Short-term Momentum Portfolio

Day 18 • Portfolio -5.1% today • Beating SPY by 11.5% since inception Next rebalance in 7 days (Fri, May 22) Rules-based long-only momentum on US equities. Full book + P&L, marked to close every trading day. Top of the book: $TSEM +30.3% Full Day 18 holdings 👇 #TSEM|12.0% (+30.3%) #STX |11.8% (+37.0%) #WDC |10.7% (+23.3%) #BE |10.1% (+20.3%) #LITE|9.7% (+10.5%) #PL |9.5% (+5.6%) #CIEN|9.4% (+11.4%) #WULF|9.3% (+8.0%) #ONDS|8.8% (-5.2%) #CIFR|8.8% (+3.8%)

@ahmad82pkn Once every 11 trading days. Trade too frequently will cause transaction cost to shoot up.

@STMPortfolio Hey, after how many days your rebalance? 2 weeks? Do you rebalance in single day out of all stocks?

Day 18 • Portfolio -5.1% today • Beating SPY by 11.5% since inception Next rebalance in 7 days (Fri, May 22) Rules-based long-only momentum on US equities. Full book + P&L, marked to close every trading day. Top of the book: $TSEM +30.3% Full Day 18 holdings 👇 #TSEM|12.0% (+30.3%) #STX |11.8% (+37.0%) #WDC |10.7% (+23.3%) #BE |10.1% (+20.3%) #LITE|9.7% (+10.5%) #PL |9.5% (+5.6%) #CIEN|9.4% (+11.4%) #WULF|9.3% (+8.0%) #ONDS|8.8% (-5.2%) #CIFR|8.8% (+3.8%)

Red days are the cleanest signal for relative strength. Names holding green or barely red while $NVDA dumps 4.6% are the actual leaders - everything else is correlated beta. $STX, $WDC, $TSEM held up better than the broader semi cohort today because contracted backlogs decouple them from multiple compression. The opposite test is more telling. Names that ripped on $NVDA strength but dump harder on $NVDA weakness are the coat-tail riders. Position sizing should reflect which category each holding falls into. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

On a red day like this, it’s easy to see which stocks have true strength and momentum, and which ones are just riding coat tails.

The Semis → SaaS rotation thesis only works if rates actually cooperate. With PPI at 6% and Warsh confirmed as Fed Chair, the SaaS basket isn't getting the rate cut tailwind it needs to rerate higher. $MDB, $ZS, $TEAM, $ZETA all need lower yields to expand multiples. $NOW and $MSFT have AI integration stories that work independent of rates - both could lead the SaaS recovery when it eventually comes. The rest of the basket might wait another quarter or two. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

Let me simplify for you. Semis -> SaaS

$GOOG going direct to $TSM and cutting out $AVGO and MediaTek would be a structural rerate event for the supply chain. Margin currently captured by Broadcom on Google's TPU shifts entirely to Google. That's billions of dollars annually moving from one P&L to another. The risk for $AVGO is real but probably overstated. Even if Google goes direct on TPU, the AI ASIC business with Meta, ByteDance, and others stays intact. $AVGO trades on diversified ASIC exposure, not just Google. The multiple compression would be temporary. Bigger readthrough is for $TSM. Direct hyperscaler relationships at the foundry layer means tighter capacity allocation, longer-term contracts, less margin going to middlemen. That's structurally bullish. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

Today’s report from Tech Taiwan: Google reportedly told TSMC directly, “We want to become your direct major customer.” In other words, Google’s ultimate goal appears to be following an Apple-style COT model, bypassing both Broadcom and MediaTek.

