Daryl (@darylspelman) header
Daryl (@darylspelman) profile picture

Hunting for exceptional companies worth owning • 20 years across buy-side & sell-side • Mainly US, China & Japan • Co-founder @TopPercentileHQ

Hong Kong
Joined October 2010
Photo by Daryl (@darylspelman)Photo by Daryl (@darylspelman)Photo by Daryl (@darylspelman)Photo by Daryl (@darylspelman)Photo by Daryl (@darylspelman)Photo by Daryl (@darylspelman)Photo by Daryl (@darylspelman)Photo by Daryl (@darylspelman)Photo by Daryl (@darylspelman)Photo by Daryl (@darylspelman)
Replying to @oguzerkan
Where I agree - $MSFT software is too vital to enterprises and too embedded to be disrupted - OpenAI exposure in Azure is a concern, but OpenAI isn’t going anywhere and AI compute demand is only going higher Where I disagree - Valuation is not fair value. You’re able to buy one of the best companies ever built for a modest fwd p/e of 22x Sometimes the markets are irrational and give you bargains. I am adding here
6
65
5,691
“From a business perspective, it’s business as usual.” Colette Kress, $NVDA earnings call KEY TAKEAWAYS FROM THE CALL ◆ FY28 growth outlook raised. Nvidia now expects 75% revenue growth, up from 70%, on a higher FY27 base. Demand still exceeds supply, but additional capacity supports the increased outlook. ◆ Vera Rubin’s record ramp. Nearly $20B in revenue during its first quarter of production shipments. Major hyperscalers are bringing systems into production in roughly half the time taken with Grace Blackwell. ◆ AI demand is broadening. Agentic AI is driving strong CPU demand. Nvidia expects Vera CPU revenue to more than double in FY28, alongside continued growth in enterprise and sovereign AI. ◆ Financial commitments are growing. Supply commitments reached $390B and guarantees $118B. Management emphasised greater selectivity in customer financing and future commitments. ◆ Leadership transition. The board confirmed a permanent CEO search covering internal and external candidates, ruled out reinstating Huang and established a National Security and Export Control Oversight Committee. WHAT ANALYSTS ASKED Joseph Moore, Morgan Stanley Q: What supports the higher FY28 growth outlook, and does demand still exceed available supply? A: Both demand and achievable supply have improved. Management remains confident in 75% growth despite the new export controls. Vivek Arya, Bank of America Q: Can Nvidia protect its competitive position and retain talent without Huang? A: Kress highlighted employee commitment, strong technical leadership and customer and supplier support. The board will conduct a thorough CEO search without rushing the appointment. Timothy Arcuri, UBS Q: Can Nvidia protect pricing power and margins as customers develop custom AI chips? A: Management reiterated its 72–73% FY28 gross-margin outlook, supported by pricing adjustments. It gave no longer-term margin target. C.J. Muse, Cantor Fitzgerald Q: Will Nvidia become more selective about financing neocloud customers? A: Yes, future support will be assessed carefully, particularly financing for land, power and data-centre infrastructure. External financing partners will play an important role. Stacy Rasgon, Bernstein Q: Are the new financing conditions in neocloud agreements a change in policy? A: No. Kress said the disclosure clarifies existing arrangements. Commitments depend on customers securing financing and getting infrastructure operational. Ben Reitzes, Melius Research Q: Could changing government policies cause an AI investment slowdown in 2028? A: Management continues to see strong demand and remains engaged with government, but offered no specific assurances about future policy. Joshua Buchalter, TD Cowen Q: How does Nvidia allocate scarce supply among hyperscalers, AI labs and neoclouds? A: Priority depends on data-centre readiness, available financing and credible customer demand. Management said pricing is consistent across customers. Jim Schneider, Goldman Sachs Q: How will Nvidia balance buybacks, supplier investment and customer financing? A: Buybacks remain a priority, alongside financing additional supplier capacity. Customer guarantees will remain selective. REMAINING QUESTION ◆ Can customer financing and data-centre construction keep pace with Nvidia’s expanded supply commitments and ambitious FY28 growth outlook?
