50 new pitches found in hedge fund reports this week

Thanks for reading Stock Analysis Compilation! Subscribe for free to receive new posts and support my work. 🆕 New on the site: every chart fund managers publish, on one page The best charts in fund letters and research are usually buried deep inside a PDF. We pull them out and put them side by side — macro, rates, earnings, valuations — each one credited to the firm that made it, with the original one click away. Free, no account needed. 👉 Browse the charts → 🔹 Alphabet (GOOGL US) by Bronte Capital Amalthea Fund 🔹 AMD (AMD US) by St. James Investment Company Value Investor’s 🔹 Astera Labs (ALAB US) by Clearbridge Investments Growth Strategy 🔹 Bentley Systems (BSY US) by Heartland Advisors Mid Cap Value Fund 🔹 Builders FirstSource, Inc. (BLDR US) by Forager Funds (Australian Shares Fund and International Shares Fund) 🔹 Cameco Corporation (CCJ CN) by Hardman Johnston Global Equity 🔹 Cavco (CVCO US) by Baillie Gifford Positive Change Strategy 🔹 Chien Kuo Construction by Stone Sentinel Capital 🔹 China Resources Beer (291 HK) by Liontrust Global Income & Growth Fund 🔹 Compagnie de l’Odet (ODET FP) by Horos AM Funds 🔹 Corebridge Financial (CRBG US) by Oakmark Fund 🔹 Hillman Solutions (HLMN US) by Heartland Value Fund 🔹 Idun Industrier (IDUN ST) by Kavaljer Quality Focus 🔹 Intel (INTC US) by Alpha Wealth Fund The Insiders Fund 🔹 Intuit (INTU US) by Lindsell Train Global Equity Fund 🔹 Kiniksa Pharmaceuticals (KNSA US) by ClearBridge Investments SMID Cap Growth Strategy 🔹 KKR (KKR US) by Lindsell Train Global Equity Fund 🔹 KLA Corporation (KLAC US) by Baillie Gifford US Equity Growth 🔹 MediaTek (2454 TT) by Artisan Emerging Markets Fund 🔹 Meta Platforms, Inc. (META US) by Jensen Investment Management Quality Growth Strategy 🔹 Microsoft (MSFT US) by Clearbridge Investments ESG Investment 🔹 Microsoft (MSFT US) by Mar Vista U.S. Quality Premier 🔹 Mobility Global by Middle Coast Investing Equity Portfolios 🔹 Nan Ya Plastics (1303 TT) by Artisan Emerging Markets Fund 🔹 NVIDIA (NVDA US) by Polen Focus Growth 🔹 OpenText (OTEX CN) by Oakmark International Fund (OAKIX) 🔹 Paycom Software (PAYC US) by Spheria Global Smaller Companies Fund 🔹 Polaris (PII US) by Bristlemoon Capital Global Fund 🔹 Quaker Houghton (KWR US) by Heartland Value Plus Fund 🔹 Recruit Holdings (6098 JP) by Baillie Gifford International Alpha 🔹 Rightmove (RMV LN) by WS Lindsell Train UK Equity Fund 🔹 SAP (SAP GR) by Clearbridge Investments International Growth EAFE Strategy 🔹 Saturn Oil & Gas by Bison Interests Energy Opportunity Fund, LP 🔹 SCIH Salt Holdings (MORTON US) by Malborough Marlborough Global High Yield Fund 🔹 Shenzhen Inovance (300124 CH) by Artisan Emerging Markets Fund 🔹 Shin-Etsu Chemical Co., Ltd. (4063 JP) by Hennessy Japan Fund 🔹 Shopify (SHOP US) by Baillie Gifford International Concentrated Growth 🔹 Shopify Inc. (SHOP US) by Baillie Gifford US Equity Growth Strategy 🔹 SK Hynix (000660 KS) by Artisan Emerging Markets Fund 🔹 SKF (SKF B ST) by Liontrust Global Income & Growth Fund 🔹 Snowflake Inc. (SNOW US) by Baillie Gifford US Equity Growth Strategy 🔹 The Hanover Insurance Group, Inc. (THG US) by Heartland Value Plus Fund 🔹 Theon International (THEON NA) by Amati Investment Funds 🔹 Tianqi Lithium (9696 HK) by ACATIS 🔹 TripAdvisor (TRIP US) by Middle Coast Investing Equity Portfolios 🔹 Universal Music Group (UMG NA) by Lindsell Train Global Equity Fund 🔹 Visa Inc (V US) by Artisan Partners Developing World Strategy 🔹 Watsco (WSO US) by Heartland Advisors Mid Cap Value Fund 🔹 Wolters Kluwer (WKL NA) by Malborough European Special Situations 🔹 Zhongji Innolight (300308 CN) by ClearBridge Investments Emerging Markets Strategy Alphabet ($GOOGL US) Fund: Bronte Capital Amalthea Fund Thesis: Alphabet is seen as a highly capable company pivoting aggressively into AI, but with an existential threat to its search business. Source: Read the original letter ↗ Analysis: We will start with the stock that worries us most and is nearest the centre of the heat: Alphabet/Google. To explain the Google’s challenge, we offer a cautionary tale about management failures that destroyed a pillar of the Australian financial system — and explain how Google has so far avoided that trap. Google has a search business that produces vast profits by displaying ads. It is threatened by AI. There is a possible future emerging where people query computers via AI, bypassing the