constantly resetting the baseline.

John Gabrieli, MIT professor of brain and cognitive sciences, shows this with simple experiments.

Ask 30 students the probability that two share a birthday.

Most people guess low.

The real answer is about 70%.

The brain trusts intuition and misses the math.

Then he reads:

sour,
candy,
sugar,
honey,
chocolate,
cake.

Then asks who heard the word “sweet.”

Hands go up.

But the word was never said.

The brain invented the missing piece from context.

That is not a small bug.

That is how perception works.

Identical lines can look different.
Identical greys can look different.
A word you never heard can feel remembered.
A salary increase can feel like freedom while your spending quietly absorbs it.

That is lifestyle inflation.

Your income rises.
Your standard resets.
Your expectations update.
Your old luxuries become basics.

Then you wonder why earning more did not make you feel ahead.

The problem is the frame.

Your brain sees a bigger salary and calls it progress.

But the real question is simpler:

what percentage did you keep?

Savings rate cuts through the illusion.

Not income.
Not status.
Not lifestyle.
Not how rich it feels.

Just:

money in,
money out,
percentage kept.

That number is hard to hallucinate.The raise will not feel like a raise for long.
That is the trap.
Your brain adapts to new income the same way it adapts to a new apartment. At first, it feels different. Then it becomes normal.
John Gabrieli, MIT professor of brain and cognitive sciences, shows why with simple experiments.
Ask 30 students the probability that two share a birthday. Most guess low. The real answer is about 70%.
The brain trusts intuition and misses the math.
Then he reads: sour, candy, sugar, honey, chocolate, cake.
Then asks who heard “sweet.”
Hands go up.
But the word was never said. The brain invented it from context and called it memory.
That is how perception works.
Identical lines can look different. Identical greys can look different. A word you never heard can feel remembered.
And a salary increase can feel like freedom while your spending quietly absorbs it.
That is lifestyle inflation.
Income rises. Standards reset. Expectations update. Old luxuries become basics.
Then you wonder why earning more did not make you feel ahead.
Your brain sees a bigger salary and calls it progress.
But the real question is simpler:
what percentage did you keep?
Savings rate cuts through the illusion.
Not income. Not status. Not how rich it feels.
Just money in, money out, percentage kept.
If your salary doubled but your savings rate did not move, you did not gain freedom.
You upgraded the cage.
Your brain adapts.
Your savings rate tells the truth.
If your salary doubled but your savings rate did not move, you did not gain freedom.

You upgraded the cage.

This is why financial progress needs math, not mood.

Your brain adapts.

Your savings rate tells the truth.

[Superior @andreysuperior]

https://t.co/6fbpq78hsH

[Cody @cody0xx]

My friend was auto-rejected by quant desks for two years.

State school.
No PhD.
No network.
Then Two Sigma offered him $650,000.
I asked what changed.
He sent me one link:

a free Cornell lecture from 2014, Steven Strogatz at a chalkboard explaining chaos.
That sounds unrelated until you watch it.
Chaos is not randomness.

It is a system that follows rules but becomes impossible to forecast because tiny errors explode over time.

A small difference in the starting point becomes a completely different future.

That is the market.
Not random enough to ignore.
Not stable enough to predict cleanly.
This is why most backtests lie.
They fit the path that happened.

But live trading gives you a path that is slightly different at the start and completely different later.
Beginners see this and try to build a smarter forecast.
Pros build systems that survive bad forecasts.

They care about:

position sizing,
drawdowns,
correlation,
execution,
slippage,
tail risk,
when to stop trading,
and how much damage one wrong regime can do.
That is what finally clicked for him.

Quant trading is not about finding the model that predicts the future.

It is about building a machine that stays alive when the future refuses to match the model.
That is why the best people sound boring.
They talk about survival.

Everyone else talks about prediction.
Bookmark this and watch the lecture below.

2:00 — what chaos actually is
22:00 — how tiny errors become different worlds
45:00 — why backtests fit the past and miss the future
58:00 — why pros manage risk instead of worshipping forecasts

[Cody @cody0xx]

The market can beat a genius model without being smarter than it.

That is the part people miss.
Elon can make Grok brilliant.
It still has to trade on a curved surface.

Maryam Mirzakhani spent her life studying those surfaces — spaces where two paths that begin almost identical can separate exponentially.
That sounds abstract until you put it inside a trading system.
Run the same bot twice.

Same model.
Same signal.
Same edge.
Same rules.

But one run gets a slightly worse fill.
One enters a thinner book.
One hits a different volatility pocket.
One gets caught in a crowded unwind.
One faces a tiny delay.
At first, the difference looks like nothing.

Then the paths split.
One account compounds.
The other dies.
That is not always because the model was wrong.