$NVDA is up 2.5% after another huge earnings beat. Revenue growth remains extraordinary, with Vera Rubin ramping rapidly, although memory costs continue to pressure margins. HEADLINE PRINT 🟢 Revenue: $115.8B vs $109.3B expected (+103% YoY) 🟢 Adj. EPS: $2.68 vs $2.47 expected WHAT CHANGED 🟢 Data Center: $107.6B vs $101.9B expected (+110% YoY) 🟢 Q4 revenue guidance: $130B vs $124.4B expected 🟡 Q4 gross margin guidance: 71–72%, down from 74.9% in Q3 WHY THIS EARNINGS MATTERS ◆ AI demand remains exceptionally strong. Vera Rubin NVL72 has started shipping, and Nvidia is guiding for another substantial increase in revenue. The business continues to deliver extraordinary growth despite supply constraints and rising memory costs. The expected margin decline remains the main financial pressure. REMAINING QUESTION ◆ Can Nvidia sustain the Vera Rubin ramp and its technology leadership through the transition away from Jensen Huang?
7
3
15
941
Daryl retweeted
Book #28: Beating the Street “When insiders are buying, it’s a good sign, unless they happen to be New England bankers.” - Peter Lynch Insider buying is often reassuring. Who should understand a business better than the people running it? But Lynch’s experience with troubled New England banks offers an important caution. Some banks appeared remarkably cheap relative to their book values. Yet beneath those valuations were deteriorating loan portfolios, with potential losses large enough to destroy much of their equity. His examples illustrate how confidence in a business can coexist with serious problems that even those closest to it may misjudge. Insiders have access to information that outside investors don’t. They can also be overly optimistic about the value of their assets or the prospects for recovery. Insider buying tells us what management believes. It doesn’t necessarily tell us whether management is right. One idea from a great investing book each week. Book #29 next Sunday.
Book #27: Investing Between the Lines “Investors who find a fogged-over communication need to ask: does the CEO not understand the business, or does he or she not want owners to understand it?” - L.J. Rittenhouse In Enron’s 2000 shareholder letter, management praised its talented people, global presence, financial strength and competitive advantages. Impressive words. But how much did they actually tell investors? Rittenhouse calls this FOG: Fact-deficient, Obfuscating Generalities. Corporate language that sounds reassuring but offers little evidence about the underlying business. Just months after publishing that letter, Enron collapsed. Vague language alone doesn’t prove dishonesty. But it can make important problems harder to see. When reading shareholder letters and listening to earnings calls, it’s worth looking beyond the confident language. What specific facts support management’s claims? And what are they avoiding explaining? One idea from a great investing book each week. Book #28 next Sunday.
6
6
2
36
10,642
Daryl retweeted
What a waste of time preparing an October market report! November made it look old in four days. An election, then yesterday’s market rout. But October was already showing where the cracks were forming. The Nasdaq 100 fell 6.7% as questions around AI spending, debt and returns started to bite. So perhaps it is still worth knowing how we got here. Final plug for the October report. Link in the reply 👇
3
2
14
777
Daryl retweeted
Book #27: Investing Between the Lines “Investors who find a fogged-over communication need to ask: does the CEO not understand the business, or does he or she not want owners to understand it?” - L.J. Rittenhouse In Enron’s 2000 shareholder letter, management praised its talented people, global presence, financial strength and competitive advantages. Impressive words. But how much did they actually tell investors? Rittenhouse calls this FOG: Fact-deficient, Obfuscating Generalities. Corporate language that sounds reassuring but offers little evidence about the underlying business. Just months after publishing that letter, Enron collapsed. Vague language alone doesn’t prove dishonesty. But it can make important problems harder to see. When reading shareholder letters and listening to earnings calls, it’s worth looking beyond the confident language. What specific facts support management’s claims? And what are they avoiding explaining? One idea from a great investing book each week. Book #28 next Sunday.