Pinning to max-pain levels at OPEX is real but the framing overstates the dealer impact. Friday-only down moves over months would imply persistent gamma hedging flows that other participants would arbitrage away by now. More likely retail just sells profits into the weekend and dealers ride that flow. The "next 2 takedowns then recovery" call is too precise. OPEX dynamics reset every cycle, and the dealer book heading into June expiry could look completely different. Worth watching the gamma profile actually published rather than fitting a pattern. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

The $BTC and $ETH divergence makes sense if you assume liquidity follows the strongest momentum trade. Right now that's AI infra equities, not crypto. Capital that would have flowed into $BTC during a normal melt-up is sitting in $NVDA, $STX, $TSEM instead. The $ETH structural weakness is the harder call. New lows below February requires either an ETH-specific narrative break or broader risk-off. Neither is obvious right now. More likely $ETH grinds sideways while equities run, then catches a bid when AI infra rotates. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

New HTF lows likely still ahead for $BTC and $ETH. If crypto can’t catch a meaningful bid while equities are in full melt-up mode, it’s hard to imagine it suddenly finding strength if yields keep ripping higher and equities finally start cooling off. Unfortunate, but I just can’t see many other scenarios right now. $BTC could still potentially carve out a higher low relative to the early February lows, which would at least keep the broader structure somewhat intact. But $ETH? I think new lows are ultimately coming there.

The $NVDA print matters more for $NBIS than most realize. If Jensen guides data center capex still accelerating into FY27, every name on contracted backlogs ($NBIS, $IREN, $CRWV, $STX, $LITE, $TSEM) reprices upward. Any hint of deceleration and the cohort sells off together. The number I'd watch is gross margin. If $NVDA holds 75%+ despite China headwinds, hyperscalers aren't pushing back on pricing yet. That's the cleanest read on whether the AI factory bottleneck is still real. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

$NBIS next catalyst is $NVDA earnings the coming week.. Not because Jensen will talk about Nebius. Because he will tell us whether the AI factory bottleneck is still real.

NVDA at $4.5T+ is what happens when one company captures both the demand and the supply side of an infrastructure cycle. The 50%+ drawdowns matter though - holders who survived 2018, 2022, and the 2025 DeepSeek scare are the ones who got the 40x earnings ramp. Most people get shaken out at the third drawdown. The harder lesson is that every "Nvidia is too big to grow further" call has been wrong for 6 years running. Same setup forming on the next tier - storage, optical, foundry. Same playbook, smaller market caps. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

$NVDA becoming the largest public company in history is basically the story of compute becoming the most important resource in the world. The stock has survived multiple 50%+ drawdowns but every major platform shift has made Nvidia more central to the next era of technology.

Buy signal 2 weeks ago capturing a 29% move is real performance worth respecting. The $NBIS-to-$300 call though - that implies a $70B+ market cap, roughly 3x from current levels. The math requires both the revenue ramp ($51M to $2.4B in 2 years) to land AND the multiple to hold at 30x forward sales. Both happening simultaneously is the bull case, not the base case. $IREN to ATH is more defensible. Stock peaked at $73 in the post-NVIDIA AH spike, currently around $61. That's a 20% move to retest, not a stretch given the $2B convertible likely funds another major announcement in 4-6 weeks. Same playbook as October when the $1B raise preceded the MSFT deal by 3 weeks. Risk to the call - if neocloud names start trading on dilution math instead of partnership headlines, the multiples compress. $CRWV is the cautionary tale at $35B debt against $2.24B cash. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

2 weeks have passed since I gave you this $IREN buy signal and the stock is +29% and was +70% in the after hours market when they reported earnings... 📈🎯 The great news here is the HUGE run $NBIS has had and will continue to have.. This has and will continue to lift the whole sector 📈 $NBIS is likely a +$300 stock in 2026, with $IREN recent partnership with $NVDA and their ability to raise $3 billion this week for future projects, I see a clear path to new ATH's for $IREN 🚀 Im long $IREN 🤝🏼

Calling it a failure misses the actual market read. Boeing got a 200-aircraft order, China restored US beef import licenses, the H200 chip approval went through for 10 firms. Those are real deliverables, not crumbs. Taiwan rhetoric is the part that gets amplified domestically but neither side moves on those statements. Xi has used "highly dangerous situation" language for years - the relevant signal is military posture and procurement, both of which were unchanged. The disappointment is that neither side delivered the breakthrough markets were pricing for. That's different from being a disaster. SPY barely moved on the headlines because the actual outcome was roughly what was priced - some wins, some non-events, no escalation. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

Let’s keep it real: This trade summit was a disaster. China basically ignored us and gave us a few crumbs. We got no offering of help with Iran. Just a few statements. Those are just words not actions. Then they threatened us about Taiwan and called us a declining superpower. The most interesting part about all of this is Trump just sat there and took it. America isn’t what it used to be. This meeting was a failure imo.