Sometimes the geometry changed.

Markets are not flat systems.

Liquidity bends the path.
Volatility bends the path.
Correlation bends the path.
Leverage bends the path.
Crowding bends the path.
Execution bends the path.

This is why backtests lie so beautifully.
They show you one clean path through a surface.

Live trading gives you a nearby path that may end in a different universe.
That is what the best quant firms pay for:
not just people who can find signals,

but people who understand how the surface changes the signal.
A weak trader asks:
“Where does the path go?”
A strong system asks:
“What surface is this path moving through?”

That is the difference.
The edge is not just the model.
The edge is knowing when the geometry will tear the model apart.

[Cody @cody0xx]

The most dangerous risk is the one too small to respect.

That is the lesson from Arnold diffusion.

Ke Zhang studies how systems that look stable can still drift apart through hidden channels.

No obvious failure mode.
No giant shock.
No clean warning sign.

Just a tiny push, repeated long enough, moving the system somewhere it was never supposed to go.

That is trading.
Your strategy usually does not die from one huge mistake.
It dies from the small assumption your backtest rounded away.

A few extra basis points of slippage.
A fee that looked harmless.
A spread that widens at the worst time.
A fill that only existed in simulation.
A tiny bias in execution.
A delay that compounds over thousands of trades.

None of these look fatal alone.

That is why they are fatal together.
The backtest says the system is stable.
Live trading quietly finds the hidden channel.

Then the account starts drifting.
Slowly at first.
Then permanently.
Then visibly.

By the time you notice, the damage is no longer a single event you can explain.
It is the accumulated cost of being slightly wrong every day.
That is why the best risk teams do not only ask:

“What can blow us up?”
They ask:
“What tiny force keeps pushing us in the wrong direction?”
Because a big risk announces itself.

A small constant risk becomes the path.
Stability is not safety.

It is just the part of the drift you have not measured yet.

[Cody @cody0xx]

The same strategy can make one trader rich and bankrupt another.

Not because one picked better trades.
Because one survived the path.
Imagine two bots.

Same signals.
Same entries.
Same exits.
Same 60% win rate.

The only difference:
one uses 3x leverage.
The other uses none.
After the same year of trades, the unlevered bot is up.

The levered bot is dead.
That is Ole Peters’ whole lesson.
Your account is not an average.
It is a sequence.

Up 50% and down 50% is not zero.
It is -25%.
Now add leverage.
A normal losing streak becomes a crater.
A recoverable drawdown becomes liquidation.
A good strategy dies before its edge has time to show up.

The levered bot was not wrong more often.

It just bet too large to survive being wrong in a row.

That is the difference between expected value and lived reality.

On paper, a bet can be profitable.
In one real account, with one real path, it can still take you to zero.

You do not get to average across thousands of alternate universes.
You get one bankroll.
So the first rule is not “maximize return.”
It is “stay alive.”

Because the edge only pays the trader who is still there to collect it.
Renaissance did not win by being right the most.

It won by never letting one path end the game.

[Cody @cody0xx]

The most dangerous chart pattern is the one that looks obvious.

Because randomness can draw beautiful shapes too.

That is the lesson hiding inside Yilin Wang’s work on the geometry of random curves.

Her field studies something most traders never think about:

how far a curve is from pure chance.

There is even a number for it.

Loewner energy measures how much structure a curve carries compared to natural random geometry.

A circle scores zero.

It is the cleanest possible case.

Markets create the opposite problem.

Every day, price draws millions of curves:

breakouts,
failed breakouts,
channels,
wedges,
spikes,
reversals,
perfect-looking setups.

The human brain sees the shape and wants meaning.
But the quant question is colder:

could randomness have drawn this exact thing anyway?
If yes, there is no edge.
Just a nice-looking accident.

That is why firms like Citadel pay extreme money for people who can separate structure from noise.
Not by vibes.
By measurement.

A trader sees a curve and says, “this is a pattern.”
A quant asks, “how unlikely is this pattern under randomness?”
That difference is everything.

The edge is not the shape.
The edge is knowing the shape should not be there.

[Cody @cody0xx]

The market does not kill you with the risk you measured.

It kills you with the risk your model rounded to zero.

That was Mandelbrot’s warning.

Wall Street built its machinery around a mild market:

normal distributions,
smooth price moves,
rare crashes,
manageable tails.

But Mandelbrot looked at real prices and saw something else.

The market is wild.

It is rough at every scale.
Minutes look like days.
Days look like months.
Months look like decades.

The violence repeats.

That is why fractals matter in finance.

They show that markets are not smooth curves with occasional surprises.

They are systems where extreme moves are part of the structure.

The bell curve says a once-in-a-century crash should be almost impossible.