search engine. There is a bleaker (for Google) future where people interact with computers as you might in Star Trek: The Next Generation: you ask the computer something with your voice, and the computer gives a nuanced answer and acts as your agent. In a voice-controlled AI world there is no space for advertisements — and no place for traditional Google Search. Google was not going to sit on their thumbs and milk their old business until it died like AMP did. Google leant into the disruption. If someone is going to disrupt your business it might as well be you. It is hard to express just how forward-thinking Google’s decisions were. But let’s lay a few out: In voice search In 2007, Google launched a toll-free number you could call that allowed them to start building a human voice search database for training software. In 2008 they launched actual voice search, first for the iPhone and later Android In 2012 they introduced voice features to partially control android phones (things like settling alarms) In AI hardware Google started chip design as early as 2007 By 2016 they had designed and built working tensor processing units (TPUs) that made them a genuine competitor to Nvidia. By 2021 they had built large numbers of TPUs that they could start putting in data centres. Today, Google is selling TPU capacity to Anthropic. The latest generation of TPUs has two versions, one optimised for “training” AI models and the other for using AI models (“inference”). In AI models and software In January 2014 (more than twelve years ago) Google acquired DeepMind. This was really a deal to acquire a team: Demis Hassabis, Shane Legg, and Mustafa Suleyman and about 50 to 75 researchers specializing in machine learning. Hassabis remains the key developer of AI at Alphabet. Shane Legg remains at Alphabet, but Mustafa Suleyman is the head of AI at Microsoft. Alas a few other key staff have defected to Anthropic. In 2016 DeepMind took on Lee Sedol, the world’s best Go player. No computer before DeepMind played Go to a level beyond a reasonable club player. This was an astonishing feat. People expected it would be decades before Go computers got to “9 Dan” level. There is a wonderful documentary on this which you can find on YouTube. Much of what became commonplace knowledge about AI is hinted at in that documentary, which John watched at the time — though he wasn’t smart enough to recognise its significance. This documentary skirts anthropomorphizing the AI. Lee Sedol concludes that the program is “creative”. In late 2017, almost as an afterthought, DeepMind released its famous chess engine, AlphaZero. This was a simple AI program that trained itself by playing millions of games against itself for just four hours, learning what does and does not work in chess. Alphazero crushed Stockfish - the seemingly unbeatable and dominant chess engine of the time. The phrase “Stockfish resigns” was uttered with amazement. AlphaZero played chess differently to other chess engines: it was aggressive, and the moves looked more “human” than the (dull) positional play of other chess engines. These games are fun to watch. In the words of Gotham Chess, “AlphaZero will sacrifice material for time, it will sacrifice material for damage, and it will sacrifice material to take your soul”. Google also purchased a substantial US based AI Lab which eventually became Google Brain. This was merged with DeepMind. This produced the seminal Transformers idea (2017) which massively improved AI language models. There are now many Google AI products, led by Gemini and NotebookLM. If you have not used any, just drop a large number of complicated documents in NotebookLM and ask it to create a podcast explaining the issues to you. It is very good. For instance, it will give you a decent explanation in English of Japanese technical manuals. More to the point, Google’s stated goal from inception was to “organize the world’s information and make it universally accessible and useful”. Now their stated goal is to take that information and make the world’s greatest superintelligence. This is an amazing pivot. It is the opposite of what AMP did. We do not think we have ever seen a better managed company. But the existential threat is still there Google has dealt with the existential threat brilliantly, by pivoting to meet it head-on. Alas, the threat remains. John’s son is a computer science graduate and more sophisticated than the