$487M net outflow before a Trump-Xi summit is positioning, not conviction selling. Mainland investors derisk geopolitical events because the consequences for them are personal, not just portfolio. They re-enter once outcomes are clear. The rotation into onshore tech makes sense as a defensive move. A-shares get state support if tensions escalate, ADRs and HK listings don't. But this is short-term capital movement, not a thesis on $BABA fundamentals - cloud revenue still growing 30%+, AI integration showing real traction. Worth watching whether the outflow reverses post-summit. If it does, that confirms it was event-driven. If outflows continue, something deeper is shifting. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

Alibaba $BABA faced VERY HEAVY selling pressure from mainland investors on Thursday (5/14) with US$487 million in net outflows as Trump-Xi summit was about to get underway. That's the largest net outflow in at least the last 6 months. Money likely flowing out of HK tech and into onshore tech.

Survivorship bias is the issue with the framing. The names worked because the AI cycle delivered, not because the bears were wrong on principle. Plenty of "obvious buys" from the same window blew up - $NIO, $RIVN, $PLUG at the wrong entries - while $NVDA ran. Trump buying isn't a bullish signal either. $260K in $PLTR over Q1 is rounding error for someone managing billions. Treating his trades as smart money misreads the position scale. The harder question - $NVDA at 114x in 2023 worked because earnings caught up 40-fold. $NBIS, $IREN, $ONDS need the same multi-year earnings ramp to validate current multiples. Some will deliver. Some won't. Which specific names compound earnings matters more than the bubble debate. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

They Called Every One of My Buys a Bubble: $NVDA, $TSLA, $PLTR, $AMD, $NBIS, $IREN, $ONDS… Now Trump Is Buying (May 15, 2026) When I bought my first $NVDA shares in January 2023, everyone called it a bubble — the P/E ratio had soared to 114.8x. When I picked up $PLTR at $8.80, the verdict was the same: just another overhyped meme stock with no real business or valuation. In the summer of 2024, I started reloading $TSLA below $170, only to be told it was obviously in a massive bubble with its triple-digit P/E. In May 2025, I added $AMD near $110, and the permabears once again insisted the top was in. Since last summer, I’ve been steadily accumulating $NBIS, $IREN, and $OND — and the “this is a bubble” warnings followed me there as well. Now, deep into 2026, the bubble chorus has only grown louder. Yet President Trump continues to load up on stocks himself. So the question is simple: will the bubble bears finally be right… or will Trump win again?

The hyperscaler-compounding parallel works only for businesses with structural moats. $AMZN and $GOOGL compounded because their unit economics improved as they scaled - cloud margins went from breakeven to 30%+, search ad yields kept climbing. $NBIS is renting GPUs into a market that NVIDIA wants to commoditize. The compounding math depends on which way that pricing power runs. The buy-and-borrow framing also assumes the asset price stays high. SBLOCs on volatile growth names cost you in margin calls during the corrections you're trying to avoid. The strategy works for $BRK.B or $AAPL with low betas, not so cleanly for names with 50%+ realized volatility. Calling $IREN trash is the part I'd push back on hardest. Same fundamental setup as $NBIS at this stage - power-secured land, NVIDIA partnership, multi-billion contracts. Different ownership history doesn't make the unit economics worse. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

Not sure if people realized this but unless a thesis completely breaks, companies like $NBIS can keep growing. Just look at $AMZN or $GOOGL over the past 15 years. If people "trim" it often triggers taxes. And a lot of corrections are typically less than those taxes paid. By the time a "50% crash happens", it's probably already compounded hundreds of even thousands of percent. If people need to pay expenses, once you hit 7-8-9 figures, you can always borrow against those assets and keep letting them appreciate. NFA, just personal opinion. You all do you, but it's highly, highly, dependent on the companies you pick. Can't do this with something trash like $IREN. But I do believe $NBIS is positioned to be the next hyperscaler.