Book #26: The Most Dangerous Trade by Richard Teitelbaum “I will follow this trade to the end of the earth.” - Bill Ackman Ackman made that vow while betting roughly $1 billion that Herbalife would collapse. He spent years investigating its business model and battling Carl Icahn, who backed the company. In 2016, Herbalife agreed to pay $200 million and overhaul its US distributor practices to settle FTC charges. Ackman’s concerns weren’t imaginary. But Herbalife survived, and its shares rose. Ackman lost money on the bet. He replaced his short with put options in 2017 to limit further losses, before exiting entirely in 2018. The case captures a difficult part of investing: identifying genuine problems in a business doesn’t guarantee you’ll profit from them. The outcome, timing and structure of a position can matter just as much as the original research. One idea from a great investing book each week. Book #27 next Sunday.
6
18
2
200
7,622
Our October Monthly Market Review is now live. AI financing fears dominated the month, with growing concerns about debt, cash burn and returns on infrastructure spending. Chip stocks came under pressure, while consumer and energy stocks found buyers. Even record memory profits weren’t enough to calm investors. See what moved markets, the key investment themes and the companies that stood out. Full report in the reply 👇
1
1
2
173
October turned the market on its head. Our October Monthly Market Review is now live. AI financing fears hit chips and speculative growth, while consumer and energy stocks found buyers. Even record memory profits couldn’t calm the tech sell-off, and the Fed paused with further hikes still in view. Full report in the reply 👇
8
2
27
1,730
The risks around AI keep expanding. Extinction risk. Huge capex. OpenAI’s $20bn revenue gap. AI debt. IPO valuations… At the same time, AI usage keeps growing. It is becoming embedded in more products and business workflows, while the infrastructure required to support it keeps expanding. So how do we navigate it as investors? First, understand who does what. Equipment enables the fabs. Fabs manufacture the chips. Memory, networking, optics and power then help turn that compute into the AI products we ultimately use. One thing I find especially useful on the map is scarcity. Where does growing demand depend on only a handful of suppliers or a genuinely constrained layer? HBM is a good example. SK hynix, Samsung and Micron dominate supply, which matters when capacity is tight. But memory is still cyclical. New capacity and Chinese competition matter too. Scarcity is valuable precisely because you need to understand what could eventually relieve it. That same lens can be applied across the stack: where is demand running into limited supply, and who controls that capacity? This week we get another useful read from the foundation. ASML and TSMC report, showing how AI demand is flowing through to the equipment and manufacturing layers underneath it all.
5
4
1
23
1,646
Daryl retweeted
Something important got buried between Micron and Nike earnings. OUR SEPTEMBER REPORT IS OUT. Inside: - What moved markets, sectors and industries in September - S&P 500 monthly statistics and market breadth - Themes from AI hardware and China tech to private credit, oil services, defence and more - The biggest post-earnings winners and losers - Which large-cap businesses improved most in quality 👇
Last trading day of September. AI safety fears, Meta’s Muse, oil disruption and rising US rates all left their mark. Our September Monthly Market Review is coming shortly. Keep an eye out for it.
2
1
13
1,271
Our September Monthly Market Review is now live. AI safety fears grabbed the headlines, but rising US rates, $110 oil and a rebound in AI hardware all shaped the month. See what moved markets, the key investment themes and the companies that stood out. Full report in the reply 👇
1
1
5
312
September was a month of extremes. Warnings of AI-driven extinction risk, 2.8m Muse downloads, a 5% 10-year yield, the Fed’s first rate hike since 2023 and $107 oil. Our September Monthly Market Review is coming shortly after the market closes today. Keep an eye out for it.
