Загрузка видео...

Не удалось загрузить видео

Возникла проблема при загрузке этого видео. Это может быть связано с временными проблемами сети или видео может быть недоступно.

На главную

a quant at Two Sigma told me something at a bar i can't stop thinking about "retail looks at price. we look at the autocorrelation of price changes - completely different signal" i asked him to explain it like i was 12 he drew on a napkin: if today's... move predicts tomorrow's move - even slightly, 53% of the time - that's an edge that edge, sized with Kelly, compounds into something insane data is free - Bloomberg and the Fed publish all of it. math takes a weekend to learn reason retail loses isn't intelligence. it's that they're reading the wrong representation of the same data "a chart hides serial correlation. a time series shows it naked" went home, ran autocorrelation tests on 3 years of SPY data found 4 patterns - statistically significant, all exploitable on a 5-day window signals aren't perfect, right 58% of the time. Kelly says that's enough data was free the whole time, framework sitting in every stats textbook. nobody pointed retail toward it they kept you staring at candlesshow more

Livsun

1,898 subscribers

149,183 просмотров • 2 месяцев назад •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

a risk analyst at two sigma told me something over drinks i still think about "our models are right 54% of the time. retail models can be right 60% and still blow up" i asked him why "you trade every day. we don't" two sigma sits completely flat roughly 40% of the year - not because signals are missing, because regime filter says math doesn't work in that environment trending regime eats mean reversion alive - ranging regime eats momentum most retail accounts blow up not from bad signals but from running the right signal in the wrong regime framework is called regime detection - published academic work, in every quant textbook since the 1990s 3 inputs to identify it: realized vol vs 30-day mean, autocorrelation of daily returns, volume profile vs 20-day average vol elevated + autocorrelation negative = mean-reversion regime - momentum will destroy you here autocorrelation positive + vol compressing = trending regime - reversion will destroy you here wrong regime turns a 60% signal into a 40% signal - you're paying the house right regime turns a 54% signal into a 65% edge - kelly compounds it into something insane bookmark this before you open another chart Dm me "Quant" for full quant course data is free, formula is public, 150 lines of python to run the whole thing they didn't hide the signals from you they hid the question you were supposed to ask first

Livsun

120,252 просмотров • 1 месяц назад

a quant at a prop firm showed me a 5x5 grid on a napkin said: > this is our entire edge. we don't predict price. we predict which box the market is in and where that box historically leads i didn't understand it for weeks. then it clicked never looked at a chart the same way since grid is called a Markov Chain transition matrix. the math is from 1906, it's in every probability textbook on earth and hedge funds use it because it asks a completely different question than retail traders ever ask retail: will this go up or down quant: what state is this market in, and where does this state typically go every market lives in one of maybe 5-6 states at any given moment tight range, volatility compression, trending with momentum, post-spike reversal, pre-breakout coil not random labels - clusters you identify from actual data using volatility, volume, and momentum readings stacked together once you have the states, you build the matrix: P(state 2 -> state 4) = 73% P(state 4 -> state 1) = 61% P(state 1 -> state 3) = 68% each cell is a historical probability. now when the market is in state 2, you're not guessing you're betting on 73% historical completion. you size it with Kelly. you take the trade when the math says to, not when it feels right i built this on BTC using 2 years of 4-hour data. identified 5 states one i labeled "volatility compression below 20-day mean for 6+ consecutive candles" transitioned to a directional move above 1.8 ATR in 71% of cases average reward/risk on those trades: 5.4 that's not prediction. that's reading a probability table the market keeps filling in for you every single day the part that should bother you: the data to build this is free. the framework is in any quant textbook python to implement it is maybe 200 lines what Renaissance Technologies has that you don't isn't secret data or proprietary signals it's this framework applied to higher-resolution data with more sophisticated state definitions you're not missing information you're asking the wrong question every single time you open a chart

a quant at a prop firm showed me a 5x5 grid on a napkin said: > this is our entire edge. we don't predict price. we predict which box the market is in and where that box historically leads i didn't understand it for weeks. then it clicked never looked at a chart the same way since grid is called a Markov Chain transition matrix. the math is from 1906, it's in every probability textbook on earth and hedge funds use it because it asks a completely different question than retail traders ever ask retail: will this go up or down quant: what state is this market in, and where does this state typically go every market lives in one of maybe 5-6 states at any given moment tight range, volatility compression, trending with momentum, post-spike reversal, pre-breakout coil not random labels - clusters you identify from actual data using volatility, volume, and momentum readings stacked together once you have the states, you build the matrix: P(state 2 -> state 4) = 73% P(state 4 -> state 1) = 61% P(state 1 -> state 3) = 68% each cell is a historical probability. now when the market is in state 2, you're not guessing you're betting on 73% historical completion. you size it with Kelly. you take the trade when the math says to, not when it feels right i built this on BTC using 2 years of 4-hour data. identified 5 states one i labeled "volatility compression below 20-day mean for 6+ consecutive candles" transitioned to a directional move above 1.8 ATR in 71% of cases average reward/risk on those trades: 5.4 that's not prediction. that's reading a probability table the market keeps filling in for you every single day the part that should bother you: the data to build this is free. the framework is in any quant textbook python to implement it is maybe 200 lines what Renaissance Technologies has that you don't isn't secret data or proprietary signals it's this framework applied to higher-resolution data with more sophisticated state definitions you're not missing information you're asking the wrong question every single time you open a chart

Livsun

188,258 просмотров • 2 месяцев назад

100 billion dollars made without predicting a single price direction not a direction call, a macro bet, or a news trade in sight they asked a completely different question than every retail trader alive retail: "will this go up?" quant: "what state is the market in, and where does that state historically lead?" same data. completely different game markets aren't random - they cycle through states: trending, ranging, compressing, reversing each state has a fixed historical probability of transitioning to the next that's the whole edge. not a signal, not an indicator - just conditional probability you don't need to know what fed will do or what earnings will print you need to know which state you're in and what that state leads to 68% of the time framework has been public since 1906 Bookmark it before it gone data is free, implementation is 200 lines of python, paper costs nothing they built 100 billion off math sitting in statistics textbooks your whole life you were staring at candles while they ran probability tables

