Vlad (Visual Sectors)

Happy Anniversary. 40% bagged + beating Blackrock's BDMIX by a mile😇 69% weekly win rate (that's the trade duration, weekly rebalancing Tuesdays)

@netrocyn @Steven_Swinford Its the reporting. Like pushing "reform won 1400 seats and theh are about to make garage PM" while actually they have less councillors than lib dem and only majority in 3 and GE in 3 years. Clear agenda and media manipulation

@Steven_Swinford So, the bit that matters. Which alternative has more support than Starmer does? None of them. They've moved too soon and plunged their party into chaos

The Labour civil war in numbers: THE ANTI-STARMER BRIGADE **87** Labour MPs have now called for Starmer to go, including 3 ministers and 6 PPSs They hail from all wings of party - around 40% moderates, 20% hard left, 36% soft left We expect number to pass 100 by end of day with more ministers going over top. Ongoing speculation about several members of the cabinet STARMER SUPPORTERS **55** Labour MPs have publicly defended Starmer as the No 10 operation begins to take effect Many of them are paid to do so - 7 Cabinet ministers have backed the prime minister and 8 other ministers, along with 9 PPSs But that figure is still remarkably low - it represents just 16% of those on the government payroll The PM also has secured the backing of 31 backbenchers

@AVMikhailova This reads as "New PM":) I thought I missed something

NEW PM tells Cabinet: As I said yesterday, I take responsibility for these election results and I take responsibility for delivering the change we promised The past 48 hours have been destabilising for government and that has a real economic cost for our country & for families

@teej_m And that will basically keep current costs intact long term. Because at the moment those 2 orders are subsidised by venture capital

The cost of machine intelligence will drop by 2 orders of magnitude or more, it’s insanity to believe otherwise. This is just the beginning.

@tonitrades_ @KanikaBK Came here to say this. Quality of the output is rubbish.

@KanikaBK Free data sounds great until you realize the real edge isn't in the data itself - it's in the cleaning, normalization, and latency. Most of these feeds have gaps or delays that'll eat your alpha before you even notice.

THIS FEELS ILLEGAL AND I LOVE IT. MOST PEOPLE DO NOT KNOW THIS EXISTS. This free GitHub repo lists every library that gives you market data across every major asset class. - Stocks - Crypto - Forex - Economic indicators No Bloomberg. No expensive APIs. Here is everything inside & why every trader & developer should bookmark it right now👇 The Awesome Systematic Trading repo is a curated directory of every library, framework, data source, and research tool a professional quant desk actually uses. This is not beginner content. This is the real infrastructure that hedge funds spend years and millions of dollars assembling. Someone just mapped it all in one place for free. The data libraries listed inside cover everything. yfinance pulls live and historical stock data from Yahoo Finance for free. AkShare covers Chinese markets. Investpy pulls from Investing.com. Quandl gives you millions of economic datasets from hundreds of publishers via a single free API. OpenBB Terminal gives you Bloomberg level research for anyone at zero cost. And FinanceDatabase, one of the libraries referenced inside, covers 300,000 plus symbols across equities, ETFs, funds, indices, currencies, and crypto. ↳ Free data libraries include yfinance, AkShare, Investpy, Quandl, OpenBB, and pandas datareader ↳ Covers stocks, crypto, forex, futures, options, CFDs, and economic indicators ↳ FinanceDatabase referenced inside covers 300,000 plus symbols across every major asset class ↳ Backtesting frameworks available in Python, Rust, C++, Go, and Java ↳ AI powered systems including Microsoft QLib, FinRL, and FinGPT all listed ↳ Live trading support across multiple brokers built into dozens of the listed frameworks ↳ 3700 plus stars from quant professionals globally and actively maintained And it goes deeper than just data. Microsoft built QLib and open sourced it here. The AI4Finance Foundation released FinRL, the first framework applying deep reinforcement learning to live trading. FinGPT put financial large language models on HuggingFace for anyone to fine tune and deploy. The tooling gap between retail traders and institutional desks is gone. The only advantages hedge funds have left are capital at scale and execution speed. Everything else is in this repo. 100% open source. MIT and Creative Commons licensed

@devinmarch @karrisaarinen This psychosis is very similar to early crypto bros. And while the crypto is a trillion industry (as is ai), I am yet to see a reasonable use case outside of illegal stuff. AI is much better in that sense but still wildly exaggerated.

@karrisaarinen i.e. because it's a tool.