Reality keeps producing those crashes every few years.
This is where elegant models become dangerous.

Black-Scholes assumes mild randomness.
Value at Risk assumes the tail behaves.
Portfolio models assume the future will resemble the measured past.
Then the wild day arrives.

LTCM had Nobel laureates, leverage, and beautiful math.
The market gave them a move their model treated as nearly impossible.

The fund was gone in weeks.
That is Mandelbrot’s lesson:
risk is not thin-tailed.
markets are not polite.
the worst day is closer than your model thinks.

A few trading days can decide decades of returns.
And the risk that ends you is usually hiding in the tail your spreadsheet made invisible.

The market is not a bell curve.

It is a fractal with teeth.

[Cody @cody0xx]

Most people do not lose the wealth game because they are lazy.

They lose because they are playing on a line while capital plays on a curve.

Albert Bartlett spent his life warning people about this:
“The greatest shortcoming of the human race is our inability to understand the exponential function.”

That quote explains money better than most finance books.
Your salary is linear.
You work.
You get paid.
You save a piece.
It adds up slowly.
But ownership compounds.

A business.
Equity.
Real estate.
A productive asset.
Those do not just add.
They can double.

At 10% a year, money roughly doubles every 7 years.

That means €10,000 can become:

€20,000 after 7 years
€40,000 after 14
€80,000 after 21
€160,000 after 28
€320,000 after 35

Same starting point.
Different function.

This is why compounding feels fake until it is too late.
For years, nothing dramatic happens.
Then the curve pulls away from everyone still thinking in straight lines.

The rich did not just out-earn you.
They owned assets sitting on the exponential curve.
The curve was never hidden.
It was free the whole time.

Most people just were not trained to see it.

[Rossst.03 @Rossst_03]

https://t.co/V3snBurQsU

[Cody @cody0xx]

The most dangerous signal is the one your backtest finds too quickly.

That is the lesson from prime numbers.

Peter Sarnak spent decades studying the boundary between randomness and hidden structure.

Primes look random.
But they are not random at all.
Every prime number is fixed forever. Deterministic. Inevitable.

And still, their distribution behaves so much like randomness that proving structure inside it is brutally hard.
Markets have the same problem, but with money attached.
A backtest can find a pattern in an afternoon.

That does not mean you found edge.
It may only mean noise had enough room to draw a pretty shape.

Clean equity curve.
Good Sharpe.
Nice entry rule.
Perfect hindsight.

That is not proof.
That is the beginning of the test.
Real edge has to survive:

out-of-sample data,
regime changes,
transaction costs,
slippage,
position limits,
crowdedness,
and the fact that the market adapts.

The hard part is not finding a pattern.
The hard part is proving the pattern is not randomness pretending to be structure.
The market is like the primes:

it looks random,
it may contain deep structure,
but almost every “obvious signal” is fake.
Edge is not pattern recognition.

Edge is pattern survival.

[Cody @cody0xx]

The blueprint for multi-agent AI is not new.

Marvin Minsky wrote it decades ago.

Today, Anthropic can pay engineers up to $900,000 to build agent systems, but the core idea was already sitting in MIT’s “Society of Mind”:

intelligence is not one brilliant brain.

It is a swarm of simple specialists.

That idea matters even more in markets.

One giant model should not be trusted to trade alone.

Markets are too noisy.
Signals conflict.
Regimes change.
Risk hides in the tail.

A stronger system is a crowd of narrow agents:

→ one reads news
→ one tracks liquidity
→ one watches macro
→ one checks filings
→ one monitors volatility
→ one searches for contradictions
→ one has veto power

No single agent needs to be a genius.

The system gets stronger because they disagree.

That is the real lesson from Minsky.

A mind is not a monolith.

It is many simple parts competing, correcting, and constraining each other.

The AI industry still sells the dream of one all-knowing model.

But the systems that survive will look more like swarms:

many dumb specialists,
shared memory,
constant verification,
and no single point of failure.

The future of AI is not one model that knows everything.

It is many agents that stop each other from being stupid.

[Roan @RohOnChain]

GPT-6 Astra is the most powerful trading agent right now.

It gives you AGI-adjacent reasoning.

I've shown the exact way to use GPT-6 Astra at its HIGHEST benchmark.

You could literally build MOST complex trading models like hedge fund with it. https://t.co/VxsUknvnSu

[Cody @cody0xx]

The best market makers try to do something that sounds impossible:

touch every direction without taking up space.
That is why Kakeya sets are such a perfect metaphor.

A Kakeya set contains a line segment in every possible direction, yet can have almost zero area.
Every direction.
Almost no footprint.
Now look at a firm like XTX Markets.