100 billion dollars made without predicting a single price direction not a direction call, a macro bet, or a news trade in sight they asked a completely different question than every retail trader alive retail: "will this go up?" quant: "what state is the market in, and where does that state historically lead?" same data. completely different game markets aren't random - they cycle through states: trending, ranging, compressing, reversing each state has a fixed historical probability of transitioning to the next that's the whole edge. not a signal, not an indicator - just conditional probability you don't need to know what fed will do or what earnings will print you need to know which state you're in and what that state leads to 68% of the time framework has been public since 1906 Bookmark it before it gone data is free, implementation is 200 lines of python, paper costs nothing they built 100 billion off math sitting in statistics textbooks your whole life you were staring at candles while they ran probability tables

Livsun

13,335 просмотров • 2 месяцев назад

a $40/month server beat a room full of analysts to the same trade by five and a half hours market opens at 9:30. his position was already in at 4am the system is a neural net trained on 11 years of tick data. it flagged the setup before the candle that "confirmed" it had even started forming this is the part retail misunderstands about ML in markets it isn't prediction in the mystical sense. it's pattern classification at a speed and scale human eyes physically cannot match the mechanics: 847,000 labeled historical setups as training data 4,200 data points per second ingested live each new state scored against every pattern the net has ever seen, in milliseconds the model isn't asking "where is price going" it's asking "how closely does the current microstructure match the conditions that preceded a move in my training set" that's a classification problem, and classification is what neural nets do better than anything else output: 3-4 candidate trades a day. he takes the top 2 by confidence score last 90 days: 71% win rate at 2.3 average risk-reward the edge isn't the architecture. the architecture is public pytorch is free, the papers are on arxiv, the network is a few hundred lines the edge is the labeling. what you feed it and how you tag the setups is the entire game retail feeds a model price and time and gets noise a desk feeds it order flow, volatility state, cross-asset context, each example hand-labeled by outcome same network. different training data. that's the whole difference retail watches the news at the open and reacts this system scored every pattern before sunrise and already decided you're not losing because your analysis is wrong you're losing to something that doesn't sleep, doesn't panic, and doesn't second-guess a probability it already computed the dataset was free. the framework was free. the compute was $40 a month the edge was never behind a paywall. it was sitting in a format almost nobody bothered to train on full breakdown in the article below

a $40/month server beat a room full of analysts to the same trade by five and a half hours market opens at 9:30. his position was already in at 4am the system is a neural net trained on 11 years of tick data. it flagged the setup before the candle that "confirmed" it had even started forming this is the part retail misunderstands about ML in markets it isn't prediction in the mystical sense. it's pattern classification at a speed and scale human eyes physically cannot match the mechanics: 847,000 labeled historical setups as training data 4,200 data points per second ingested live each new state scored against every pattern the net has ever seen, in milliseconds the model isn't asking "where is price going" it's asking "how closely does the current microstructure match the conditions that preceded a move in my training set" that's a classification problem, and classification is what neural nets do better than anything else output: 3-4 candidate trades a day. he takes the top 2 by confidence score last 90 days: 71% win rate at 2.3 average risk-reward the edge isn't the architecture. the architecture is public pytorch is free, the papers are on arxiv, the network is a few hundred lines the edge is the labeling. what you feed it and how you tag the setups is the entire game retail feeds a model price and time and gets noise a desk feeds it order flow, volatility state, cross-asset context, each example hand-labeled by outcome same network. different training data. that's the whole difference retail watches the news at the open and reacts this system scored every pattern before sunrise and already decided you're not losing because your analysis is wrong you're losing to something that doesn't sleep, doesn't panic, and doesn't second-guess a probability it already computed the dataset was free. the framework was free. the compute was $40 a month the edge was never behind a paywall. it was sitting in a format almost nobody bothered to train on full breakdown in the article below

delost

20,790 просмотров • 26 дней назад

a trader in Shanghai has been running 71% annual returns since 2019 without ever touching a Western exchange nobody outside Weibo knows his name he doesn't manage outside capital, never went on a podcast, never posted a P&L screenshot pause at 0:34 - look at the monitor behind him on the right that's not a price chart. that's a 6x6 state transition matrix built from 11 years of Chinese A-share data he found something in 2018 that every quant textbook describes but almost nobody applies CSI 300 price-state transitions are predictable at a level that makes S&P pattern noise look clean by comparison he mapped 6 market states: trending-up, trending-down, range-tight, range-wide, vol-compression, spike-decay then calculated every historical transition probability across 11 years of 30-minute bars: trending-up -> stays trending: 63% vol-compression -> spike-decay: 78% range-tight -> breaks directional: 71% now he's not predicting direction. he's entering when math says 71% historical completion, sizing with Kelly, closing in under 30 minutes 28 min avg hold, worst month -3.1%, best year +94% framework is markov's from 1906. A-share data is free on WIND Terminal implementation: roughly 180 lines of python insight was never about math - it was about where to aim it Chinese A-shares have thinner institutional algo penetration than US equities. patterns don't get arbed out as fast statistical edges persist for months longer than they would on SPY retail in Shanghai trades on gut. US quants are chasing S&P microstructure nobody was running transition matrices on Chinese state sequences at any real scale he aimed a 119-year-old framework at a market nobody was watching and held the edge for 6 years bookmark this before it becomes obvious math is free, data costs nothing what took time was realizing the most exploitable market wasn't the most-watched one they kept you watching SPY candles while the cleanest probability table on earth sat untouched in Shanghai