average person. He hardly uses Google search any more. His default search is Claude, and a typical question might be to work out different ways of getting from Lyon to Geneva (on holidays) when he needs to be there around 6pm tomorrow. This is much faster than Google search, produces a nuanced answer (giving options at various price points), and denies Google a high-value revenue search. How much is Google spending? Google’s spending is truly breathtaking: they announced $180-190 billion for 2026 and guided for a substantial increase in 2027. The 2026 spend exceeds their cumulative profits over the entire 2010s. Their total AI spend will be larger than their cumulative profits throughout history. The whole application of Google’s achievements, at least in a financial sense, has been to fund a push into AI. Google has even raised cash in 2026. We repeat: one of the most cash-generative businesses in history has had to raise cash. What do we think about this? A summary This presents us with a difficult decision. We love Google as a company and think its management team may be amongst the best assembled in the history of capitalism. But the stock is abnormally risky. The core business is being disrupted, and that disruption may be total. The company’s reaction is to spend its entire accumulated profits on disrupting itself with no certainty about the outcome, or even whether there is any profit at the end. We keep trimming the stock. And we know in the end we will be squeamish about the outcome. At the one end Google is the provider of the world’s finest superintelligence, and is the central company in displacing a good slice of the world’s labor (while privatising the profits). At the other end, the company will have its core business destroyed and will have wasted all of the money it made during the good years. We own it. We genuinely have no idea what it is worth. If somebody tells you that they do, they are pretending to know more about the future than they do about the past or present. Access our full research database on Alphabet AMD ($AMD US) Fund: St. James Investment Company Value Investor’s Thesis: AMD gained share by using a fabless model and efficient Zen chips to challenge Intel’s former moat. Source: Read the original letter ↗ Analysis: AMD’s answer, beginning in 2008, was to go “fabless” and later stake everything on a clean-sheet architecture (Zen) manufactured at TSMC’s facilities. Because AMD could not match Intel’s capital spending, it spun off its factories and outsourced its chip production to TSMC. This innovation removed the requirement that a challenger be a world-class manufacturer while TSMC’s rise made the fabless chip company possible. A competitor no longer had to control the entire channel; it could buy access to the best process while Intel shouldered the cost of the entire production value chain. [1] Intel planned to launch its first 10-nanometer (nm) CPUs in 2016, but its development never arrived on schedule. The delay pushed Intel away from its long-running Tick-Tock cadence (a two-year process-development cycle), kept the company at 14 nm for years, and helped TSMC and AMD narrow the manufacturing gap. While TSMC shipped 7 nm and then 5 nm, Intel spent years refreshing its 14 nm products. By the time “Intel 7” (a rebranded 10 nm process) reached critical production volume, its competitors had already closed the performance and efficiency gap. Intel assumed its largest customers would never leave the x86 island, but its dominant moat no longer existed. Apple was once Intel’s most prestigious PC customer, yet beginning in 2020, it transitioned the entire Mac line to its own ARM-based M-series chips. Hyperscalers, a market segment Intel once completely dominated, began dual-sourcing AMD’s EPYC processors and later designed their own silicon. Microsoft promoted Windows on ARM, while OEMs that had treated Intel as the default began viewing it as just one option. The moat that had once protected Intel’s business quickly disappeared. x86 compatibility had once been a genuine barrier to competition: rewriting the world’s software was unthinkable while Intel chips kept getting faster. Once process leadership vanished and new platforms (phones, then Macs, then parts of the cloud) no longer required x86, Intel’s structural moats began to erode. Compilers, emulation, and new operating-system support removed the assumption that “everything important