Long-Term Pick

😃 #ADBE is down 20% over the last year. The reason is a big bet: Adobe gives its AI tools away for free to build a bigger user base. It works: free Creative users grew from 50M to 90M in a year, Acrobat and Express passed 850M, and AI-first recurring revenue tripled to over $500M. Management says this lowers revenue growth in the second half, and the payback comes only in 2027. The CEO and CFO are changing at the same time. In my view, Adobe is a strong business going through a hard year; I "hodl" them in my portfolio. Chart: EPS Consensus vs Price Dan from @longtermpick

📱 #RDDT is joining the S&P 500 next week. This means index funds and ETFs that track the S&P 500 now must buy the stock. That's extra demand. But the bigger effect should be later. Big passive investors will now hold Reddit long-term, not just retail traders - that should make the stock more stable. My analysis 🌐 Chart: EPS Consensus vs Price Dan from @longtermpick

🏃‍♀️ #ONON One-Pager The fastest-growing sportswear brand fell 14% in one day. I read the report, and I think the market read it wrong. The full analysis is on Tuesday, as always. Dan from @longtermpick

😳 The Shiller P/E (price vs 10Y average earnings) for the S&P 500 #SPX just hit 42.4. In a century and a half of data, only the dot-com peak of 2000 was higher - about 44. The 1929 top was 31. But here's the problem I see with this indicator: it has been "expensive" for most of the last decade, and selling on it would have cost you the whole bull market. It also looks backward - today's earnings are growing 30%+, which older averages don't capture. So I treat it like the other warning charts (concentration at 41%, record tech inflows): not a reason to sell, but a reason to have realistic expectations. From these levels, future returns are usually modest. Dan from @longtermpick

😎 In early 2023, #NVDA earned $12M of profit per day. Today it's $445M. And by fiscal 2029, analysts expect almost $1.3B every single day. That's a 100x jump in six years. My analysis. Dan from @longtermpick

🧪 Today, GPUs talk to each other through copper. It's cheap and reliable, but only over 1-2 meters. As chips get faster, the industry moves to co-packaged optics (CPO) - light-based links built into the chip package. The transition is slow. The chart shows copper holding until ~2028 and CPO winning only around 2030. #MRVL earns on both sides: copper now and CPO later. Dan from @longtermpick

😳 One more chart for the "how hot is this market" thesis. BofA tracks yearly inflows into tech funds since 2015. Most years brought $20-70B. In 2026 the line goes almost vertical: $216B annualized - roughly three times the previous record. It means the crowd is now fully in, and crowded trades fall hardest on bad news. For me this is a sentiment warning, not a sell signal. I keep my AI positions, but this is not the moment to add aggressively. Dan from @longtermpick

🏃‍♀️ I’m writing an analysis on On Holding (#ONON), which is undervalued, and I see their ads on Meta’s (#META) social media (Threads), which is also undervalued. Dan from @longtermpick

🏰 Top holdings of the VanEck Wide Moat ETF #MOAT - a fund that only buys companies with strong long-term competitive advantages, trading below fair value. The surprise leader is Veeva Systems #VEEV at 2.78%, ahead of Microsoft. Then LPL Financial, Schwab, Masco, Airbnb, Bristol-Myers, Broadcom, Nvidia and Palo Alto. Notice the structure: every position is ~2.5-2.8%, the top 10 is only a quarter of the fund. The opposite of today's S&P 500 #SPX, where ten AI names are 41% of the index. You stay in quality businesses - #MSFT, #NVDA, #ABNB - but without betting everything on one theme. I wrote before that at this level of concentration, adding something outside the Big 10 is common sense. Dan from @longtermpick

😇 A small but important signal: Berkshire Hathaway's #BRK cash pile just declined for the first time in four years. The mountain grew from ~$100B in 2022 to almost $380B. Quarter after quarter Buffett was selling stocks and accumulating Treasury bills. Now the pile dropped to ~$360B, which means Berkshire finally became a net buyer of stocks. This is also the first big move of the new era - Greg Abel took over as CEO this year, and his first signal is "we see things to buy". Dan from @longtermpick

💼 LTP Portfolio #001: Introducing My Portfolio Since December 2021, the portfolio is up +101.45% vs +67.48% for the S&P 500. That is +16.12% annualized with a beta of 0.95. Right now: 10 positions, ~40% in treasuries, and I just started moving that cash into stocks - #MSFT in July, #NVDA and #ONON in early August. LTP Portfolio #001 💼 Dan from @longtermpick

⚙️ Four stages of AI and how much memory each one eats. Training and basic chatbots mostly need HBM. But the next stages change the picture: advanced inference (long context, RAG) pushes NAND/SSD demand to "Extreme", and agentic AI (where we are heading now) does the same for conventional DRAM, with HDD demand coming back too. The AI memory story is not just HBM anymore. Agents with memory and long workflows need the whole storage stack. Bernstein tags #SNDK for NAND and #STX for hard drives. Dan from @longtermpick