a trader in Shanghai has been running 71% annual returns since 2019 without ever touching a Western exchange nobody outside Weibo knows his name he doesn't manage outside capital, never went on a podcast, never posted a P&L screenshot pause at 0:34 - look at the monitor behind him on the right that's not a price chart. that's a 6x6 state transition matrix built from 11 years of Chinese A-share data he found something in 2018 that every quant textbook describes but almost nobody applies CSI 300 price-state transitions are predictable at a level that makes S&P pattern noise look clean by comparison he mapped 6 market states: trending-up, trending-down, range-tight, range-wide, vol-compression, spike-decay then calculated every historical transition probability across 11 years of 30-minute bars: trending-up -> stays trending: 63% vol-compression -> spike-decay: 78% range-tight -> breaks directional: 71% now he's not predicting direction. he's entering when math says 71% historical completion, sizing with Kelly, closing in under 30 minutes 28 min avg hold, worst month -3.1%, best year +94% framework is markov's from 1906. A-share data is free on WIND Terminal implementation: roughly 180 lines of python insight was never about math - it was about where to aim it Chinese A-shares have thinner institutional algo penetration than US equities. patterns don't get arbed out as fast statistical edges persist for months longer than they would on SPY retail in Shanghai trades on gut. US quants are chasing S&P microstructure nobody was running transition matrices on Chinese state sequences at any real scale he aimed a 119-year-old framework at a market nobody was watching and held the edge for 6 years bookmark this before it becomes obvious math is free, data costs nothing what took time was realizing the most exploitable market wasn't the most-watched one they kept you watching SPY candles while the cleanest probability table on earth sat untouched in Shanghai

Livsun

25,048 просмотров • 1 месяц назад

Alan Oppenheim, MIT professor: "Jump Trading pays quants $650K to pull one real signal out of market noise. I wrote the textbook on that exact math and taught it for free." this free MIT lecture is the math the whole "quant edge" is built on, and almost nobody selling you a course will name it. the man teaching it wrote the signal-processing textbook every engineer on earth learned from, and he put the lectures online for nothing. here is the idea. a price chart is a signal buried in noise. most of what you see is random static, and hidden inside is a faint, repeating pattern with real mathematics for pulling it out. that is not a metaphor. jim simons was a codebreaker who did exactly this to enemy transmissions, then pointed the same math at the market and made a hundred billion dollars. none of it is new. oppenheim taught this in 1987, the math is older, and every lecture is free. same story i told in the article above. the edge was never a secret formula. it was knowing a market is a noisy signal, and having the tools to read it. here is what the gurus skip. the math to separate signal from noise is free and everyone has it. the edge is cleaner data than the next desk, and the discipline to trust a faint signal across thousands of trades instead of chasing the static. the filter is free. the patience to use it is not.

Alan Oppenheim, MIT professor: "Jump Trading pays quants $650K to pull one real signal out of market noise. I wrote the textbook on that exact math and taught it for free." this free MIT lecture is the math the whole "quant edge" is built on, and almost nobody selling you a course will name it. the man teaching it wrote the signal-processing textbook every engineer on earth learned from, and he put the lectures online for nothing. here is the idea. a price chart is a signal buried in noise. most of what you see is random static, and hidden inside is a faint, repeating pattern with real mathematics for pulling it out. that is not a metaphor. jim simons was a codebreaker who did exactly this to enemy transmissions, then pointed the same math at the market and made a hundred billion dollars. none of it is new. oppenheim taught this in 1987, the math is older, and every lecture is free. same story i told in the article above. the edge was never a secret formula. it was knowing a market is a noisy signal, and having the tools to read it. here is what the gurus skip. the math to separate signal from noise is free and everyone has it. the edge is cleaner data than the next desk, and the discipline to trust a faint signal across thousands of trades instead of chasing the static. the filter is free. the patience to use it is not.

Rossst.03

269,381 просмотров • 1 месяц назад

a russian mathematician solved trading in 1906 wall street found the paper 70 years later, built billion-dollar funds around it, never mentioned it to retail andrei markov proved markets aren't random - they're state machines trending, ranging, reversing - each state has a fixed historical probability of shifting to the next build a transition matrix from real data: trending -> stays trending: 68% trending -> flips to range: 21% trending -> reverses: 11% now you're not predicting direction. you're entering on 68% historical completion identify your state, size with kelly, take the trade when math says yes that's the edge. the whole edge renaissance has run this since 1988. 37 years of 66% annual returns paper is free, data is free, implementation is 200 lines of python Bookmark and use in your own strategy they kept you staring at candles while they ran probability tables

a russian mathematician solved trading in 1906 wall street found the paper 70 years later, built billion-dollar funds around it, never mentioned it to retail andrei markov proved markets aren't random - they're state machines trending, ranging, reversing - each state has a fixed historical probability of shifting to the next build a transition matrix from real data: trending -> stays trending: 68% trending -> flips to range: 21% trending -> reverses: 11% now you're not predicting direction. you're entering on 68% historical completion identify your state, size with kelly, take the trade when math says yes that's the edge. the whole edge renaissance has run this since 1988. 37 years of 66% annual returns paper is free, data is free, implementation is 200 lines of python Bookmark and use in your own strategy they kept you staring at candles while they ran probability tables

Livsun

376,556 просмотров • 2 месяцев назад

Quant on Polymarket turned $313 into $296,000 in one month - and it never predicted Bitcoin's direction once. 65% win rate on 15-minute Up/Down markets. The edge wasn't the forecast. It was the order book. Wallet proof: It's the one thing retail never sees and bots live on: full bid/ask depth on any BTC/ETH/SOL Up/Down market, at any second in its history, down to 300ms - Binance depth beside it, replayable tick-by-tick on real fills. I've been pulling it from PolyHistorical Here's why it matters. Bloomberg ran the numbers: the top 1% of wallets, mostly bots, took 80%+ of all profits. Retail lost $131M net. Same predictions - better data. What that data shows: pull the depth on a BTC 5-minute market and watch the last minute. Resting bids at T-60 sit around $1,210. At T-10, $240. At T-3, $90. 95% of the liquidity you think you can hit vanishes before resolution. A $500 buy at T-30 eats ~4.2¢ of slippage - half the edge, gone. One tested pattern: when depth leans 65/35 while price sits flat, it reverts within 60s 73% of the time. Median lag from depth to price: 11 seconds. 312 signals in 30 days. The depth wasn't lagging. It was waiting. That's the whole game here - not calling the candle, but seeing depth shift before price does. This wallet did exactly that. Every level, every 300ms, 13,000+ resolved markets, free to replay.