A common dynamic I observe with AI: it feels most impressive when you don’t know much about the subject, don’t care or don’t have a clear idea of what the you want. This applies across design, code, legal, and more. If I don’t know code very well, every piece of code it writes feels very impressive. Once you know what something should feel or look like, it becomes almost impossible to guide AI there. And you definitely can’t one-shot it.

@TENNarrates @karrisaarinen @devinmarch Nothing intelligent there pal.

@karrisaarinen @devinmarch It's not some sort of dumb screwdriver. It's called Artificial "Intelligence".

@DeryaTR_ Shortest path to a stroke.

Over the past 2-3 months, I’ve been sleeping an average of 5-6 hours, 2 hours less than usual & have cut all entertainment to almost zero, including on weekends & yet still don’t have nearly enough time to work with AI. I am just scratching the surface of what is possible with AI

@TrueSlazac Well, using a tool, having evidence it works, having no elasticity in regards of price changes 10x and company valuations that form a bubble are 4 absolutely uncorrelated things. And I am not even an anti AI guy. Its just a tool, a good one but with limited application

The main reason anti-AI people online annoy me isn't because of misplaced morals, it's the delusion How can you think a tool used by over half of businesses that is very obviously a productivity boost could be a bubble that will disappear in a few years?

Can LLM-based trading systems increase market volatility? The idea is quite simple: 1. LLMs tend to act as "belief homogenizer" - producing similar outcomes based on similar trusted sources 2. This creates clustering of trades 3. The more crowded clusters become, the higher the resonance Came across several interesting studies that support this idea. 1. SSRN paper “Strategic Homogeneity in Trading: Amplifying Market Volatility and the Emergence of Mini-Crashes” argues that convergence of trading strategies can amplify market volatility 2. "Algorithmic Trading and Liquidity Commonality" paper finds that increased algorithmic trading contributed to stronger liquidity co-movement: 5-minute liquidity co-movement rose by 30–50%, and daily liquidity co-movement was 90% higher. In plain English, automation did not just change individual trades; it made liquidity conditions move together more 3. While the previous 2 explore fundamental processes and are not exactly referring to LLMs, this third one actually supports the common ground. "When ChatGPT Stops Talking: GenAI-induced Retail Herding and Systematic Risk" by Xiaoxue He. The finding is that ChatGPT acts as a "belief homogenizer," synchronizing retail beliefs and trading decisions Basically, the more AI systems are in place, the sharper the swings. And while the systems might not be wrong at all, amplitude may cause risk management systems to break down and trigger response from stop loss positions and market makers. 2 quick options here: 1. Use short-based volatility based position sizing. 6-12 months max 2. Play around with portfolio composition. Market neutral framework has done wonders for us lately

@imPenny2x Resources are scarce. Production might be cheap but resources will stay monopolized (or should I say oligopolozed). That means inflation on resources will make everything else unaffordable even if production is extremely cheap

99% of people really do not understand abundance as Elon describes it. The fundamental reason is that they don’t understand compound growth. Same people who would probably pick 1 million dollars today over a penny that doubles in value every day for 30 days. It’s a bad choice by the way. You lose out on millions. Imagine if that doubling object was a labor producing robot instead of a penny. Compounding labor. It’s actually crazy if you try and wrap your mind around it. So Elon mentions Universl High Income and the midwits flip a lid. “The elites won’t share” You don’t get it. They won’t need to share. They will make everything so cheap, it is effectively free. Charities will have immense resources to distribute. Unfathomable intelligence will exist to help optimize production and distribution. An unfathomably large labor pool will exist that operates on solar power exclusively. The public work projects that are erected will be unseen before levels of breathtaking. I think we are incredibly blessed to steward this new age of abundance. Can you see it now? Can you see the future?

@Swiss4Peace @ApoStructura Jesus christ.. ai bros are even further down the line compared to NFT bros. AI burns resources, it doesnt create them.

@visualsectors @ApoStructura Yes. Oil is no longer an issue when you electrify everything for cheap apart from polymeres. Most other resources are pretty abundant, the bottleneck currently is that it is too expensive to get them out of the ground and (re) mprocess them. That will long term no longer apply.

"If everyone gets material abundance. No one has material abundance." This is a dumb argument masquerading as a clever one. The key here is that AI/robotics enables an explosion of productivity, and UHI is a way to spread it to those who will lose their job.

@Swiss4Peace @ApoStructura Supply of oil? Wood? Aluminium? Read again.

@visualsectors @ApoStructura Why should inflation eat whatever income is paid out. Sure demand will increase. But supply will grow exponentially more. (This is based on actually achieving AGI, not just some AI job losses in some fields.)

