Auren

real money moves in silence. i follow the footsteps everyone else is too loud to hear

[Auren @AurenPro]

@Ox_Cheetah Worth knowing the dealer often makes more in the finance office than on the car itself, because they can mark up the bank's rate by up to 2 or 2.5 points and keep the spread for the whole 84 months.

[Auren @AurenPro]

@reidgrows The farm test works on way more than bitcoin, run it on half the things people online call passive income and most of them fail the same way.

[Auren @AurenPro]

@lctrnotes A flat decade from PE 22 is survivable if you never have to sell in it, and that's the part that worries me more, since people are already pulling from brokerage accounts to pay for groceries.

[Auren @AurenPro]

@Beaver_0x The part nobody screenshots is what the $1.3 billion cost him, because the same Coke bought with borrowed money at 7% compounds into a very different number than Coke bought on insurance float.

[Auren @AurenPro]

@0x_Squirrel And the $39 million started with a guy sitting in the insurance department in Lincoln reading reports the clerks had to dig out for him.

[Auren @AurenPro]

3:15. Buffett explains where the $177.5 billion machine started.

January 1951, a Saturday. He's 20 and walks into GEICO because his hero Ben Graham chairs the board. A man named Lorimer Davidson gives him four hours on how an insurer makes money.

At 3:47 he says the one insurance course he took in school "had no value to me." He can't even remember the textbook.

At 5:00 he tells you what he did instead. He sat in the Department of Insurance in Lincoln and kept asking for examination reports the clerks had to dig out from the bottom of the stacks.

Float was never a secret. The filings were public, and the math fits on a napkin. He was the one person reading them.

At 5:31 he gives the advice for anyone starting cold: read twenty years of Berkshire annual reports, insurance section only.

My piece below is about what that Saturday turned into. Free capital underneath every stock he ever bought, and why most people today spend the gap instead of letting it work.

[Auren @AurenPro]

https://t.co/FJ62YCiqf0

[Auren @AurenPro]

@0xLupenn This lecture has been free since 2005 and the reason it didn't make everyone a quant is the same reason most people never finish lecture 6.

[Auren @AurenPro]

@lorneseth She was selling insurance on a crash without any of the reserves her brother never let himself run without

[Auren @AurenPro]

@Allx032 Anyone who has raised money when desperate knows this one, you fight over the rate for weeks and hand over the upside in a clause you read once.

[Auren @AurenPro]

@0xCortez Funny how people argue for hours over 1% of expected return and never check the 1% they're guaranteed to lose every year.

[Auren @AurenPro]

@Beaver_0x 22-year-olds priced an unknown object at a third of its value in 90 seconds with zero information. Most retail investors do worse with a Bloomberg terminal and six months of research.

[Auren @AurenPro]

@0x_Squirrel The 1% looks small until you see it compound against you for 30 years straight.

[Auren @AurenPro]

Shannon's 28% for 30 years wasn't about the filter. He sized bets using Kelly - the formula he helped build with John Kelly at Bell Labs. Most people chasing better signals are solving the wrong problem. The edge was never in detection. It was in how much you put down when you actually had something.

[Auren @AurenPro]

@fomoliver Schwarzman measured one quarter out and the math said sell. Fink measured one decade out and the math said dilute. Same numbers, different time horizon - $240 million vs $62 billion difference. The spreadsheet doesn't tell you which window to use. That's the actual decision.

[Auren @AurenPro]

@finillin_dev The 50.75% tilt is real but the post skips the hard part - you need thousands of trades before that edge separates from noise. Most people don't quit because they lack math. They quit because 600 trades in, a 51% edge looks exactly like a coin flip and feels worse.

[Auren @AurenPro]

Cover proved something most portfolio managers still refuse to accept - the return wasn't in the stocks. It was in the rebalancing. Two mediocre names, daily trim-and-add, 70x over 20 years. The math doesn't care about your thesis on the company. It cares about variance and how you harvest it.

[Auren @AurenPro]

16:35 - Druckenmiller repeats a line he caught from Reed Hastings. Manage for quarterly earnings and you are dead.

He compounded around 30% a year for thirty years without a losing one, and he spends this talk saying that one thing a dozen ways. The whole system plays for the next quarter, and the few who play for the next decade quietly take everything.