Quant on Polymarket turned $313 into $296,000 in one month - and it never predicted Bitcoin's direction once. 65% win rate on 15-minute Up/Down markets. The edge wasn't the forecast. It was the order book. Wallet proof: It's the one thing retail never sees and bots live on: full bid/ask depth on any BTC/ETH/SOL Up/Down market, at any second in its history, down to 300ms - Binance depth beside it, replayable tick-by-tick on real fills. I've been pulling it from PolyHistorical Here's why it matters. Bloomberg ran the numbers: the top 1% of wallets, mostly bots, took 80%+ of all profits. Retail lost $131M net. Same predictions - better data. What that data shows: pull the depth on a BTC 5-minute market and watch the last minute. Resting bids at T-60 sit around $1,210. At T-10, $240. At T-3, $90. 95% of the liquidity you think you can hit vanishes before resolution. A $500 buy at T-30 eats ~4.2¢ of slippage - half the edge, gone. One tested pattern: when depth leans 65/35 while price sits flat, it reverts within 60s 73% of the time. Median lag from depth to price: 11 seconds. 312 signals in 30 days. The depth wasn't lagging. It was waiting. That's the whole game here - not calling the candle, but seeing depth shift before price does. This wallet did exactly that. Every level, every 300ms, 13,000+ resolved markets, free to replay.

cvxv666

22,348 просмотров • 2 месяцев назад

hedge funds spent $2M to build what you could replicate in a weekend they just knew it was possible first - that was their only advantage a researcher at d.e. shaw handed me a paper in 2023 two years of quant labor, internal "discovery" - original source was an MIT lecture from 1998, public, free, searchable mechanism is cointegration two assets drift independently, but their spread behaves like a rubber band - stationary, mean-reverting, measurable retail asks: will this go up quant asks: is this spread 2.3 standard deviations outside its 5-year mean - and what's historical reversion rate from here completely different question. second one has an answer run it on GLD/GDX, SPY/QQQ, XLE/XOM - data goes back 20 years, free on yahoo finance spread reverts within 8 trading days 71% of the time when it crosses 2 sigma size with Kelly on that 71%, take the trade when math clears, exit at mean paper they spent two years on is engle & granger, 1987 - won a nobel prize Bookmark before it deleted it's in every econometrics textbook on earth, chapter 4 they didn't hide the math from you they just never told you it existed

hedge funds spent $2M to build what you could replicate in a weekend they just knew it was possible first - that was their only advantage a researcher at d.e. shaw handed me a paper in 2023 two years of quant labor, internal "discovery" - original source was an MIT lecture from 1998, public, free, searchable mechanism is cointegration two assets drift independently, but their spread behaves like a rubber band - stationary, mean-reverting, measurable retail asks: will this go up quant asks: is this spread 2.3 standard deviations outside its 5-year mean - and what's historical reversion rate from here completely different question. second one has an answer run it on GLD/GDX, SPY/QQQ, XLE/XOM - data goes back 20 years, free on yahoo finance spread reverts within 8 trading days 71% of the time when it crosses 2 sigma size with Kelly on that 71%, take the trade when math clears, exit at mean paper they spent two years on is engle & granger, 1987 - won a nobel prize Bookmark before it deleted it's in every econometrics textbook on earth, chapter 4 they didn't hide the math from you they just never told you it existed

Livsun

26,816 просмотров • 2 месяцев назад

The biggest marketplace on Earth won't be the assets humans own. It'll be the data AI needs. Everything that can be digitised is getting pulled in to train the next models. Every scientific dataset, every piece of research sitting in a university, every sensor reading off every John Deere tractor in a field. All of it feeds the machine, and almost none of it is visible to us. What comes out is an agentic economy running at machine speed. You plug your data into a vault, and agents monetise it on your behalf, trading with other agents in real time, in a marketplace we never see. The company sitting on proprietary data becomes part of it without even knowing. This is what the Economic Singularity actually looks like from the inside. The real economy stops being the one we can see and price, and moves somewhere we can't follow. You own a piece of it, or you get priced out of it. Jordi Visser.

The biggest marketplace on Earth won't be the assets humans own. It'll be the data AI needs. Everything that can be digitised is getting pulled in to train the next models. Every scientific dataset, every piece of research sitting in a university, every sensor reading off every John Deere tractor in a field. All of it feeds the machine, and almost none of it is visible to us. What comes out is an agentic economy running at machine speed. You plug your data into a vault, and agents monetise it on your behalf, trading with other agents in real time, in a marketplace we never see. The company sitting on proprietary data becomes part of it without even knowing. This is what the Economic Singularity actually looks like from the inside. The real economy stops being the one we can see and price, and moves somewhere we can't follow. You own a piece of it, or you get priced out of it. Jordi Visser.

Raoul Pal

78,736 просмотров • 19 дней назад

quants don't predict markets. they stopped trying years ago retail is still asking "where is price going?" - quants switched to a different question entirely "given the current state, what's the probability distribution of outcomes" that's not a semantics game. it changes the math entirely stochastic processes model how a system evolves probabilistically - not picking a direction, mapping where each path leads and how likely it is ornstein-uhlenbeck does this for mean reversion: drift rate back to equilibrium, variance around that path same price data everybody has. different equation math is a century old, free in every applied probability textbook Bookmark and learn this before it gone you've been drawing trend lines while they ran probability distributions on same numbers it was never about data it was the question you were taught to ask

quants don't predict markets. they stopped trying years ago retail is still asking "where is price going?" - quants switched to a different question entirely "given the current state, what's the probability distribution of outcomes" that's not a semantics game. it changes the math entirely stochastic processes model how a system evolves probabilistically - not picking a direction, mapping where each path leads and how likely it is ornstein-uhlenbeck does this for mean reversion: drift rate back to equilibrium, variance around that path same price data everybody has. different equation math is a century old, free in every applied probability textbook Bookmark and learn this before it gone you've been drawing trend lines while they ran probability distributions on same numbers it was never about data it was the question you were taught to ask

