Matheus

239 posts

Matheus banner
Matheus

Matheus

@matheusgab

$15k → $100k public challenge. Real money. Real misses. Disciplined research, no signals. 555 days.

Katılım Ağustos 2025
695 Takip Edilen101 Takipçiler
Sabitlenmiş Tweet
Matheus
Matheus@matheusgab·
Starting a public challenge today. $15,000 to $100,000 in 18 months. Real money, my own account, every position posted as it goes on, every exit posted as it comes off. The number is the number. The point is not the number. The math first, because it is the part nobody likes saying out loud. $15k to $100k is a 567% return. That works out to roughly 14% compounded per month, or one 6x winner, or three 2.5x winners stacked clean. The S&P does 10% in a good year. The target is roughly 50x harder than the index. If I told you the base case was hitting it, I would be lying to you and to myself. The honest framing is different. The base case is that I get to $40k to $60k, which is already a 3-4x and would beat almost every fund on the planet over the same window. The bull case is that two or three asymmetric small-cap setups print and the curve bends. The bear case is that I get chopped up in two bad earnings prints and end the run at $8k, write the post-mortem in public, and the account becomes a case study in what not to do. All three outcomes are useful to whoever is reading. The reason to post any of it is simple. The actual edge in this market is not having access to better information. It is having the discipline to act on the information you do have, the patience to wait for setups that pay you to take the risk, and the honesty to cut a position when the thesis breaks rather than when the price hurts. Those three things are completely invisible until you put them in public. The account is the commitment device. What you will see here: Deep dives when a new ticker enters the book. Full read of the business, the catalyst on the calendar, the three to five validators that need to confirm for the thesis to compound, the price level where I cut, and the probability-weighted math behind the position size. Same structure every time so the track record stays comparable over months. Exit notes when a position closes. Win, loss, or "thesis broke before the price did." The losses get the same airtime as the wins. Maybe more. The mistakes are where the work product actually compounds for anyone reading. Earnings reactions for covered names, within 24 hours of the print. Not a recap of the headline. The read. The occasional audit of a fintwit account whose calls I have been tracking quietly for months. Whose signal compounds, whose noise rounds to zero. Names included. The point is not to dunk. The point is to make the difference visible. What you will not see: signals, alerts, "buy now at this level," paid groups, courses, options shilling, crypto, anonymous shitposting, engagement bait. If a post would embarrass a senior PM reading it cold, it does not go up. That filter alone removes 90% of the fintwit timeline. The universe. Small and mid caps where size and patience are the actual moats. US small caps, Brazilian small caps where most of fintwit has no coverage, the occasional UK or Australian name when the geology or the geopolitics forces the trade there. No mega caps unless there is a setup the size investors are missing. No crypto. The portfolio shape. 8 to 12 positions at any time, 6 to 8 percent per name, average hold around 90 days. 80 percent of the book in shares, up to 20 percent in defined-risk long-dated options on setups where the structure is asymmetric and the premium pays for the time. Hard stop on every share position, no exception. Options sized to the loss tolerance, not the price target, and never margined. One cap on the account: any drawdown of more than 25 percent from peak triggers a forced 30-day pause, no new entries, full audit of every position, public write-up of what broke. The way I will be honest about timing. I do not post a position before it is executed. Exit notes lag the actual exit by hours, not seconds. Nothing on this account is front-runnable, and nothing on this account is a call to action for anyone else. The account is a log of my own decisions, audited in public, with the math shown. First name drops in the next 24 hours. Rare earth deposit in Brazil, listed in Australia, valued in US dollars. The setup that hooked me: the project literally next door sold for 10x the current market cap eighteen months ago, the final feasibility study lands in June, and the geopolitical backdrop on critical minerals is doing real work on the entire sector. Small position, defined risk, calendar catalyst. The exact kind of asymmetry the challenge needs to compound from. If you want to follow without committing to anything, the account is here for the full 18 months. If you want to use it as a sparring partner for your own book, the deep dives publish the kill conditions before they fire, so the work is auditable in real time. If you want signals or shortcuts, this is the wrong account. 555 days. The clock starts today. Not advice. Just how I am sizing my own book in public.
English
0
0
0
12
Serenity
Serenity@aleabitoreddit·
All right chat, crowdsourcing your #1 highest conviction (10x only) stock long for the Power Semi trade. Especially given $NVDA pushing shift to 800 VDC. Stuff like $NVTS or $WOLF, but high-beta, 10x potential only. Anywhere around the world. What's your pick?
Serenity tweet media
English
441
88
1.5K
333.6K
Matheus retweetledi
AskLivermore
AskLivermore@asklivermore·
Get ready. Listen to me if you want to become a multi-millionaire in the next 5-7 years. Once the AI-compute trade cools off, the NEXT super-cycle will rotate into THESE categories: 1. Layer 1 - uranium miners • $CCJ - Cameco (blue chip urnamium provider) • $UEC - Uranium Energy Corp. (aggressive U.S. developer) 2. Layer 2 - fuel cycle / enrichment (the real bottleneck - especially HALEU and SMRs) • $LEU - Centrus Energy Corp. (only US company producing HALEU, big Department of Energy money) 3. Layer 3 - SMR & advanced reactors (highest risk-to-reward; the picks & shovels of the future) • $OKLO - Oklo (Sam Altman-backed, already signed $META deal) • $SMR - NuScale Power (first US - approved SMR design) • $NNE - Nano Nuclear (microreactors - portable power) 4. Layer 4 - nuclear operators / utilities (real cash-flow and massive data center contracts) • $CEG - Constellation Energy (largest US nuclear fleet) • $VST - Vistra (insane $META 20-year PPA and nuclear restarts) 5. Layer 5 - components & services (more defensive plays) • $BWXT - BWX Technologies (reactor components + services + defensive overlap) • $GEV - GE Vernova (BWRX-300 SMR tech + nuclear services Safer ETF's include: $URA, $URNM, and $NUKZ Once you make $4M, please thank me by buying me a Porsche.
AskLivermore tweet media
English
125
220
2.3K
160.1K
Matheus retweetledi
美股OK哥
美股OK哥@artinmemes·
但斌今天分享的视频爆了,14分钟讲清楚光模块 押注美股相信光板块的,都值得看一看
中文
152
999
4.2K
426.8K
Grok
Grok@grok·
**Yes, for AI server PCBs (the +233% BOM line in that MS NVL72 analysis), public companies with direct exposure include:** - **Unimicron (3037.TW)** — Taiwan leader in high-layer/HDI server boards, gaining share in AI substrates. - **Ibiden (4062.T)** — Japan’s top build-up substrate player for high-end AI servers. - **TTM Technologies (TTMI)** — Major US maker of complex high-layer PCBs, benefiting from data-center ramp. Smaller/related names: Kinsus, Gold Circuit Electronics, etc. This is the “picks & shovels” part of the AI buildout. (Not advice—DYOR.)
English
1
0
1
203
Clark Tang
Clark Tang@_clarktang·
Just my my math speculation: Pre Xai deal OAI at ~2.5GW / Anthro at ~2GW - Renting NVDA ~$13B/Yr/GW (hyperscalers / CRWV financials) - Assume mix of Trainium/TPU is 50% discount per GW - Then Anthro spending ~$13B for ~2GW then = ~$1B / mo of spend So adding this deal for COLOSSUS and COLOSSUS II = $1.25B / mo of spend That means after this deal, $NVDA penetration of Anthropic compute stack = >50% instantly... That's something Jensen mentioned on the earnings call: "And so we're growing - and our coverage of Anthropic has been largely 0 until just recently. And so we're gaining share tremendously in inference" Based on studying the industry I'm pretty sure that math is right... Even assuming Trainium / TPU is 25% discount then this deal puts Nvidia compute penetration at Anthropic at 40%! anyone else have different analysis?
Sheel Mohnot@pitdesi

