Hi all,
Giles Capital just passed 2,048 subscribers and from next Monday (24 August), it’s going paid. Firstly, thank you. Many of you filled in the survey and it shaped this directly. What you value most is the selection and the summaries, that it’s handpicked and organised and increasingly you want more of our own analysis and conviction on the best ideas. That’s where Giles Capital is heading next.
Every week I filter through 30 to 50 company write-ups, several fund letters and a stack of articles, curating well over 1,000 investment ideas a year down to the best of them, sorted and summarised so you get them in one place. Going paid lets me go further: checking the details, flagging what looks wrong, marking what's worth looking into, and over time add more of my own analysis and a searchable database of everything covered.
It’s likely some of you will unsubscribe after this, and that’s completely fine. You should only pay for something you like, and something you want in your inbox every week. It’s not really about the money. It’s about keeping this going, making it better, and rewarding the readers who genuinely value the work rather than trying to be everything to everyone.
A free email will stay - the midweek Reads keeps landing every week and will get better, with selected company reports as a taste of the deeper work. But from the 24 August, the Monday weekly becomes for paid subscribers only: every write-up, the summaries, the fund letters, my own analysis and conviction on the best ideas, plus the full archive going forward.
To thank the readers who back Giles Capital early, there’s an early rate of $20 a month, or $200 a year, which works out at about $16.60 a month. Become a paid subscriber by 30 August to lock it in. After 30 August, it will be $25 a month, or $250 a year.
There’s also a new Partner tier, at $50 a month or $500 a year, for the few who want to back the newsletter directly. Partners get a large say in where the newsletter goes, feature requests, and can request a custom newsletter made to their own criteria. If you already support Giles Capital as a paid subscriber, thank you, you keep your current rate and nothing changes. And if you’re a student, or the price is a barrier where you live, just send me a message and I’ll sort something out.
Thanks so much for your time. Now, on to this week’s update..
Welcome to another edition of Giles Capital Weekly. Global bond yields are climbing on their own terms: analysts are flagging threats beyond central bank guidance, with $70 billion in shadow credit backstops for AI companies now registering as a systemic concern for debt markets. AI infrastructure costs are filtering into the real economy, with the UK reporting ‘chipflation’ as compute-driven price pressure reaches consumers. Iran's diplomatic window closed without a deal, leaving the Hormuz Strait and energy markets in a holding pattern. For value investors, rising cost of capital tightens the field to businesses that earn rather than borrow their returns.
Our top picks this week are Acid Investments on Global Tax Free, Korea's leading VAT refund operator, which earned a 41.5% return on invested capital last year and launches a Japan joint venture this November, and Rijnberk InvestInsights on Uber Technologies, where 208 million monthly active consumers and trips up 18% sit alongside a stock price that still prices in robotaxi disruption as existential rather than the partnership the company is actively building.
StockOpine’s Newsletter on Booking Holdings (🇺🇸 BKNG US - US$150bn)
All the anxiety surrounding LLM-driven search disruption has overlooked Booking's loyalty programme: more than half of all room nights booked. Net income up 118%; 27% margins last quarter.
Rijnberk InvestInsights on Uber Technologies (🇺🇸 UBER US - US$145bn) TOP PICK
The market is pricing robotaxi disruption the operating data do not support. Two hundred and eight million monthly consumers; trips up 18%. Down 21% over twelve months.
The Finance Corner on Nike (🇺🇸 NKE US - US$62bn)
Nike optimised for scale and ceded shelf space to Hoka and On. Down 75% from peak; insiders now buying. The reversal is underway; revenue is still flat.
P14 Capital on Owlet (🇺🇸 OWLT US - US$159m)
Revenue approaching $130 million and growing 30%, at a $159 million market cap. The pivot to health subscriptions is complete; approaching breakeven. The case grows stronger each quarter.
GHGInvest on Berkshire Hathaway (🇺🇸 BRK.B US - US$1.1tn)
Not a story about its largest holdings. At thirteen times trailing earnings, a cash pile exceeding $300 billion sits against a $1.1 trillion market cap. Greg Abel's first full year.