1
1
10
442
Daryl retweeted
Our September Monthly Market Review is now live. September was green where AI hardware mattered most, and red almost everywhere else. Nasdaq, KOSPI and Nikkei all gained, while most major markets fell. Full report link in the reply 👇
3
1
11
720
$MU Micron earnings tomorrow night. AI demand has driven one of the strongest memory upcycles in its history. Now the test is how durable it can be. Stock: up 273% YTD | up 14% over 1 month | up 15.4% at last earnings The revenue chart captures the setup. The current surge is far larger than previous upcycles, just as investors are trying to work out whether AI has changed the shape of the memory cycle. WHAT HAPPENED AT LAST EARNINGS ◆ Micron said DRAM and NAND supply-demand would remain tight beyond calendar 2027. WHAT HAPPENED LATELY ◆ Samsung said HBM could take nearly 30% of industry DRAM wafer capacity in 2027, up from around 20% now, which could keep conventional DRAM supply constrained. At the same time, China’s CXMT has moved its latest DRAM platform into mass production, adding a potential source of new supply. WHAT MATTERS THIS EARNINGS ◆ Whether FY27 pricing and margins show the memory cycle is still strengthening rather than starting to turn. ◆ Whether HBM4 and multi-year customer agreements are making this cycle more durable than previous ones.
6
1
1
15
1,390
$CCL is up 11% after earnings. Pricing held up far better than expected despite higher fuel costs. HEADLINE PRINT 🟢 Adj. EPS: $1.43 vs ~$1.35 prior guidance 🟢 Revenue: $8.4B, up 3% YoY WHAT CHANGED 🟢 Net yields: +2.4% vs ~1.2% prior guidance 🟢 2027 bookings: record occupancy and pricing WHY THIS EARNINGS MATTERS ◆ The pricing concern eased considerably. Carnival doubled its Q3 yield guidance, while 2027 bookings remain at record occupancy and pricing. Higher fuel costs are still a drag, but demand is holding up.
$CCL earnings tomorrow morning. Carnival has become a clean test of how durable cruise pricing really is as costs rise. Stock: down 27.5% YTD | down 10.6% over 1 month | down 4.2% at last earnings The selloff has changed the setup. CCL still screens as one of the stronger businesses in Travel Services, but its valuation is now among the most attractive in the group. WHAT HAPPENED AT LAST EARNINGS ◆ Carnival posted another record quarter for net yields and was already 93% booked for 2026 at historically high prices. Full-year constant-currency net-yield guidance fell from 2.75% to 1.75% as geopolitical disruption weighed on the outlook. WHAT HAPPENED LATELY ◆ Higher fuel costs and weaker close-in Caribbean pricing have become the main concerns heading into this report. WHAT MATTERS THIS EARNINGS ◆ Whether recent Caribbean pricing pressure is temporary or the first sign Carnival’s yield momentum is starting to slow. ◆ The extent to which higher fuel costs start eating into earnings if demand and pricing remain firm.
1
1
11
659
$CCL earnings tomorrow morning. Carnival has become a clean test of how durable cruise pricing really is as costs rise. Stock: down 27.5% YTD | down 10.6% over 1 month | down 4.2% at last earnings The selloff has changed the setup. CCL still screens as one of the stronger businesses in Travel Services, but its valuation is now among the most attractive in the group. WHAT HAPPENED AT LAST EARNINGS ◆ Carnival posted another record quarter for net yields and was already 93% booked for 2026 at historically high prices. Full-year constant-currency net-yield guidance fell from 2.75% to 1.75% as geopolitical disruption weighed on the outlook. WHAT HAPPENED LATELY ◆ Higher fuel costs and weaker close-in Caribbean pricing have become the main concerns heading into this report. WHAT MATTERS THIS EARNINGS ◆ Whether recent Caribbean pricing pressure is temporary or the first sign Carnival’s yield momentum is starting to slow. ◆ The extent to which higher fuel costs start eating into earnings if demand and pricing remain firm.
4
1
1
11
1,327
I never chase dividends at the expense of compounding. A 5% yield can be expensive if it means giving up a business capable of growing much faster. My rough rule is to look for companies where earnings can double every five years. That’s about 15% annual growth. Take Microsoft. Revenue has compounded 14.6% a year over five years and ROE is 33.2%. It pays a dividend too, but that’s hardly why I’d own it. Buffett put the capital-allocation case well: “For every dollar retained by the corporation, at least one dollar of [additional] market value will be created for owners.” There is another side though. For investors living off a portfolio, there is real psychological comfort in cash arriving without having to decide what to sell. Vanguard founder John Bogle: “Bet on the dividends, and not on the market price.” I’m in the compounding-first camp. How do you think about it?