runs on Intel.” AMD did not need to destroy x86; it only needed to make “good enough, cheaper, better efficiency” available on the same instruction set, and then let customers decide. Access our full research database on AMD Astera Labs ($ALAB US) Fund: Clearbridge Investments Growth Strategy Thesis: Astera Labs provides high-speed connectivity for AI data centers and could benefit from a broader AI connectivity platform and Amazon’s Tranium ramp. Source: Read the original letter ↗ Analysis: We also initiated and built a position in Astera Labs, which provides high-speed connectivity solutions that help enable data to move quickly and reliably between GPUs, CPUs, memory, storage and networking components within AI data centers. Astera is evolving into a broader AI connectivity platform and could be a beneficiary of Amazon’s ramp of its Tranium AI accelerator chips. Together, these positions diversify our AI exposure as AI value capture broadens beyond logic chips. Access our full research database on Astera Labs 🔓 Unlock the full research database This week’s ideas are a sample. The real edge is the full searchable archive : The Associate ($19/mo) gives you every pitch we’ve indexed — 3,000+ across 300+ funds — searchable by fund, sector or ticker, plus the entire quarterly-letter archive. The weekend of PDF-hunting, already done for you. The Rainmaker ($29/mo) adds Warren AI to ask questions across the whole corpus in plain English. Start your 7-day free trial → Bentley Systems ($BSY US) Fund: Heartland Advisors Mid Cap Value Fund Thesis: Bentley Systems provides critical infrastructure engineering software and is viewed as having a strong moat, high switching costs, and long-term sales and free cash flow upside. Source: Read the original letter ↗ Analysis: Technology. Bentley Systems (BSY) , a newly initiated position during the quarter, is an example of a stock getting heavily discounted by a negative top-down industry narrative, with defensible and compelling long-term prospects. To be clear, we are wary of the outlook for many software products, especially those that serve administrative functions and are prone to faster switching decisions in a widening competitive landscape. Our research has centered on “vertical software” companies that provide end-to-end solutions for a specific industry because domain expertise drives higher customer loyalty and switching costs. Among the vertical software companies in our universe, we find BSY’s characteristics especially attractive from both a “moat” and opportunity perspective. The company provides engineering software for critical infrastructure around the world for public works and natural resources end markets. Over 40 years of proprietary data and deep integration has made BSY the industry standard for infrastructure modeling and simulation. Their software represents roughly five basis points (0.05%) of total project cost for mission critical infrastructure, with BSY’s engineering customers facing severe engineering labor challenges and their asset owner customers relying on BSY’s software to ensure reliability and minimize expenses while avoiding a high cost of failure. With end market demand buoyed by a need to address structural underinvestment, we expect BSY sales and free cash flow to accelerate further as customers consume more of their offerings for AI-oriented workloads. Access our full research database on Bentley Systems Builders FirstSource, Inc. ($BLDR US) Fund: Forager Funds (Australian Shares Fund and International Shares Fund) Thesis: Builders FirstSource distributes building materials to US home builders and is tied to the housing cycle and interest rates. Source: Read the original letter ↗ Analysis: Getting out of jail cost money at Builders FirstSource (NYSE:BLDR), the largest distributor of building materials, roof frames, windows and lumber to US home builders. It was the Fund’s biggest detractor this quarter. By September, the Fund had sold its remaining shares. When the Fund first invested in July 2025, interest rates looked to be peaking, and lower rates would mean more houses being built. Early this year, that seemed to be playing out. US mortgage rates fell towards 6%, and most of the Fund’s shares were bought in the months that followed. And then interest rates turned aggressively upwards. With US 30-year mortgage rates now well above 7%, many fewer new homes will be built. The company’s