📱 Reddit: 61% Growth, Priced for 20% #RDDT ended the June quarter with revenue up 61%, a net margin of 31%, and a FCF margin of 32%. It is now trading 21% lower than it was a year ago. › The growth at this scale is exceptional. The company reported $805 million in quarterly revenue, an increase of 61%, marking the eighth consecutive quarter of over 60% growth, which puts it on a $3.2 billion run-rate. This revenue growth, combined with the EBITDA margin, categorizes Reddit as a "Rule of 104" company. › Profitability came quickly. The gross margin is 91.3%, the adjusted EBITDA margin is 43%, and FCF reached $1.02 billion over the past year. Revenue per employee has just surpassed $1 million. › There is a clear path for monetization. The average revenue per user (ARPU) of $6.18 increased by 36%. The international market presents an opportunity, as 59% of users contribute only 21% of revenue, earning $2.26 per user compared to $11.85 in the US. › The balance sheet is strong. $2.79 billion in cash against $21 million in debt, and buybacks have already begun, two years after the IPO. › The stock multiple has dropped. It stands at 18.7x of Price/Fwd Earnings compared to a 47.9x average since the IPO, which is below the -1 standard deviation band. Additionally, the PEG ratio is 0.40, while consensus anticipates a 47% growth in EPS. › The bear case highlights some concerns. Daily users in the US saw a sequential decline, search referrals are unpredictable as Google implements AI responses, and nearly all licensing value comes from just two partners. The expected slowdown is set at around 48% for the next quarter after experiencing 61% growth. Full free analysis 🌐 Dan from @longtermpick

⬆️ With ~90% of S&P 500 companies reported, Q2 EPS is growing about 30% YoY. That excludes one-time investment gains at Alphabet and Amazon (with them it's almost 50%). Back on July 1, consensus expected only 22%. And 76% of companies beat estimates - the best rate since 2021. Where the growth is: Technology +72% on +36% sales, Energy +148%, Materials +36%, Financials +21%. Health Care is the only sector shrinking (-7%). Technology: profits growing twice as fast as sales means margins are expanding - AI is showing up in real earnings, not just in capex plans. #SPX Dan from @longtermpick

📱 #RDDT One-Pager. The full analysis is on Tuesday. Dan from @longtermpick

📱 Reddit: Starting a Deep Dive I'm preparing a detailed analysis of #RDDT, and the numbers I see so far are hard to ignore. Q2 2026: revenue $804.9M, up 61%, with a 91% gross margin. The company turned profitable only a year ago, and its net margin is already 31%. Higher than #META! Same with ROIC: 24% vs Meta's 19%. The stock is up 224% in three years. Three times the S&P 500. But it's down ~45% from its peak on fears that AI chatbots will steal its traffic. As a result, Reddit now trades at 18.7x forward earnings, the same price as Meta, while analysts expect ~47% EPS growth per year. PEG is 0.40. Full analysis coming soon 🌐 Valuation: › Price/Fwd Earnings: 18.8x (3Y avg: 47.9x) › Price/Fwd Sales: 8.0x › PEG: 0.40 Growth: › Sales 2026-28: $3.38B -> $5.50B (+53%, +31%, +24%) › EPS 2026-28: $7.56 -> $11.79 Quality: › ROIC: 23.9% › Net Margin: 31.4% › Debt/Equity: 0.06 › FCF: $1.02B Dan from @longtermpick

😬 A chart from UBS that continues the capex story. For years, Amazon, Alphabet, Meta, Microsoft and Oracle barely touched the bond market - $20-50B of new debt a year. In 2025 that jumped to ~$110B, and 2026 should exceed $250B. That's ~14% of all investment-grade bonds issued. One dollar in seven now comes from five tech companies. The message is simple: cash flow is no longer enough. Even the richest companies in the world have to borrow to keep up with the AI race. I don't see danger for #MSFT, #GOOGL or #AMZN. Their balance sheets can carry this easily. But the AI trade is slowly changing its nature: it now runs on leverage, and leverage always makes the downside sharper. Together with Nvidia's CDS chart from before, this is the second sign that credit markets are becoming the place to watch. Dan from @longtermpick

⚙️ BlackRock ranked the 30 most critical companies for AI. The surprise is at the top: Micron #MU is 1, ahead of AMD (2), TSMC (3) and Nvidia only at 4. Also interesting: storage names like Western Digital (10) and Seagate (11) rank above Microsoft (13), and Google, Amazon and Meta are not on the list at all. The logic here is criticality, not size. GPUs exist in volume, but HBM memory is sold out for years ahead - you can't build an AI cluster without it. That's why the memory and storage names crowd the top of the list. This matches my view from the Micron post before: memory is the tightest bottleneck in AI right now. Dan from @longtermpick

🐈 A strange combination in #META right now. The stock is down 21% over the last year. At the same time, analysts expect 21% EPS growth for the next five years - above Meta's own historical average of 16%. Faster growth, lower price: PEG has dropped to 0.89 against a long-term mean of 1.54. Why so cheap? The market is punishing Meta for its huge AI capex (~60% of sales) and for a slow 2026 - EPS grows only ~5% this year as depreciation from all those data centers kicks in. But then estimates show 12% in 2027 and 19% in 2028, with sales growing ~20% a year the whole time. I think this is one of the best risk/rewards among the big AI players: you pay 18.7x forward earnings for a company growing sales 20%+ - if the AI spending works even partly, the stock is simply mispriced. Valuation (LTM / NTM): › P/E: 22.2x / 18.7x › EV/Sales: 6.7x / 5.5x › EV/EBITDA: 13.4x / 9.6x › PEG: 0.89 (vs 1.54 mean) Growth: › Sales 2026-2028: $254B -> $360B › EPS 2026-2028: $31.06 -> $41.63 › EPS Fwd 5Y CAGR: 21.0% Capital Structure: › Total Debt: $112.3B › Cash & Inv.: $90.3B Dan from @longtermpick