Livsun

13,315 просмотров • 2 месяцев назад

🚨 Jensen Huang says everyone panicked about the AI data when MOST training data was never REAL to begin with. Ilya Sutskever told the industry pre-training was over. "Ilya said, 'We're out of data,' or something like that. 'Pre-training is over,' or something like that," Huang says. "The industry panicked, you know, that this is the end of AI." "And of course, of course that's obviously not true. We're gonna keep on scaling the amount of data that we have to train with." "A lot of that data is probably gonna be synthetic." That's where the panic came from — synthetic data sounds like cheating. "Most of the data that we are training, that we teach each other with, inform each other with, is synthetic." "It's synthetic because it didn't come out of nature." "You created it. I'm consuming it. I modify it, augment it, I regenerate it, somebody else consumes it." The textbook in your hand is synthetic. The post you're reading is synthetic. The lecture you took is synthetic. Nature didn't make any of it. Humans did. AI just learned to do the same thing — faster. "Training is now limited by compute," Huang says. "Data is now limited by compute." The data wall wasn't a wall. It was a mirror. If you're new here, follow @AiEvolutio for the latest on ChatGPT, Claude, and the AI tools shaping how we work and create. — Jensen Huang ( NVIDIA ), NVIDIA CEO, on Lex Fridman's ( Lex Fridman ) podcast

🚨 Jensen Huang says everyone panicked about the AI data when MOST training data was never REAL to begin with. Ilya Sutskever told the industry pre-training was over. "Ilya said, 'We're out of data,' or something like that. 'Pre-training is over,' or something like that," Huang says. "The industry panicked, you know, that this is the end of AI." "And of course, of course that's obviously not true. We're gonna keep on scaling the amount of data that we have to train with." "A lot of that data is probably gonna be synthetic." That's where the panic came from — synthetic data sounds like cheating. "Most of the data that we are training, that we teach each other with, inform each other with, is synthetic." "It's synthetic because it didn't come out of nature." "You created it. I'm consuming it. I modify it, augment it, I regenerate it, somebody else consumes it." The textbook in your hand is synthetic. The post you're reading is synthetic. The lecture you took is synthetic. Nature didn't make any of it. Humans did. AI just learned to do the same thing — faster. "Training is now limited by compute," Huang says. "Data is now limited by compute." The data wall wasn't a wall. It was a mirror. If you're new here, follow @AiEvolutio for the latest on ChatGPT, Claude, and the AI tools shaping how we work and create. — Jensen Huang ( NVIDIA ), NVIDIA CEO, on Lex Fridman's ( Lex Fridman ) podcast

AI Evolution

15,565 просмотров • 2 месяцев назад

Two Sigma runs 500 stocks through one formula and looks at 5 numbers math has been public since 1901 PCA - principal component analysis - run it on the S&P and entire market compresses 87% of all price movement. 5 hidden variables not sectors, not earnings - 5 eigenvectors. the hidden skeleton underneath price > market pressure - a tide moving every name in sync > growth vs value rotation > size divergence, mega cap vs small cap > momentum regime - trending or mean reverting > volatility clustering when eigenvector 4 flips, every momentum strategy on earth breaks at the same time same force rotating under all 500 names at once - not coincidence, that's structure retail reads 500 charts. quant desks read 5 transitions data: free. code: 300 lines of python. framework: 1901 been sitting in every linear algebra textbook for 120 years they just never pointed you at the right chapter

Two Sigma runs 500 stocks through one formula and looks at 5 numbers math has been public since 1901 PCA - principal component analysis - run it on the S&P and entire market compresses 87% of all price movement. 5 hidden variables not sectors, not earnings - 5 eigenvectors. the hidden skeleton underneath price > market pressure - a tide moving every name in sync > growth vs value rotation > size divergence, mega cap vs small cap > momentum regime - trending or mean reverting > volatility clustering when eigenvector 4 flips, every momentum strategy on earth breaks at the same time same force rotating under all 500 names at once - not coincidence, that's structure retail reads 500 charts. quant desks read 5 transitions data: free. code: 300 lines of python. framework: 1901 been sitting in every linear algebra textbook for 120 years they just never pointed you at the right chapter

Livsun

13,957 просмотров • 1 месяц назад

a researcher at a systematic fund showed me a 6x6 matrix on a whiteboard said: > this is our whole system. we don't forecast price. we classify which cell the market occupies and where that cell historically transitions i stared at it for days before it made sense. then everything changed never interpreted a chart the same way again the matrix is called a Markov Chain transition table. the theory dates back to 1913, it's in every introductory statistics textbook ever written and systematic funds rely on it because it poses a fundamentally different question than what retail traders ever think to ask retail: is this going higher or lower systematic: what regime is this market in, and where does this regime historically resolve every market sits in one of roughly 4-6 regimes at any point in time narrow consolidation, expanding volatility, trending with acceleration, post-reversal drift, pre-expansion squeeze not arbitrary categories - clusters you extract from real data using volatility, volume, and trend strength layered together once you define the regimes, you construct the table: P(regime 3 → regime 5) = 71% P(regime 1 → regime 3) = 64% P(regime 2 → regime 4) = 69% each cell is a historical frequency. now when the market sits in regime 3, you're not speculating you're trading on 71% historical resolution. you scale it with Kelly. you execute when the math confirms, not when it feels convincing i constructed this on ETH using 3 years of 1-hour data. isolated 6 regimes one i named "volatility squeeze below 15-day average for 8+ consecutive bars" transitioned to a directional breakout exceeding 2.2 ATR in 74% of cases average reward/risk on those setups: 4.8 that's not forecasting. that's reading a probability matrix the market populates for you every single session the part that should concern you: the data to construct this is free. the methodology is in any quantitative finance textbook python to build it is maybe 180 lines what Citadel has that you don't isn't classified data or secret algorithms it's this methodology applied to tick-level data with more granular regime definitions you're not lacking information you're framing the wrong question every single time you open a chart