@zg_dev @econcallum I am yet to see a compelling reason for crypto to exist. Its only still around because it hasnt reached the level of rug pull that its supposed to

Crypto led to huge ecosystems being built, but few of them infiltrated mainstream workflows (other than stablecoins). It has always felt “parallel” not “integrated”. Mainstream financial services actively tries to keep it that way AI is quickly leading to massive new software ecosystems and tooling, and it’s directly affecting mainstream development workflows (and has from day 1). Mainstream enterprise unlike crypto is actively adopting (or being forced to adopt through shareholder pressure) Certainly bubble-esque vibes, and may companies will flop, but the core ROI is definitely there, just a matter of companies figuring out what the token spend threshold is, and tools/infrastructure becoming more cost effective to reduce the denominator of that equation But the trillions of dollars riding on it is certainly a big bet that needs a big payoff to breakeven

The bear case on AI is NOT that "AI doesn’t work". It clearly does. The bear case is this: Silicon Valley in recent years has an extremely poor record of understanding how humans actually use tech. In the past five years: Bitcoin as payments, NFTs as art, the metaverse, VR headsets. Every time the tech "worked". But mass adoption did not happen. In retrospect, it seems obvious that people wouldn't want to use bitcoin to buy stuff in the metaverse. But as recently as 2021 many people earnestly believed it. Here's the bigger problem. Bitcoin, metaverse etc were consumer products. Relatively simple. By contrast, a big part of AI is targeted at businesses. These are WAY more difficult to understand. Businesses are the aggregation of thousands of different people, all doing things that even people within the business don't understand. This makes prediction way more difficult. Then you get the question of whether AI adoption is actually profitable. Again, no one actually has a clue. So far companies are spending loads on AI inference. Costs are rising. But there are VERY few instances of companies seeing higher profits as a result of AI use. The notion that "once AI is good enough, profitable adoption at scale will follow" is a MASSIVE bet with trillions of dollars riding on it.

@paris_steff @econcallum This was so clearly written by ai🤣

@econcallum You’re framing AI as a product adoption problem. It’s not. It’s a cost-compression layer. Value won’t show as revenue first… it shows as margin pressure everywhere else. The mistake isn’t overestimating AI demand. It’s underestimating how it reprices everything simultaneously.

@econcallum Brilliant brilliant text. And we'll know soon enough. And we'll sell old school stuff to AI dropouts who have ruined their internal processes, teams and product by adopting unproven tech

@elonmusk Look at Cuba and North Korea and think again

Universal HIGH INCOME via checks issued by the Federal government is the best way to deal with unemployment caused by AI. AI/robotics will produce goods & services far in excess of the increase in the money supply, so there will not be inflation.

LLMs might accelerate alpha decay in trading Here's how it works in a nutshell: - free basic data - conventional "top of mind" indicators and strategies - something LLMs know and trust because of multiple sources - very similar backtest fitting All of these combined lead to generating signals that use very similar logic. This crowding effect will cause 2 things: alpha decay and volatility spike. I'll expand on vol next time, alpha decay is a big enough topic One recent paper in The Journal of Finance shows that anomaly returns are concentrated in the first month after information release and then decay soon after. In other words, even before LLMs, the monetization window for many signals was already narrow. Now add LLMs to the research stack. A 2025 paper on LLM-driven alpha mining makes the risk explicit: LLM-based approaches can rely too heavily on existing knowledge, generating homogeneous factors that worsen crowding and accelerate decay. That is the part I think the market is underestimating. If many teams are drawing from overlapping public data, familiar factor logic, and similar backtesting workflows, LLMs help discover signals faster. But they also make it easier to produce the same kinds of signals faster. My bet: LLMs will not kill all alpha. But they will compress the shelf life of copyable alpha. So what can traders do about it? Option A: accelerate strategies launch. Have a pipeline, be ready to halt underperformers. Expensive, time consuming, capital heavy Option B: add alternative data, create/buy prop indicators, work on portfolio composition

@TheSpaceSoviet But there can be zero gains for months

I'm increasingly convinced that there is never going to be a major crash again. The market is so divorced from the "real" economy that 90% of the population could be eating dirt and the S&P would still be noting record gains.

@YosarianTwo The more I learn about LLMs, the more I see it as a great tool for people who know what they are doing. And those people are never going to lose their job.

"AI is not ever going to be intelligent or a threat to us, it's a glorified chatbot, but it's going to steal all our jobs" is a weirdly common opinion. Do people think it's likely LLM's can do so many jobs they cause mass unemployment without being a general intelligence?