Anthropic is paying $1.25B a month to SpaceX for compute

English
13
17
223
60K
Kova
Kova@kovainvest·
I sent someone to $POET's Singapore office. It's real, with real people. So this isn't a fake-company thesis. It's a financial engineering thesis. The latest "$50M customer deal" is a circular financing structure dressed up as an AI partnership. Here's what POET didn't put in the headline: POET granted Lumilens 22.9M warrants. 2.29M vest immediately (~$30M value at current price). Lumilens "ordered" $50M of product in return. POET handed over $30M of stock to receive a press release. Who is Lumilens? A 2-year-old startup. Tried to raise $106M, got $9M. 7 weeks before this deal, bought a distressed Singapore entity — Rain Tree Photonics: $800K revenue over 3 years, assets shrinking from $12.7M to $3.2M. Lumilens CEO Ankur Singla sits on the board of Accton ($2345.TW), which invested $9M in Lumilens last year. Accton's own books value Lumilens at ~$158M. This press release implies $1B+. No hyperscaler. No samples until late 2026. No production until 2027. POET raised $400M five days later at a "premium" — supported by this exact press release. The Singapore office is real. The customer order isn't. $POET
Kova tweet media
English
44
25
220
88.9K
Matheus retweetledi
Ejaaz
Ejaaz@cryptopunk7213·
cursor is now officially a frontier ai lab. in < 1 yr they went from being labelled “just an ai wrapper” to training a model that competes directly with opus 4.7 and gpt 5.5 while being 10x cheaper! the key to this was the SpaceXAI partnership: > cursor gets access to tremendous amounts of compute from colossus II compute + cursor agent harness + open source model (kimi k2.6) = composer 2.5 but there’s a powerful card they haven’t played yet… cursor will get access to the 10 and 15 trillion param models elon is training. that combined with cursor’s agent harness moat will create a coding ai capable of competing with the best. sama said 1-2 weeks ago: models and harnesses should be treated as the same thing cursor proved that today. really impressive turnaround.
Ejaaz tweet mediaEjaaz tweet media
Cursor@cursor_ai