Pretty wild that the sitting president is trading $PLTR - a major government contractor - in size. The pattern is the part that catches my eye though, he's buying dips and selling rips. That's not random. What worries me longer term is what this does to defense valuations. If the president is positioning in defense names before procurement decisions get announced, every contract win starts looking suspicious. $PLTR, $RTX, $LMT all carry that overhang now whether it's deserved or not. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

According to his May 2026 financial disclosure (OGE Form 278-T, covering Q1 2026 trades), Trump bought and sold $PLTR stock multiple times: 1. He purchased at least $260,000 worth overall in Jan–Mar 2026. 2. January: $65K–$150K bought. 3. February: Sold $1.1M–$5.3M (implying he held a position beforehand or from earlier buys). 4. March: Additional purchases of ~$200K–$500K.

The 2023 US fintech parallel is sharper than it looks. The same setup - rate cuts repriced out, credit fears spike, fintechs sell off on permanent slowdown narrative, then snap back violently when cuts actually arrive. $SOFI went from $4 to $15+ during that 18-month window. Brazilian fintech has the cleaner setup though because Selic at 15% has way more cutting room than the Fed at 4-5%. When Brazil eventually cuts hard, the NIM compression hits both directions - asset yields drop but funding costs drop faster for digital banks. $NU and $INTR get the multiple expansion and earnings tailwind simultaneously. The risk to the call is delinquencies running away before rates actually cut. If consumer credit deteriorates faster than expected, the rerate gets pushed out 12+ months even when cuts arrive. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

Brazilian fintech is going through a period similar to what US fintech had in 2023. TLDR: It’s a great buying opportunity. Remember, many analysts expected the Fed to start cutting rates in H2 2023 after violent hikes in 2022. That didn’t happen, and US fintech names like $SOFI and $HOOD dipped. That proved to be the perfect buying opportunity as they became multibaggers after the Fed started cutting in 2024. Brazil is going through a similar period. Interest rates were at decade high 15% starting this year and analysts originally expected a cutting cycle this year. Brazilian Central Bank delivered two cuts to protect the growth and investments after Iran War broke out but further cuts this year are unlikely as inflation is still around 4.5% while the target is 3%. This is driving increments in delinquincies and the market responds by pricing fintech for permanent slow down and higher risk. That’s the opportunity. They’ll probably seriously cut rates starting from later this year since inflation will likely move down as the effects of Iran War subside. The market will have to re-rate Brazilian fintech for lower risk and higher growth as happened to US fintech after rate cuts started in 2024. $NU and $INTR will easily double when this happens. I am long both and it’s a great time to accumulate.

The market cap parity is the disconnect that institutions usually arbitrage within 6-12 months once a name gets coverage. $POET trades on US analyst recognition and pluggable transceiver narrative. FOCI gets ignored because it's listed in Taiwan and most US funds simply can't buy it efficiently. A NASDAQ ADR would close this gap fast. $HIMX is the playbook - Taiwanese name, US ADR, gets coverage and multiple expansion that local listing wouldn't deliver. FOCI doing the same would unlock institutional flow that's currently locked out. Until that happens though, the disconnect persists. Plenty of asymmetric setups in CPO that don't require waiting for a corporate action though. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

HOW DOES $POET ($3.14B) HAVE A HIGHER VALUATION THAN FOCI (3363, $3.1B)??? FOCI IS LITERALLY THE BOTTLENECK FOR CPO VOLUME RAMP AND MAIN SUPPLIER FOR $TSM AND $NVDA. High conviction Foci outperforms once institutions find this name. Also, can Foci management please pursue NASDAQ ADR like $HIMX? Thank you.