Long-term Investing on Alphabet (🇺🇸 GOOGL US - US$4.2tn)
Whether AI disrupts search queries or not is subject to interpretation. What's clear: queries just hit an all-time high. Revenue up 24%, cloud up 82%. P/E at seventeen times.
The Oak Bloke on Harbour Energy (🇬🇧 HBR LN - £4.1bn)
Current valuation makes no sense unless the windfall tax is permanent. Strip it out: $2.86 billion in first-half cash generation prices at close to 6x.
Iggy on Investing Newsletter on Interlife General Insurance (🇬🇷 INLIF GR - €124m)
Compounded 22% per year for a decade. At 4.5 times earnings and 0.77 times book, an MSCI Greece upgrade is the near-term catalyst. Already cheap without one.
Hated Moats on Novo Nordisk (🇩🇰 NVO US - US$200bn)
Down 42% over twelve months. Revenue falls as US GLP-1 prices reset in 2026. At eleven times trailing earnings, permanent impairment is the only thesis that justifies this price.
TacticzHazel’s Substack on Taiwan Semiconductor (🇹🇼 TSM US - US$2.1tn)
TSMC has durable competitive advantages in a world where every AI dollar eventually reaches the foundry that makes the chips. July revenue up 44.7%; seven-month figure up 37%.
Best Anchor Stocks on Nintendo (🇯🇵 7974 JP - ¥10.3tn)
Operating profit up 150%, partly from tariff refunds. Switch 2 units fell 34% from launch; the case rests on an IP catalogue no competitor can touch and software margins.
Capytal Management on Huishang Bank (🇨🇳 3698 HK - US$8.3bn)
All the anxiety surrounding Chinese bank credit quality overlooked Huishang: bad loans at 0.98% and falling. Three and a half times earnings, 6% yield. Anhui is home to CXMT.
Global Equities on CTOS Digital (🇲🇾 CTOS MK - US$370m)
Think of CTOS as the infrastructure layer beneath Malaysian credit: invisible from the outside, impossible to remove from within. Free cash flow yield 7%; PE exit pending.
Acid Investments on Global Tax Free and Geumhwa Plant (🇰🇷 204620 KS, 🇰🇷 036190 KS - US$325m, US$130m) TOP PICK
The valuation makes no sense unless Korea's leading VAT refund operator stops earning 41.5% on invested capital. Domestic revenue up 30%; Japan joint venture launches November.
JPARCVUE on GS Yuasa (🇯🇵 6674 JP - ¥710bn)
Japan's market leader in batteries for automotive and grid storage. At seventeen times earnings, every unit of domestic electrification capacity passes through this supply chain. Revenue growing alongside infrastructure demand.
JPARCVUE on Nakanishi (🇯🇵 7716 JP - ¥180bn)
Dominant global position in dental handpieces and surgical micro-motors used in every major market, regardless of brand. At ¥180 billion market cap, revenue grows as dental access expands globally.
Cohong Lane on Bank of China (🇨🇳 3988 HK - US$63bn)
H-shares at 5.7 times earnings and 4.7% yield; A-shares command a structural premium. One bank, two prices. The investment case rests on that gap compressing.
AI Proem on Tencent (🇨🇳 0700 HK - US$450bn)
Revenue up 9%, profits up 12%, at fifteen times forward earnings. I imagine regulators are less confused about what Tencent is than the market has been since 2022.
Crack The Market - Who Powers AI, Part 1: The Anatomy of a Gigawatt
The full infrastructure stack behind one gigawatt of AI compute: roughly $35 billion in total capex, 20% energy-related, with $1 trillion in cumulative mechanical and electrical spend projected through 2029. The binding constraint today is the interconnection queue, now 2.3 terawatts deep, not capital availability or permitting.(22 min read)
stf research - Why I Keep Buying Japan's AI Supply Chain
Maruwa, Union Tool, and MEC hold 50-80% global share in substrate materials and precision tooling required by advanced PCB production. They trade at 20-35x forward earnings, below Taiwan and Korea equivalents. The thesis: foreign investors are misclassifying these as legacy industrials rather than technology franchises embedded in the AI supply chain.(12 min read)
The Intellectual Edge - The best explanation wins.
