quantfold

Quant trader · Wall Street alpha through mathematics & AI

[quantfold @quant_fold]

Michael Milken built a market worth hundreds of billions out of bonds Wall Street called trash.
Then the government gave him 10 years for it

He wasn't guessing either At Drexel Burnham Lambert, Milken ran the numbers on corporate bonds Wall Street had written off as junk, and found the market was mispricing the risk - default rates didn't justify the discount

He built an entire market out of that one insight, high-yield bonds, worth hundreds of billions, funding companies that couldn't get a traditional loan

Then came the same RICO investigation that raided Thorp's fund Milken pled guilty in 1990 to six felony counts. $600 million in fines. Ten years in prison

He served 22 months In February 2020, a sitting president pardoned him, three decades after the case that ended his career

The math about mispriced risk wasn't the crime that took him down - the specific trades around it were.
And the market he built outlived every one of his prosecutors

[quantfold @quant_fold]

Ed Thorp, 20 years, 20% annual returns after fees, zero losing quarters. He took Kelly's formula out of a physics paper and used it to beat blackjack first, then beat Wall Street

Before Thorp, professional gamblers just felt their way through bets. Casinos assumed the house edge was unbeatable, full stop

Thorp read Kelly's paper, built the math into a card-counting system, wrote Beat the Dealer, and casinos across Nevada started barring him at the door. So he did the same thing to the market instead

1974, Princeton/Newport Partners. Kelly sizing on every trade. 20% annualized after fees. Two decades. Not one losing quarter

The fund never blew up on its own math. It got shut down chasing Michael Milken's junk bond scandal at Drexel - Thorp was cleared, the firm wasn't spared

The formula never failed him. Someone else's crime did.

[quantfold @quant_fold]

@andreysuperior New alpha !
thanks for post

[quantfold @quant_fold]

Ed Thorp, 20 years, 20% annual returns after fees, zero losing quarters. He took Kelly's formula out of a physics paper and used it to beat blackjack first, then beat Wall Street

Before Thorp, professional gamblers just felt their way through bets. Casinos assumed the house edge was unbeatable, full stop

Thorp read Kelly's paper, built the math into a card-counting system, wrote Beat the Dealer, and casinos across Nevada started barring him at the door. So he did the same thing to the market instead

1974, Princeton/Newport Partners. Kelly sizing on every trade. 20% annualized after fees. Two decades. Not one losing quarter

The fund never blew up on its own math. It got shut down chasing Michael Milken's junk bond scandal at Drexel - Thorp was cleared, the firm wasn't spared

The formula never failed him. Someone else's crime did.

[quantfold @quant_fold]

@0x_Ito It’s strange that I didn’t see this before – it’s a gem video, thanks

[quantfold @quant_fold]

@fomoliver Thanks for breaking down the timings in the video.
It’s really handy

[quantfold @quant_fold]

@kryneeex An interesting point of view
That’s worth bearing in mind

[quantfold @quant_fold]

Ed Thorp, 20 years, 20% annual returns after fees, zero losing quarters. He took Kelly's formula out of a physics paper and used it to beat blackjack first, then beat Wall Street

Before Thorp, professional gamblers just felt their way through bets. Casinos assumed the house edge was unbeatable, full stop

Thorp read Kelly's paper, built the math into a card-counting system, wrote Beat the Dealer, and casinos across Nevada started barring him at the door. So he did the same thing to the market instead

1974, Princeton/Newport Partners. Kelly sizing on every trade. 20% annualized after fees. Two decades. Not one losing quarter

The fund never blew up on its own math. It got shut down chasing Michael Milken's junk bond scandal at Drexel - Thorp was cleared, the firm wasn't spared

The formula never failed him. Someone else's crime did.