Over $250B traded daily, but not as one giant directional bet.

The goal is different:

be everywhere,
quote constantly,
absorb flow,
rebalance inventory,
hedge exposure,
and avoid carrying the wrong risk.

From the outside, the book looks enormous.
Inside, the objective is to keep net exposure as thin as possible.
That is the market maker’s version of the Kakeya problem.

Maximum directional contact.
Minimum occupied risk
Most people think edge comes from taking a big position and being right.

But in market making, the edge often comes from the opposite:
being present in every direction,
getting paid for flow,
and disappearing before exposure becomes a thesis.

The money is not always in size.
Sometimes it is in the near-zero position that still touches everything.

[Rossst.03 @Rossst_03]

https://t.co/zeqJCFDNYZ

[Cody @cody0xx]

Wall Street pays $500,000 for a single rare discipline: distinguishing a repeatable statistical edge from an exceptionally lucky sequence of draws.

A backtest showing a 1.53 Sharpe ratio tricks almost everyone, because variance on a hot streak produces a chart that looks indistinguishable from genius.

Gerd Gigerenzer, director at the Max Planck Institute, has spent decades demonstrating how even seasoned professionals suffer from statistical blindness.

His work hinges on a vital boundary most market participants fail to draw:

Risk vs. Uncertainty: Risk belongs to closed systems with fixed, known probability distributions, like a roulette wheel. Uncertainty belongs to the real world and open markets, where the underlying probability parameters are never truly known.

The Illusion of Precision: A backtest spits out comforting metrics: Sharpe 1.53, maximum drawdown 8.8%. It gives the seductive impression of a closed-risk calculation, while masking raw market uncertainty.

Small-Sample Deception: Three years of strong daily returns is not a definitive proof of alpha. It is easily a random coin-flipper riding a favorable regime.

The Seduction of Metrics: The greatest danger in quantitative finance is never the unknown. It is placing blind faith in a precise number that was designed around flawed assumptions.

The actual edge is never the output of an optimized backtest.

The edge belongs to the risk operator who looks directly at a 1.53 Sharpe and asks the one question the metric actively conceals:

Did we engineer an authentic edge, or did we just survive a lucky draw?

[Furiokao @furiokao]

https://t.co/8wXADZ29zF

[Cody @cody0xx]

A mathematician shortened World War II by two years and saved an estimated 14 million lives by decoding a hidden pattern across millions of intercepted messages. The nation he saved chemically castrated him for his sexuality. By 41, he was dead.

His name was Alan Turing.

He broke the Enigma code the German military believed was unbreakable, and in the process, conceptualized the universal computing architecture that powers the screen you are reading right now.

The core realization behind his work dismantles everything school taught about the subject:

Arithmetic vs. Pure Math: School teaches calculation, rote memorization, and manual counting. Real mathematics is the science of structural relationships and abstract patterns that no brute-force method can uncover.

The Power of Pure Logic: Breaking an impossible cipher didn't come from faster computation; it came from constructing logical constraints that eliminated billions of incorrect permutations in seconds.

The Foundation of the Modern World: The same formal logic that ended a war laid the exact theoretical framework for modern algorithms, machine intelligence, and quantitative modeling.

A lead quant once noted that he requires every new researcher to study Turing’s approach before they touch a single live risk model.

Most people think they are bad at math because they were only ever taught the mechanical accounting version.

Turing demonstrated what real mathematics actually is: an unfair advantage in perceiving hidden structures before anyone else realizes they exist.

[Potato @Potatxxl]

https://t.co/iSxk8U2YAD

[Cody @cody0xx]

@andreysuperior The “how many engines are you running?” question is brutal because it cuts through income vanity.

A high salary with one engine is still fragile.

Real wealth starts when cash flow, equity, skills, and capital begin feeding each other instead of depending on one paycheck.

[Cody @cody0xx]

This is the exact layer crypto was missing.
Not another alpha feed.
Not another influencer call.
A pre-hype risk filter that asks the boring questions first:
who holds the supply,
where is the liquidity,
what changed before launch,
and can one wallet end the game?
Most rugs don’t need genius to spot.
They need speed.

[Cody @cody0xx]

The real unlock is not replacing Bloomberg with one tool.
It is unbundling the terminal into agents:
one watches filings,
one tracks news,
one reads social flow,
one builds charts,
one summarizes signal,
one alerts when something actually matters.
Bloomberg was a screen.
This is becoming a system.

[Cody @cody0xx]

@andreysuperior This is the number most people avoid because it removes the story.
Not “I make good money.”
Not “I’m pretty responsible.”
Not “this was an unusual year.”
Just:
what came in,
what stayed,
what leaked.
Your real savings rate is the receipt for your actual behavior.