David Deutsch's framework on hard-to-vary explanations, applied to investing. The best investors don't collect rules of thumb; they find genuine explanations for why something is mispriced. Fallibilism, not authority, is the correct epistemic posture. Observation is always theory-laden, which changes how much weight to give management's own narrative.(18 min read)
TSCS - It's Not The Poor. It's You.
The cost floor for median American life has risen to 75% of gross wages, driven by rent at 67% of the floor. Electricity will be the next pressure as AI-driven grid buildout outpaces capacity, the author argues, forcing Washington toward supply-side reflation rather than rate cuts.(18 min read)
Elliot’s Musings - Which AI Buildouts Actually Earn Their Cost of Capital
ROIIC analysis across seven hyperscalers and neoclouds. Deployed capital earns above cost everywhere except neoclouds, which fail on any measure. Meta and Oracle sit at hurdle rates. The next four earnings prints will settle whether the buildout is value-creating or accounting fiction.(18 min read)
Panda Perspectives - Earning the Multiple
Leonid Mironov dissects why the S&P 500 earns a 20x multiple while MSCI China trades at 12.5x. The American premium is built on 13.9% net margins, structural buybacks, and automatic wage-linked equity inflows. China is replicating the playbook deliberately, with the scale to pull it off.(18 min read)
Baiguan - China Insights, Data, Context - Eel, Caviar and Electricity
China's manufacturing logic has migrated into agriculture and services. Eel farming, caviar production, and healthcare platforms all follow the same pattern: standardise inputs, processes, and outputs, powered by cheap electricity. A short piece that reframes how China's industrial surplus generates capacity in categories that look nothing like manufacturing.(12 min read)
FT Alphaville - AI people need people
Google's organizational reshuffle, with Demis Hassabis stepping back and Jeff Dean departing, examined as a symptom of tension between research culture and commercial urgency. Despite inventing the transformer, Google's fortunes depend more on retaining individual talent than deploying AI. The core insight applies well beyond Google.(8 min read)
The Intellectual Investor - Chess Nights at Red Robin
Vitaliy Katsenelson on Thursday night chess meetups at a Red Robin in Denver. A personal essay on the community that forms around the game, playing with his daughter Hannah, and how chess functions as intellectual exercise and genuine family time. Not about investing, but worth eight minutes.(8 min read)
Five of six letters this quarter describe markets rewarding AI proximity over fundamentals, frustration that cuts across style boxes. The more interesting pattern is where each manager is looking instead: AI-adjacent industrials held for years, energy pipelines, and deep-value compounders the index has no reason to notice.