[quantfold @quant_fold]

John Kelly Jr., Bell Labs physicist one equation, 1956 built more billionaires than Harvard

He wasn't solving gambling
He was fixing noisy phone lines at Bell Labs
One paper, not about money, about signal versus noise
One line inside it became the most copied formula in trading history

f* = edge / odds

Ed Thorp used it to beat blackjack, then Wall Street
Claude Shannon ran it on the stock market in his spare time
Every fund that's survived three decades sizes its bets by some version of it
Kelly never saw any of it
March 18, 1965, a stroke on a Manhattan sidewalk, six packs a day, he was 41

The paper still doesn't mention gambling once

[quantfold @quant_fold]

Harry Markopolos, a quant analyst, spent 9 years warning the SEC about the largest fraud in history.
29 red flags. 21 pages. They ignored him five times

He wasn't guessing 1999, Markopolos was asked to replicate Madoff's returns for his own firm. He couldn't. Nobody could
Not because the strategy was too good, because the math didn't work

He ran the numbers and saw only two explanations Madoff was either the greatest trader alive, or running the biggest fraud in history

He picked the second one, and started building the case
By 2005 he had 29 separate red flags, mathematical proof the returns were impossible, laid out in a 21-page memo titled "The World's Largest Hedge Fund is a Fraud"

He sent it to the SEC. Then sent it again. Five times, over nine years. He was so certain Madoff's money touched the Russian mob that he checked under his car for bombs and carried a gun

The SEC opened a file, made a few calls, closed it.
Every time. December 2008, the fraud collapsed on its own, exactly the way Markopolos predicted years earlier

He was right the whole time. Nobody with the power to stop it wanted to hear it

[quantfold @quant_fold]

@stat1x_x Thank you, I’ll look into this

[quantfold @quant_fold]

Harry Markopolos, a quant analyst, spent 9 years warning the SEC about the largest fraud in history.
29 red flags. 21 pages. They ignored him five times

He wasn't guessing 1999, Markopolos was asked to replicate Madoff's returns for his own firm. He couldn't. Nobody could
Not because the strategy was too good, because the math didn't work

He ran the numbers and saw only two explanations Madoff was either the greatest trader alive, or running the biggest fraud in history

He picked the second one, and started building the case
By 2005 he had 29 separate red flags, mathematical proof the returns were impossible, laid out in a 21-page memo titled "The World's Largest Hedge Fund is a Fraud"

He sent it to the SEC. Then sent it again. Five times, over nine years. He was so certain Madoff's money touched the Russian mob that he checked under his car for bombs and carried a gun

The SEC opened a file, made a few calls, closed it.
Every time. December 2008, the fraud collapsed on its own, exactly the way Markopolos predicted years earlier

He was right the whole time. Nobody with the power to stop it wanted to hear it

[Noxie @0xNox1e]

https://t.co/q8ZLVGcPsc

[quantfold @quant_fold]

@A1xen_ yep

[quantfold @quant_fold]

@0x_Ito yes man

[quantfold @quant_fold]

@0x_Ito Youre welcome

[quantfold @quant_fold]

Stanley Druckenmiller. One trading rule, 30 years, never broken. In 2023, he broke it

Paul Tudor Jones asked him what his own obsession with the national debt had cost him
"After the crash of 1987 I was so worried about debt to GDP, it caused me to be short the stock market for the entire decade of the 90s"
Druckenmiller's rule for 30 years: never let my obsession with the debt interfere with my trading, because I felt it would never have market impact

Then he explains what changed his mind

Argentina, from 2020 to 2023, their central bank financed 60% of their deficit. They got 122% inflation The Fed, by the way, financed 60% of ours - just a kind of a fun fact

By 2043 interest expense will be 7% of GDP
That is 144% of all current discretionary spending
So the politicians telling you they're not going to cut entitlements it's just an outright lie

The position he's actually in right now
I bought a massive leverage position in two years
I'm short 30 years. I'm actually long fixed income for the first time since 2020

[Сarm1ne @carm1nee]

https://t.co/AEvpm1M5Sd

[quantfold @quant_fold]

@maximumxspace That’s right, they lead to great victories

[quantfold @quant_fold]

@0xCybronix that’s exactly right

[quantfold @quant_fold]

@0xCybronix technology is grazy

[quantfold @quant_fold]

@0x_Ito He’s a cool bloke I could do with learning a thing or two from him

[quantfold @quant_fold]

@Vektrix It’s a shame they don’t teach this sort of thing in schools and universities

[quantfold @quant_fold]

@stat1x_x wow, cool