outlook got worse each quarter. Operating leverage means it is earning much less profit on each sale than our original expectations, its leverage has increased relative to earnings, and it has stopped buying back its own shares. Builders FirstSource is well placed to forge good returns over the full cycle. The shares are worth revisiting at giveaway prices or following compelling signs that the US housing market is turning. Both feel a long way off from here. Access our full research database on Builders FirstSource, Inc. Cameco Corporation ($CCJ CN) Fund: Hardman Johnston Global Equity Thesis: Cameco Corporation is positioned to benefit from new demand for upstream nuclear fuel driven by AI data-center needs and energy diversification. Source: Read the original letter ↗ Analysis: Flowing from the AI buildout, as well as the impact of unresolved conflict in the Middle East, are secular shifts in energy and the infrastructure that supports it. We are witnessing new demand, in part fueled by the enormous requirements of data centers, and in part by energy diversification. That means opportunity in fields like upstream nuclear, fed by Cameco, the world’s second-largest uranium producer. Access our full research database on Cameco Corporation Cavco ($CVCO US) Fund: Baillie Gifford Positive Change Strategy Thesis: Cavco manufactures factory-built homes and is owned for its leading scale and expertise in a growing affordable-housing market with expected accelerating earnings growth. Source: Read the original letter ↗ Analysis: By contrast, we bought Cavco, one of the leading manufacturers of factory-built homes in the US. Its homes can help address a shortage of affordable housing in North America. Manufactured affordable housing is a growing market, and Cavco’s scale, manufacturing expertise and distribution mean it’s among the leading players, having delivered consistent mid-teens earnings growth over a number of years. We think that growth should accelerate over the medium term, supported by regulatory reform as affordable housing becomes a policy priority. Access our full research database on Cavco Chien Kuo Construction Fund: Stone Sentinel Capital Thesis: Chien Kuo Construction is a cheap Taiwanese contractor with a valuable investment portfolio, float-like deposits, improving operations, and dividend-supported upside. Source: Read the original letter ↗ Analysis: Chien Kuo Construction is a new position initiated earlier this year. The company is a Taiwanese specialist contractor managing a large portfolio of investments that sum, after haircuts, to roughly 80% of its market cap. The construction business is hence undervalued at 4x PE on top of 33% insider ownership, a 6.5% dividend yield, and shrewd management. The current version of Chien Kuo is engineered by Wu, who became Chairman in 2019. Under Chairman Wu, the company moved away from speculative residential work toward insurance-led commercial complexes and government social housing. These are negotiated, capability-driven contracts with price-index escalation clauses that guard against inflation. Large construction projects typically involve milestone deposits from customers who pay Chien Kuo before it commences construction. Before the deposits are paid for labor and materials, Chien Kuo invests them, effectively utilizing deposits as what an insurer would do with float. About 70% of the investment portfolio is invested in money market funds and investment-grade bonds, yielding an effective yield near 5%. Another 10% is in Taiwan public equities. And the balance is in alternative investments including stakes in private companies, multi-asset funds, and private equity funds. The construction business has improved significantly under Chairman Wu. Since he took charge in 2019, revenues and margins have grown consistently, except during Covid. Within 6 years, revenues increased by 50%, gross margin by ~2x, and operating margin by ~7x. EBT, which includes interest income from the investment portfolio, saw margins expand by 11x. ROE grew from 1% to 18%. The investment portfolio effectively funds the company and hedges the bottom line during downturns. When Covid froze most construction, interest income and sales of securities cushioned the profit decline. In 2021-23, while EBIT fell 80% over 2 years, EBT declined only by 30%. Haircutting the equities and alternative positions by 10-50%, normalizing customer deposits (ie