4
1
14
489
Earnings season is starting to pick up, with a couple of major names on the calendar this week. I’ll focus on the memory cycle, a major consumer name still trying to turn itself around, and cruise pricing. Here are the 3 earnings I’m watching this week: $CCL $MU $NKE $CCL (Tuesday, before open): Carnival is still booking well ahead at historically high prices, but concerns around Caribbean pricing and higher fuel costs have put yield growth back in focus. Last earnings reaction: down 4.2% ◆ Question: Are the recent pricing concerns temporary, or is Carnival’s yield momentum starting to slow? $MU (Wednesday, after close): Memory pricing and margins have surged as AI demand collides with constrained supply, with Micron now entering earnings around record profitability. Last earnings reaction: up 15.4% ◆ Question: Is there any sign the memory cycle is starting to turn as Micron heads into FY27? $NKE (Thursday, after close): Nike is still in the middle of a difficult turnaround. Performance products and wholesale are improving, but sportswear and Greater China remain major weak spots. Last earnings reaction: up 3.8% ◆ Question: Can the recovery broaden enough to offset the continued weakness in China and sportswear? Every earnings report reshapes the Top Percentile rankings
1
11
515
Emma is one of my favorite follows on X and I’m hoping to interview her and her husband for the podcast in the coming months! Top Percentile is also among my top investing websites bookmarked along with Stock Analysis, Quartr, Investors Grow, Fiscal.ai, Insider Flow and others.
I’m not here to tell you which stock to buy. I think I have something more useful to share. This is a quick recap on what happens on this account and what I do. I occasionally share what I have in my portfolio, but that’s hardly the point. Most of what I share here starts with a tool we built called Top Percentile. It reads every single SEC filing and converts the data into a quality score for every company, every day. It then ranks them, telling you who is #1, #2, #3... all the way through the market. $TSM currently holds the #1 spot, after taking over from $NVDA during the last earnings season. That’s our superpower. Why should you care? Because it helps you understand exactly what you are buying, including its strengths and weaknesses, and more importantly helps you see the whole board and who actually stands out. This is where some of our best investments have come from. So, whether you come here with a stock idea or find one through the data, we’re here to help you evaluate it using our unique dataset and build your own conviction. You still need to do a bit of work and ask the questions, but we’ve done the hard part for you. The site is there for you to explore. And this is what I do on this account: Post about earnings because ultimately they drive the site’s quality scores. But I focus specifically on why the earnings matter, with a relevant chart. Walk you through the recent tensions around a company and what I think is worth looking at. Share comprehensive monthly market reviews to get you up to speed with what happened, why it happened and hopefully improve your understanding of the market along the way. Build visual industry stacks. AI and every branch of it, humanoids, space and more. Mostly to give you a simple perspective on who does what and where everyone sits. Share a glimpse of my investing library every Sunday. It also gives you some sense of what went into building Top Percentile. Give you brief weekly updates, occasional thoughts on market news, rare introductions like this one and a tiny bit of my personal life. Finally, I always love questions, so ask me anything. I don’t know every stock, but our site often does. And soon, we’re going global. Ah, and everything on the site is free for individual investors. Link in bio. We have separate paid B2B services for funds. Happy weekend everyone.