a researcher at a systematic fund showed me a 6x6 matrix on a whiteboard said: > this is our whole system. we don't forecast price. we classify which cell the market occupies and where that cell historically transitions i stared at it for days before it made sense. then everything changed never interpreted a chart the same way again the matrix is called a Markov Chain transition table. the theory dates back to 1913, it's in every introductory statistics textbook ever written and systematic funds rely on it because it poses a fundamentally different question than what retail traders ever think to ask retail: is this going higher or lower systematic: what regime is this market in, and where does this regime historically resolve every market sits in one of roughly 4-6 regimes at any point in time narrow consolidation, expanding volatility, trending with acceleration, post-reversal drift, pre-expansion squeeze not arbitrary categories - clusters you extract from real data using volatility, volume, and trend strength layered together once you define the regimes, you construct the table: P(regime 3 → regime 5) = 71% P(regime 1 → regime 3) = 64% P(regime 2 → regime 4) = 69% each cell is a historical frequency. now when the market sits in regime 3, you're not speculating you're trading on 71% historical resolution. you scale it with Kelly. you execute when the math confirms, not when it feels convincing i constructed this on ETH using 3 years of 1-hour data. isolated 6 regimes one i named "volatility squeeze below 15-day average for 8+ consecutive bars" transitioned to a directional breakout exceeding 2.2 ATR in 74% of cases average reward/risk on those setups: 4.8 that's not forecasting. that's reading a probability matrix the market populates for you every single session the part that should concern you: the data to construct this is free. the methodology is in any quantitative finance textbook python to build it is maybe 180 lines what Citadel has that you don't isn't classified data or secret algorithms it's this methodology applied to tick-level data with more granular regime definitions you're not lacking information you're framing the wrong question every single time you open a chart

Hrundel75 🐷

12,088 просмотров • 26 дней назад

Congrats if you watched this video a few months ago, I think we were one of the first channels to spot this diminishing MVRV pattern on $BTC. The technicals and on chain data all indicated that we were starting to form a top. Price of #Bitcoin at the time was $111,000.

Congrats if you watched this video a few months ago, I think we were one of the first channels to spot this diminishing MVRV pattern on $BTC. The technicals and on chain data all indicated that we were starting to form a top. Price of #Bitcoin at the time was $111,000.

SmartMoneyCrypto

61,149 просмотров • 8 месяцев назад

quant at D.E. Shaw told me his best strategy loses 6 out of every 10 trades showed me the ledger. didn't believe him until i saw the equity curve retail optimizes for win rate. quants optimize for expected value - completely different targets that gap is why hedge funds extract money from retail every single day math is literally 4th grade: EV = (win% x avg win) - (loss% x avg loss) a 39% win rate at 3:1 reward/risk beats an 80% win rate at 1:2, every single time retail either doesn't know this, or ignores it because losing 61% of trades feels wrong so they cut winners early to "lock in gains" and hold losers because "it'll come back" both moves destroy expected value. both are emotional, not mathematical citadel calls systematic position sizing its edge. retail calls it going with your gut Kelly Criterion fixes it - bet (EV / odds) of your capital per trade, sized from your own historical data ran 2 years of backtests on a 40% win-rate strategy: > Kelly-sized positions: +340% > same signals, gut-sized: +71% same market, same entries and exits only variable: math decided how much to bet Bookmark this before you open your next position they kept you counting candles while they ran expected value tables Reply below what you think

quant at D.E. Shaw told me his best strategy loses 6 out of every 10 trades showed me the ledger. didn't believe him until i saw the equity curve retail optimizes for win rate. quants optimize for expected value - completely different targets that gap is why hedge funds extract money from retail every single day math is literally 4th grade: EV = (win% x avg win) - (loss% x avg loss) a 39% win rate at 3:1 reward/risk beats an 80% win rate at 1:2, every single time retail either doesn't know this, or ignores it because losing 61% of trades feels wrong so they cut winners early to "lock in gains" and hold losers because "it'll come back" both moves destroy expected value. both are emotional, not mathematical citadel calls systematic position sizing its edge. retail calls it going with your gut Kelly Criterion fixes it - bet (EV / odds) of your capital per trade, sized from your own historical data ran 2 years of backtests on a 40% win-rate strategy: > Kelly-sized positions: +340% > same signals, gut-sized: +71% same market, same entries and exits only variable: math decided how much to bet Bookmark this before you open your next position they kept you counting candles while they ran expected value tables Reply below what you think

Livsun

14,773 просмотров • 1 месяц назад

a hotel front desk clerk in nashville figured out why markets move exactly when they do not direction, not news - the actual mechanism of why a move happens at all he works overnight shift, 11pm to 7am. lobby goes quiet after midnight, nothing but a monitor and a wifi connection question that started it: why does volatility cluster he'd read it in passing - options dealers cause price moves they didn't intend spent 6 hours across two nights searching, wrote everything into a google doc called "options thing" here's what he found when you buy a call option from a dealer, dealer has a new problem. they sold you the right to buy shares at a certain price if stock moves up, your option gains value and dealer owes you money. to protect themselves they have to buy shares immediately - no discretion, no delay amount they have to buy at every price level is published every second for free - it's open interest on the options chain. every brokerage shows it he built a spreadsheet every morning at 9:29am, one minute before open, he pulled SPY's options chain and calculated where dealers were most exposed marked strikes with heaviest call open interest. watched what happened in first 30 minutes of trading day 12 he stopped breathing for a second price moved to the strike with heaviest dealer exposure 73% of the time in the first 45 minutes not because of a chart pattern, not because of any signal because 400 dealers ran the same hedge calculation at open, and all of them had to buy the same shares at the same time he started calling it gravity price pulls toward certain strikes when dealer positioning is heavy enough - not prediction, mechanics math has a name: gamma exposure, or GEX SpotGamma built a whole company surfacing it. Squeeze Metrics published an academic paper on mechanics in 2018 python implementation is around 400 lines, nothing but the options chain you already have he built it in google colab over 3 weekends, free, working only on nights the lobby was empty tracked it against 60 days of live SPY data on negative GEX days - dealers short gamma, forced to amplify moves - average daily range expanded 2.8x on positive GEX days, 63% of sessions closed within half a percent of open this is not a signal. it's a regime classifier negative GEX: something moves big today, whichever direction gets started. buy straddles, size up, let dealers carry it positive GEX: nothing moves today. dealers kill every attempt before it gets 2 points sell premium, collect theta, sleep at month 4 he went live. $4,200 account, pure options, no directional bet six months later: $4,200 became $19,800 he still works overnight shift. told me about it in the lobby at 3am when i asked what he was typing google doc still says "options thing" - he never renamed it i asked why he never shared this. he looked at the lobby doors and said "who would believe a hotel clerk" data is free, formula is public, wall street has run this since 2017 they assumed retail would never think to read options flow as a mechanical map of where price has to go they were right about retail. they weren't right about him bookmark this and go build it market tells you exactly where it's going. you just have to stop reading the wrong layer Write your thought below