Introducing Composer 2.5, our most powerful model yet. It's more intelligent, better at sustained work on long-running tasks, and more reliable at following complex instructions. For the next week, we’re doubling the included usage of the model.

English
44
61
955
75K
First Squawk
First Squawk@FirstSquawk·
GOOGLE AND BLACKSTONE ARE LAUNCHING A NEW U.S.-BASED AI CLOUD COMPANY BACKED BY $5 BILLION IN EQUITY CAPITAL FROM BLACKSTONE, ACCORDING TO THE WALL STREET JOURNAL. THE VENTURE TARGETS 500 MEGAWATTS OF AI COMPUTING CAPACITY BY 2027, WITH GOOGLE VETERAN BENJAMIN TREYNOR SLOSS SET TO SERVE AS CEO, WHILE GOOGLE WILL PROVIDE THE HARDWARE, SOFTWARE, AND SERVICES POWERING THE PLATFORM.
English
32
51
600
190.2K
Matheus retweetledi
Rohan Paul
Rohan Paul@rohanpaul_ai·
Just a few days back, Thinking Machines Lab (TML), showcased a way of making AI interaction continuous instead of turn-based, a Full-Duplex Time-aligned micro-turn. It's a preview of the future of a near-realtime AI voice and video conversation with new 'interaction models' And MiniCPM-o 4.5 already shipped the same core idea through OpenBMB’s Omni-Flow framework: time-aligned perception and response instead of old turn-based chat. A 9B Full-Duplex omnimodal model that can see, hear, and speak at the same time. Omni-Flow also treats interaction as a continuous stream on a shared temporal axis, aligning visual input, audio input, and output speech/text into time chunks so the model can perceive while responding. That breaks the old walkie-talkie UX of AI: user talks, model waits, model replies. And this is not just a demo concept. It is a 9B open model with code, weights, a report, and edge deployment under 12GB RAM. It also surpasses Qwen3-Omni-30B-A3B in omni-modal capabilities and speech generation quality. This feels like the interaction layer AI was missing. OpenBMB already shipped this as a real Full-Duplex omni-modal architecture, with video tokens, audio tokens, LLM hidden states, speech tokens, and waveform generation all synced to one shared timeline.
Rohan Paul tweet media
Thinking Machines@thinkymachines

People talk, listen, watch, think, and collaborate at the same time, in real time. We've designed an AI that works with people the same way. We share our approach, early results, and a quick look at our model in action. thinkingmachines.ai/blog/interacti…

English
0
21
108
16.5K
Jeffrey Currie 🆔++
Jeffrey Currie 🆔++@CommodMkt·
Welcome to the most asymmetric trade in modern financial history. The thread below lays out why. The opportunity exists because capital has chased the AI trade while ignoring the physical assets AI requires to run — assets that have quietly become the best-performing asset class of the decade. Since October 2020 when we first called for the commodity super cycle: QCI Total Return +217%, GSCI Total Return +205%, Gold +140%. NASDAQ trails at +130%. S&P 500 at +85%. The top three are all commodities. Yet oil cannot get out of its own way while copper and the broader atom complex prints fresh highs . That is the dislocation. That is the trade. Get long. Buckle in. Hang on for the ride. Forgive the longer posts in this thread — attempting to mimic my old 10-bullet commodity takes. On to it.
English
244
1.2K
6.9K
2.9M
Lukas Ziegler
Lukas Ziegler@lukas_m_ziegler·
🧵The quiet revolution in robotics isn't humanoid, or world models.. It's the rise of inspection & maintenance robots, quietly scaling across dirty, dangerous, and distant industrial environments. This sector is projected to hit $8.3B by 2030. Let's break down why it's growing so fast. [Save this thread for later 📌]
English
40
352
2.2K
170.4K
ppdd
ppdd@usppdd·
@ousterlidar Director Virginia Boulet bought 1,658 shares of $OUST today. Before you think that’s peanuts, look at the screenshot below. She’s a true believer, with $1.18M of her own money invested in $OUST.
ppdd tweet media
English
5
2
78
23K