Watch 13:48. He sets IBM next to Amazon. IBM has missed its earnings three times since 2006, guards every quarter, shrinks R&D as a share of sales, and buys back 43 billion in stock at an average price of 189 that now trades at 142. Amazon missed nine of its last nineteen quarters and did not care, poured the money back into the business, and ran R&D from 5% of sales to 10%. One protected the quarter. The other built the decade.

At 7:50 he describes his own method, and it is stranger than the buy-and-hold crowd assumes. He manages for the long term through a series of short-term positions, and every one of his biggest returns came in chaos, not calm.

Then 2:04, the sentence the rest of it rests on. Government, business, the Fed, money managers, all of them now managing for the short term, and that is the disease.

Loud money chases the print, defends the quarter, buys its own stock at the top to flatter a number. Quiet money thinks in decades, sits through the noise, and moves when the chaos finally shows up.

You are not behind because you picked the wrong stock. You are behind because you have been keeping score on a clock that was never yours to win on.

[Tigerflow @tigerfl0w]

https://t.co/Ye9wCJ9Jiv

[Auren @AurenPro]

51:12 - watch this part twice.

López de Prado runs a simulation. Sharpe ratio of the underlying strategy: zero. No edge. Pure noise.

He repeats the backtest 100 times. The expected maximum Sharpe that emerges from nothing: 2.5.

That number would get a fund launched. It would pass most due diligence. It would look, on paper, like a career-defining discovery. And it is arithmetic fiction.

This article builds the only thing that catches it - the filter.

Most of what I see on this platform about "AI trading agents" is about the generator. The scanner. The swarm. How many strategies it finds per hour. That is the easy half and everybody knows it. The hard half is that a machine like this will hand you statistically significant, beautifully backtested, completely fake results - and it will do it confidently, forever, unless you build the thing that says no.

The article gives you the thing that says no.

t-stat above 3.0, not 2.0 - one parameter shift that cuts false positives from fifty per thousand to one. Walk-forward, not one long backtest. Costs subtracted before you judge anything. A trial log where every hypothesis is recorded - because a Sharpe of 2.0 found on the first try and a Sharpe of 2.0 found on the four hundredth are not the same finding.

At 9:36 López de Prado explains why a single person cannot do this alone anymore. The level of competition requires an industrial process - a strategy to produce strategies. The article translates that into four roles you can build over a weekend: Scanner, Researcher, Validator, Reporter.

At 41:27 he explains why Google engineers fail at finance. The data is non-IID - every observation is contaminated by every other. Face recognition assumes independence. Markets do not. That is why plugging financial data into a general-purpose ML model is not a shortcut. It is a trap.

And at 53:05 - the line that should be framed on the wall of every quant who backtests anything:

"The better you get at backtesting, the less useful it is."

Read the article. Then watch the full 75 minutes. The math is free. The willingness to build the filter before the generator is the only barrier left.

[Auren @AurenPro]

exactly - income is what the world decides you're worth, savings rate is what you decide you're worth to yourself

[Auren @AurenPro]

I ran one subtraction and one division on my bank account last month.

That single calculation told me more about my financial future than my salary, my job title, or my portfolio ever did.

A person earning $60,000 who saves 25% keeps $15,000 a year.
A person earning $150,000 who saves 5% keeps $7,500.
The one making less than half as much is building wealth twice as fast. This isn't motivation. It's arithmetic.

Your salary doubled last year? Congratulations. But if your spending doubled with it - you didn't get richer. You got further from freedom.

Morgan Housel calls it "purchasing independence." Every dollar you don't spend is a dollar that works for you until you die.

Most people have never calculated their savings rate. Not once. The formula is embarrassingly simple - one subtraction, one division. And the number it gives you maps directly to how many years stand between you and never needing to work again.

The article breaks down the exact table, the three traps that silently kill your rate, and why cutting $500 in spending is mathematically more powerful than earning $500 more.

Your salary is a vanity metric. Your savings rate is the only number that's entirely yours.

[Superior @andreysuperior]

https://t.co/6fbpq78hsH

[Auren @AurenPro]

@Beaver_0x the course is free, the textbook is sixty dollars, and the only real barrier is twenty-six hours of silence most people will never trade their scroll time for