a hotel front desk clerk in nashville figured out why markets move exactly when they do not direction, not news - the actual mechanism of why a move happens at all he works overnight shift, 11pm to 7am. lobby goes quiet after midnight, nothing but a monitor and a wifi connection question that started it: why does volatility cluster he'd read it in passing - options dealers cause price moves they didn't intend spent 6 hours across two nights searching, wrote everything into a google doc called "options thing" here's what he found when you buy a call option from a dealer, dealer has a new problem. they sold you the right to buy shares at a certain price if stock moves up, your option gains value and dealer owes you money. to protect themselves they have to buy shares immediately - no discretion, no delay amount they have to buy at every price level is published every second for free - it's open interest on the options chain. every brokerage shows it he built a spreadsheet every morning at 9:29am, one minute before open, he pulled SPY's options chain and calculated where dealers were most exposed marked strikes with heaviest call open interest. watched what happened in first 30 minutes of trading day 12 he stopped breathing for a second price moved to the strike with heaviest dealer exposure 73% of the time in the first 45 minutes not because of a chart pattern, not because of any signal because 400 dealers ran the same hedge calculation at open, and all of them had to buy the same shares at the same time he started calling it gravity price pulls toward certain strikes when dealer positioning is heavy enough - not prediction, mechanics math has a name: gamma exposure, or GEX SpotGamma built a whole company surfacing it. Squeeze Metrics published an academic paper on mechanics in 2018 python implementation is around 400 lines, nothing but the options chain you already have he built it in google colab over 3 weekends, free, working only on nights the lobby was empty tracked it against 60 days of live SPY data on negative GEX days - dealers short gamma, forced to amplify moves - average daily range expanded 2.8x on positive GEX days, 63% of sessions closed within half a percent of open this is not a signal. it's a regime classifier negative GEX: something moves big today, whichever direction gets started. buy straddles, size up, let dealers carry it positive GEX: nothing moves today. dealers kill every attempt before it gets 2 points sell premium, collect theta, sleep at month 4 he went live. $4,200 account, pure options, no directional bet six months later: $4,200 became $19,800 he still works overnight shift. told me about it in the lobby at 3am when i asked what he was typing google doc still says "options thing" - he never renamed it i asked why he never shared this. he looked at the lobby doors and said "who would believe a hotel clerk" data is free, formula is public, wall street has run this since 2017 they assumed retail would never think to read options flow as a mechanical map of where price has to go they were right about retail. they weren't right about him bookmark this and go build it market tells you exactly where it's going. you just have to stop reading the wrong layer Write your thought below

Livsun

200,378 просмотров • 1 месяц назад

quant at aqr sat next to me at a conference and said something that rewired how i think about markets "we don't analyze 500 stocks. we run 5 factors that explain all 500 at once" i asked him to draw it out take your entire stock universe - 500 names, 3 years of daily returns - and run PCA the 500 collapse into 5 factor 1: broad market momentum - explains 58% of all price movement by itself factor 2: sector rotation - where institutional money is actually flowing factor 3: rate sensitivity - duration risk bleeding across the equity curve factor 4: momentum tilt - which names are accelerating vs mean-reverting factor 5: credit spread stress - signal hiding inside equity vol if you know these 5 numbers, you know what 500 stocks are about to do retail watches 500 charts and sees noise. quant desks watch 5 eigenvalues and see structure math is called Principal Component Analysis, in every intro stats course on earth data to run it is free. python implementation is ~120 lines renaissance, two sigma, aqr - running this since before most retail traders could name a stock reason nobody told you isn't that it's proprietary it's that telling you costs them their edge Bookmark this market was never 500 moving parts it was always 5 - hiding inside 500

quant at aqr sat next to me at a conference and said something that rewired how i think about markets "we don't analyze 500 stocks. we run 5 factors that explain all 500 at once" i asked him to draw it out take your entire stock universe - 500 names, 3 years of daily returns - and run PCA the 500 collapse into 5 factor 1: broad market momentum - explains 58% of all price movement by itself factor 2: sector rotation - where institutional money is actually flowing factor 3: rate sensitivity - duration risk bleeding across the equity curve factor 4: momentum tilt - which names are accelerating vs mean-reverting factor 5: credit spread stress - signal hiding inside equity vol if you know these 5 numbers, you know what 500 stocks are about to do retail watches 500 charts and sees noise. quant desks watch 5 eigenvalues and see structure math is called Principal Component Analysis, in every intro stats course on earth data to run it is free. python implementation is ~120 lines renaissance, two sigma, aqr - running this since before most retail traders could name a stock reason nobody told you isn't that it's proprietary it's that telling you costs them their edge Bookmark this market was never 500 moving parts it was always 5 - hiding inside 500