float), and accounting for tax on repatriation and the recent dividend paid result in a portfolio of NT$8b, which is roughly 80% of recent market cap. This implies that the construction business, which earns NT$490m mid-cycle, is valued at about NT$2b or 4.1x PE. If the segment grows topline by low single digits and maintains margins, an 8-10x multiple would be fair, implying NT$2-3b in additional value. That’s not all. The company paid NT$3.30 per share in dividends earlier this year, implying a 6.5% yield. A 3-year IRR is roughly in the high teens to 20% after summing the re-rating of the construction segment (20-30% over 3 years), the dividend yield, and 5% EPS growth. As mentioned, any good investment can’t be without contradictions. Construction is inherently a cyclical and rate-sensitive business. The top 5 customers also accounted for 80% of revenues. Fortunately, nearly 80% of the backlog relates to government- and insurer-funded infrastructure projects ranging from social housing to train stations, which are less rate-sensitive than residential projects. 4 out of the top 5 customers are government agencies that are likely more creditworthy than private developers. Unlike an insurer’s, Chien Kuo’s float is not built into its industry. It depends on winning large projects with milestone deposits, and a history of shrewd management and project selection has delivered. Whether they may continue to do so depends on continuity of Wu’s management culture, which has an enviable but short track record of 6 years. Succession risk comprises two layers. Chairman Wu is probably in his late 60s with 8-10 years of leadership left at best. Wu was also appointed by the controlling Chen family who owns ~33% of the company, so Wu’s tenure is determined by the Chen family. Family-controlled companies have a history of appointing family members as top leaders. The Chen family’s willingness to let Wu run the business through technical merit signals that the family prioritizes value creation over dynasty preservation. Access our full research database on Chien Kuo Construction China Resources Beer ($291 HK) Fund: Liontrust Global Income & Growth Fund Thesis: China Resources Beer has distribution, scale and a Heineken partnership, and its premium brands are still gaining share and growing rapidly despite weak consumption. Source: Read the original letter ↗ Analysis: We initiated a small position in China Resources Beer (CR Beer) late last year. Overall beer consumption is not growing in China, but the mix profile is changing. CR Beer has many advantages over its peers. These include widespread distribution and scale in production. Since 2018 it also has a partnership with Heineken, a key shareholder. As can be the case, we were too early. 2026 has been a challenge with continued weak consumption and a severe typhoon over the summer. But CR Beer’s sales have held up better than the market. Its premium brands have gained market share and continue to grow double digits. If it can continue to take costs out and emerge stronger from tough times, we are happy to be patient with companies. But their actions must be closely watched. Access our full research database on China Resources Beer Compagnie de l’Odet ($ODET FP) Fund: Horos AM Funds Thesis: Compagnie de l’Odet is the central Bolloré family holding company, protected by cross-holdings that also obscure its underlying value. Source: Read the original letter ↗ Analysis: For its part, Compagnie de l’Odet (”Odet”) is the holding company through which the Bolloré family controls the group. The origin of this structure dates back to the 1980s and 1990s, when Vincent Bolloré rebuilt the family business and created a chain of intermediate companies that allowed him to retain control with little capital. Through this process, Odet ended up establishing itself as the central holding company of the group. As we have seen, Bolloré is, in turn, the controlling shareholder of Odet, a cross-holding that acts as a defensive mechanism for the family, shielding it against any takeover bid. Of course, that cross-holding also masks the true value of Odet. This, together with the group’s recent cash movements (Odet is the main recipient of the extraordinary dividend recently distributed by Bolloré), encouraged us to build a position in this other family vehicle. Access our full research database on Compagnie de l’Odet


Read more

Read original article