6
2
14
1,640
I’m not here to tell you which stock to buy. I think I have something more useful to share. This is a quick recap on what happens on this account and what I do. I occasionally share what I have in my portfolio, but that’s hardly the point. Most of what I share here starts with a tool we built called Top Percentile. It reads every single SEC filing and converts the data into a quality score for every company, every day. It then ranks them, telling you who is #1, #2, #3... all the way through the market. $TSM currently holds the #1 spot, after taking over from $NVDA during the last earnings season. That’s our superpower. Why should you care? Because it helps you understand exactly what you are buying, including its strengths and weaknesses, and more importantly helps you see the whole board and who actually stands out. This is where some of our best investments have come from. So, whether you come here with a stock idea or find one through the data, we’re here to help you evaluate it using our unique dataset and build your own conviction. You still need to do a bit of work and ask the questions, but we’ve done the hard part for you. The site is there for you to explore. And this is what I do on this account: Post about earnings because ultimately they drive the site’s quality scores. But I focus specifically on why the earnings matter, with a relevant chart. Walk you through the recent tensions around a company and what I think is worth looking at. Share comprehensive monthly market reviews to get you up to speed with what happened, why it happened and hopefully improve your understanding of the market along the way. Build visual industry stacks. AI and every branch of it, humanoids, space and more. Mostly to give you a simple perspective on who does what and where everyone sits. Share a glimpse of my investing library every Sunday. It also gives you some sense of what went into building Top Percentile. Give you brief weekly updates, occasional thoughts on market news, rare introductions like this one and a tiny bit of my personal life. Finally, I always love questions, so ask me anything. I don’t know every stock, but our site often does. And soon, we’re going global. Ah, and everything on the site is free for individual investors. Link in bio. We have separate paid B2B services for funds. Happy weekend everyone.
5
2
3
23
3,688
There is something special about these 12 companies. They are heavily out of favour, but still look surprisingly attractive on business quality and valuation. That gap is often where emotions take over, and this may create some of the best opportunities. When fear takes hold around a company, the story can become more powerful than the numbers. Once we believe it, we start interpreting the facts through it. ServiceNow is a good example. The stock fell roughly 65% from its high as AI disruption fears swept through software. We track companies every day on two things: how strong the fundamentals are, and how attractive the valuation is. When both are green, that is where I start looking for quality on sale. Then comes the harder part: go through them one by one and ask what the fear actually means for the business. Is the future genuinely in danger? You may look at some of these names and think: surely these companies can’t still be that good. But that reaction is exactly the point. Sometimes the story around a stock starts shaping our judgement before we even look at the facts. $NVO: Lilly took the lead. Pricing is being squeezed and investors are already looking ahead to semaglutide patent expiry in the early 2030s. $TCOM: Beijing hit the hotel lock-in. A $770m antitrust penalty forces Trip. com to stop exclusive hotel arrangements and gives rivals an opening. $DECK: HOKA is growing up. Growth has slowed sharply from its exceptional years, while consumers remain cautious. $ADBE: AI threatens the moat. Figma, Canva and generative tools make investors question whether Adobe stays indispensable. $TTD: The perfect-execution story broke. Growth slowed and management admits it needs to fix execution and upgrade the platform. $UPWK: AI can replace the work. Lower-complexity freelancing is already shifting toward automation. $TME: No obvious headline fear. Revenue and profit are still growing, although social-entertainment revenue continues to shrink. $LULU: Lost its cool. Leggings fell ~20%, Americas sales are shrinking, and the new CEO inherits a brand still searching for a fix. $SFM: The growth party ended. Comparable sales turned negative just as a CEO handoff begins. $YELP: AI wants the front door to local search. Yelp has to prove it still owns the click, not just the data behind it. $PDD: The cheap-parcel model got hit. Customs advantages are disappearing while competition at home keeps profits under pressure. $INFY: AI eats the staffing model. Clients want more for less, automation threatens billable hours, and growth expectations have come down. Pick the names that surprise you. Read the bull and bear cases. Decide which fears are temporary, which are structural, and where the market may have gone too far. Bookmark it. You will learn just by looking back at these in a few months.
2
2
15
705
Good old $NKE has two very different challengers: $DECK and $ONON. All three are more than 50% below their highs. Nike has the scale: 8x Deckers and 12x On by revenue. Deckers has the economics: nearly 3x Nike’s operating margin. On has the growth: 18% vs Nike essentially flat, and more than 2x Deckers. Which one would you rather own from here?
5
2
11
861