Livsun

22,073 просмотров • 1 месяц назад

7 years of trading options. Lost over $30,000. Blown accounts I can't even count. There was a point I seriously considered quitting forever. I was making less than $ 30k at my 9-5. Struggling financially. Struggling mentally. Trading was supposed to be the way out. Instead it became the thing that was destroying me. I remember one point, it was 2am. I was staring at my brokerage screen after blowing $500 on a revenge trade. Kept refreshing hoping the number would change. It never did. Too embarrassed to tell anyone. Dodging family questions. Starting to wonder if maybe I just wasn't smart enough. Maybe these traders on social media had something I didn't. I tried everything. Every YouTube strategy. Every Discord. Spent thousands on courses. Scalping. Swing trading. Signals. Indicators. All of it. Every time - further in the hole. Then one January morning. Down $4,000 year to date. Sat on the edge of my bed, head in my hands. Told myself I was done. And then I realized something. I had been treating trading like gambling. Forcing trades. Oversizing. Chasing. Revenge trading. I scrapped everything. Started over with tiny positions and a checklist . One rule: if a trade doesn't meet every condition, I don't take it. Doesn't matter how good it looks. The edge was being boringly selective, not finding a better setup. That's when everything changed. 7 years later. I'm consistently profitable. I still have a full-time job (a better one now). The centerpiece of that checklist is something most retail traders have never heard of. It's called GEX - Gamma Exposure. Every time you buy an options contract, someone has to be on the other side of that trade. That someone is almost always a market maker - firms like Citadel, SIG, Wolverine. They don't want directional risk. They just want to collect the spread. So the moment they sell you that contract, they immediately go into the market and hedge their position by buying or selling the underlying shares. That hedging is not optional. They have to do it. Continuously. All day long. As price moves. At certain price levels there is a massive concentration of forced buying or forced selling from those dealers. Price reacts to those levels because dealers are mechanically obligated to transact there. That's a GEX level. Not a line you drew on a chart. Not a pattern you spotted. A level backed by real institutional positioning data. I pull the GEX map every single morning using ITMatrix HQ before I open a single chart. It shows me exactly where the major levels are, what kind of environment I'm trading in, and where the clean air is between levels. So instead of guessing where price might react - I know where it has to. Because it almost always behaves the exact same way! But GEX is only one piece of it. The checklist has four components total. And all four have to be green before I place a single trade. Miss one - I wait. That's what makes it an A+ setup. I used it to turn $1,000 into $21,000, and I consistently collect around 5k every month from trading alone! If all this sounds complicated, watch the video below. The checklist is FREE btw - I recently turned it into a web app find it in the comments below ↓

7 years of trading options. Lost over $30,000. Blown accounts I can't even count. There was a point I seriously considered quitting forever. I was making less than $ 30k at my 9-5. Struggling financially. Struggling mentally. Trading was supposed to be the way out. Instead it became the thing that was destroying me. I remember one point, it was 2am. I was staring at my brokerage screen after blowing $500 on a revenge trade. Kept refreshing hoping the number would change. It never did. Too embarrassed to tell anyone. Dodging family questions. Starting to wonder if maybe I just wasn't smart enough. Maybe these traders on social media had something I didn't. I tried everything. Every YouTube strategy. Every Discord. Spent thousands on courses. Scalping. Swing trading. Signals. Indicators. All of it. Every time - further in the hole. Then one January morning. Down $4,000 year to date. Sat on the edge of my bed, head in my hands. Told myself I was done. And then I realized something. I had been treating trading like gambling. Forcing trades. Oversizing. Chasing. Revenge trading. I scrapped everything. Started over with tiny positions and a checklist . One rule: if a trade doesn't meet every condition, I don't take it. Doesn't matter how good it looks. The edge was being boringly selective, not finding a better setup. That's when everything changed. 7 years later. I'm consistently profitable. I still have a full-time job (a better one now). The centerpiece of that checklist is something most retail traders have never heard of. It's called GEX - Gamma Exposure. Every time you buy an options contract, someone has to be on the other side of that trade. That someone is almost always a market maker - firms like Citadel, SIG, Wolverine. They don't want directional risk. They just want to collect the spread. So the moment they sell you that contract, they immediately go into the market and hedge their position by buying or selling the underlying shares. That hedging is not optional. They have to do it. Continuously. All day long. As price moves. At certain price levels there is a massive concentration of forced buying or forced selling from those dealers. Price reacts to those levels because dealers are mechanically obligated to transact there. That's a GEX level. Not a line you drew on a chart. Not a pattern you spotted. A level backed by real institutional positioning data. I pull the GEX map every single morning using ITMatrix HQ before I open a single chart. It shows me exactly where the major levels are, what kind of environment I'm trading in, and where the clean air is between levels. So instead of guessing where price might react - I know where it has to. Because it almost always behaves the exact same way! But GEX is only one piece of it. The checklist has four components total. And all four have to be green before I place a single trade. Miss one - I wait. That's what makes it an A+ setup. I used it to turn $1,000 into $21,000, and I consistently collect around 5k every month from trading alone! If all this sounds complicated, watch the video below. The checklist is FREE btw - I recently turned it into a web app find it in the comments below ↓

Nick Ireland

25,939 просмотров • 3 месяцев назад

MIT has been teaching this publicly for 50 years. retail traders still draw trendlines andrei markov published the framework in 1906. hedge funds found it in the 70s built billion-dollar operations around it, never told retail it existed markets aren't random - they're state machines trending, ranging, reversing. each state has a fixed historical probability of shifting to the next you build a transition matrix from real price data: trending -> stays trending: 68% trending -> flips to range: 21% trending -> reverses: 11% now you're not predicting direction. you're entering on 68% historical completion that's not a signal, that's conditional probability compounding every single trade math is free, data is free, implementation is ~200 lines of python what's not free: knowing this framework exists while everyone around you draws support levels and hopes renaissance has run this since 1988. paper was always public edge was never secret. just never pointed out

MIT has been teaching this publicly for 50 years. retail traders still draw trendlines andrei markov published the framework in 1906. hedge funds found it in the 70s built billion-dollar operations around it, never told retail it existed markets aren't random - they're state machines trending, ranging, reversing. each state has a fixed historical probability of shifting to the next you build a transition matrix from real price data: trending -> stays trending: 68% trending -> flips to range: 21% trending -> reverses: 11% now you're not predicting direction. you're entering on 68% historical completion that's not a signal, that's conditional probability compounding every single trade math is free, data is free, implementation is ~200 lines of python what's not free: knowing this framework exists while everyone around you draws support levels and hopes renaissance has run this since 1988. paper was always public edge was never secret. just never pointed out

Livsun

13,963 просмотров • 2 месяцев назад