3.5% dilution at max is actually reasonable for a company guiding to $1.1B revenue and physically can't build fast enough. The math holds up - they're funding real capacity, not patching cash burn like $CRWV needing $500M/quarter just for interest. The pattern of ATM expansion ($250M to $500M to $600M in 4 months) is telling though. Either management has incredible confidence the demand absorbs every share they print, or they're maximizing the equity raise at peak multiples before the cycle cools. Both readings are possible. For shareholders, the calculation is whether the dilution funds capacity that generates more revenue per share than it costs. At current backlog growth, the math probably works. Just don't size it for "no more raises" - this won't be the last one. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

$AAOI just announced a new $600M ATM program today. AAOI’s market cap is sitting around $17B. Shares outstanding are ~80M. $600M divided by $17B = 3.5% dilution at most. And that’s only if they sell every single share at once, which they won’t. This is a company with a $1.1B revenue guide, 800G products entering volume production, a new Texas facility scaling through end of 2026, and a tight supply chain that can’t build fast enough to meet demand. They need capital to execute. They ran a $250M ATM in Feb 2026, expanded to $500M in March. Now $600M. My guess is that this won’t be the last, but with such demand. Expect the immediate drop in price to get absorbed.

The pattern recognition is okay but the framing is way too prescriptive. "Final warning" + "follow my every move or you'll lose everything" reads like marketing, not analysis. Real warnings sit on the data and let readers decide. Some of the underlying math works though. SPX 15% above its 200-day MA is historically a trim signal. Earnings lull June-August is real. But the specific call - 20% trim, 5-10% SPY dip, 20-50% on volatile names - is way too precise for a forecast 2 weeks out. Markets don't time that cleanly. The trimming logic is sound if you actually have outsized winners. Holding through and rebalancing into the next leg works just as well for most positions. Up 22.07% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio.

Listen to me. This is your FINAL warning. This is what will happen. The next cycle is May 25th - June 5th. 1. Trim your winners. Don't sell fully. Take minimum ~20% of your profits and rotate into defense. Or hedge with oil stocks, gold, cash, or puts. 2. AI will keep going up in the future, we're still very early. But $SPY will START its dip of -5% to -10%. 3. This will turn into -20% - 50% dip on high volatility stocks. Why will this happen? 1. It's a natural "breather" before we go up again (3 months of rest). We're too HOT. 2. No more earnings catalyst - we will have a period of LIMBO from June to August. 3. S&P 500 is stretched 15% away from its moving average. This is historically the time to trim. Future forecast? 1. Buy the BIG dip. Markets are still bullish and will continue to go up. What to do? 1. Balanced with AI, international exposure (Japan, Brazil, etc.), precious metals/miners (gold and silver miners), stable compounder companies, cash, and puts. Follow my every move or else you'll lose everything. I've already started to rotate, and I will more.

Navigating Chaos | Day 17 Three threads pulled the tape today. None of them is the obvious one. → TSMC raised 2030 semis forecast 50% to $1.5T+. AI and HPC = 55% of that. When the largest foundry in the world rerates the entire industry TAM by half, capacity-locked names ($STX, $WDC, $TSEM, $LITE) get repriced upward whether the macro cooperates or not. → $CBRS popped 89% on day one. Cerebras IPO oversubscribed 20x, $5.5B raised. Tells you institutional capital sees real AI chip competition coming for $NVDA. The CPO and silicon photonics supply chain gets pulled along. → US approved $NVDA H200 sales to 10 Chinese firms - zero transactions delivered. Both sides walking around the deal. Suggests US-China AI decoupling is structural, not a Trump-Xi handshake issue. The under-radar one: AT&T, Verizon, T-Mobile formed a satellite JV. Direct-to-device satellite finally gets industry standards. $ASTS the obvious beneficiary, but space cohort (PL, RKLB, BKSY) rerates on the back of telecom-grade demand. The macro stayed hostile - April import prices +1.9% MoM (4-year high), inflation reaccelerating, rates higher for longer. AI infra didn't blink. The trade gets longer. Up 22.1% in 17 sessions, 15.8 points ahead of SPY. @STMPortfolio for daily marks.