Heartland Value Plus Fund: +19.25% Q2 2026 vs Russell 2000 Value +17.19%, gains driven by AI-adjacent industrial positions held long before the trade arrived. Letter discusses:
Littelfuse (🇺🇸 LFUS US, ~US$3bn) - Existing: high-voltage protection products now ~20% of revenue from AI data centers; bookings +20% YoY
FirstCash Holdings (🇺🇸 FCFS US, ~US$4bn) - Existing: pawn fees +39% YoY as high gold prices lift loan values; guidance raised
Century Communities (🇺🇸 CCS US, ~US$1bn) - Existing: trading 0.81x book, well below recent homebuilder takeouts at 1.1-1.2x
Long Cast Advisers: +20% Q2 2026, +19% YTD, and 343% cumulative since inception in 2015; concentrated small and micro-cap SMA with no AI, leverage, or passive-fund exposure. Letter discusses:
Pro-Dex (🇺🇸 PDEX US, ~US$50m) - Existing: surgical effectors for orthopedic robots; Zimmer CVRs could add $1.15-$4.50 EPS by 2028-2030 if Monogram revenue hurdles hit
NRC Group (🇺🇸 NRC US, ~US$150m) - New: record contract backlog; two largest competitors merged in a $6.5bn PE deal, rationalising the industry
Matrix Service (🇺🇸 MTRX US, ~US$300m) - Existing: June year-end results pending; fund expects record profitability and above-consensus cash earnings
Legacy Ridge Capital Management: +15.7% net H1 2026 vs S&P 500 +10.0%; portfolio of eight companies in airlines, energy pipelines, and royalty businesses, with the Iran War used as a lens for durable pricing power rather than a short-term disruption. Letter discusses:
Allegiant Travel (🇺🇸 ALGT US, ~US$1.5bn) - Existing: completed acquisition of Sun Country; Spirit Airlines liquidation removes the last irrational capacity actor from the market
Plains All American Pipeline (🇺🇸 PAA US, ~US$9bn) - Existing: capex guide raised 21% for 2026; US energy export volumes at record as Strait of Hormuz closure validates domestic infrastructure
Signia Small Cap Value Strategy: +27.82% YTD net Q2 2026 vs Russell 2000 Value +23.00%; received its first institutional allocation ($30M) in H1 2026, with the letter flagging evidence of early large-cap to small-cap rotation. Letter discusses:
Cracker Barrel (🇺🇸 CBRL US, US$670m) - New: added at ~$30/share; owns land and buildings for 358 of 650+ stores, with real estate value estimated at $1.2-$1.4bn; traffic improving after management reversed the logo redesign
Itron (🇺🇸 ITRI US, ~US$10bn) - New: utility meter solutions; Outcomes segment +22.1% YoY; transitioning to recurring revenue model at ~12x 2027 EPS
Sycamore Mid Cap Value Equity Strategy: +9.6% net Q2 2026 vs Russell Midcap Value +13.83%, with approximately 85% of underperformance explained by the decision not to own Technology Hardware, which returned +144.5% in the quarter. Letter discusses:
Molina Healthcare (🇺🇸 MOH US, ~US$15bn) - Existing: Q1 2026 results signaled worst of medical cost ratio deterioration behind the company; CMS risk-adjustment data viewed favorably
Steel Dynamics (🇺🇸 STLD US, ~US$18bn) - Existing: metal spreads widened; 25% Section 232 tariffs reduced import competition; benefits from data center construction and onshoring
Devon Energy (🇺🇸 DVN US, ~US$18bn) - Existing: under pressure from crude price decline post-Coterra acquisition; first post-merger earnings due August 2026; thesis intact
Oakmark Fund: +2.45% Q2 2026 vs S&P 500 +15.20%, consistent underperformance from the fund's refusal to chase AI-adjacent concentration; Oakmark has returned 12.66% annualised since inception in 1991 vs the S&P 500's 11.02%. Letter discusses:
Booking Holdings (🇺🇸 BKNG US, US$150bn) - New: initiated near a ten-year trough valuation; LLM traffic to travel sites remains negligible; loyalty program accounts for more than 50% of room nights booked
Delta Air Lines (🇺🇸 DAL US, ~US$25bn) - Existing: passing through Iran War fuel cost spikes faster than historical norms; premium leisure and corporate travel demand healthy
Equitable Holdings (🇺🇸 EQH US, ~US$11bn) - New: pending merger with Corebridge Financial; trades below 6x 2027 distributable cash flow; more than half of cash now in capital-light segments
Thanks to Insider Monkey for compiling the investment letters which are featured here.
Thanks for reading Value Investing by Giles Capital! This post is public so feel free to share it.
Currently, I feature free content and collaborate with select paid newsletters. If you're a writer producing insightful investment analysis, whether free or behind a paywall, I'd love to discuss potential features. What would you like to see in future editions? Let me know in the comments.
Disclaimer: This newsletter is for informational purposes only and not investment advice. The intro reflects my views, while investment summaries are my interpretations of original authors' analyses. Information may not be fully verified and is subject to correction. Original authors' complete views may differ. Always do your own research before making investing decisions.
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.