The Latticework Monday Morning Briefing is our weekly “Guide to the Markets” for long term-oriented investors. It is sent on a separate mailing list (complimentary to members). If you do not wish to receive it, opt out here.
Check out the agenda for Latticework 2026 (Chicago, November 10-11). The summit features fireside chats, panel discussions, idea pitches, and a group dinner. Seating is limited. Register now.
Listen to the Latticework Monday Morning Call:
Ideas from our member newsletter
We highlight a few stock write-ups from the latest Weekly Inspiration:
Exor (Amsterdam: EXO), the Agnelli family holding company run by John Elkann, trades at a 57% discount to a pro-forma net asset value of ~€167 a share and has committed €500 million to a buyback that retires 3.5% of its capital. The interim report showed ~€1.2 billion of unlisted investment sales, about €5.90 a share, of which ~€750 million settles after June 30: debt and equity in the media business GEDI for €249 million, the proposed sale of Welltec for €488 million on Andrew Brown’s calculation, Lifenet for €209 million and assorted smaller pieces. Counting the ~€1.5 billion still to arrive from Iveco and Welltec, he puts Exor in net cash of ~€1.2 billion, which leaves scope for more retirement of equity while the discount persists; at these prices a €50 million disposal buys back €116 million of underlying value. His complaint is Lingotto, the externally managed vehicles that make up €21.30 a share, or 13% of attributable value, nearly 2.5x what remains of Stellantis, with disclosure he calls opaque; 13-F filings suggest the US holdings of its two main listed funds have gained ~9.5% since June 30, worth ~€1.53 a share of NAV. The discount, in his reading, prices the concentration in Ferrari, the opacity of Lingotto and the fear of a new ~€4 billion commitment in healthcare, where Elkann’s record is spotty.
Meta (Nasdaq: META) brings a consumer AI agent to ~3.6 billion daily users of its apps: Muse, released September 8, reached 2.8 million downloads in its first twelve days, a pace reportedly faster than ChatGPT’s, and Meta does not need the best model, only one within shouting distance of the free ones. The shares have risen 36% in a month to $748. Sanjiv writes the piece as a possible mistake of omission: on August 11 he declined to buy at ~$780 and named ~$550 as an entry point, the shares touched that level on August 18, and he did not act. Muse runs on Meta’s Muse Spark model and completes tasks rather than answering prompts: booking travel, shopping on Amazon or Shopify, ordering from DoorDash. The free tier allows 100 million tokens a week, with paid tiers at $20 and $100 a month, and one user reported that it found and bought an auto policy saving $3,500 a year in five minutes. He doubts consumers will pay $20 a month; the ultimate case is for sellers to pay Meta a small take-rate on the transactions Muse conducts. Amazon blocked the agent from its site on the day of writing, as it has other shopping agents, because agents ignore the advertising behind its ~$76 billion of annual ad revenue; Meta’s shares rose ~12% anyway. Rivals from Amazon to Apple may answer with agents of their own, so he calls the durability of Muse’s lead unclear.
Bellway (London: BWY) has compounded total shareholder returns at 12.9% a year since 1991, builds ~5% of the UK’s houses, and trades at 0.67x tangible book value, close to its 21-year low and against an average of 1.15x, and at a 33% discount to a liquidation value that counts only land, work in progress and cash less every liability. The UK housing downturn has run since 2019: completions are down 21%, house prices are 10% behind inflation since 2022, and Bellway’s adjusted earnings per share fell 56% from 436p in FY19 to an estimated 192p in FY26 as gross margins went from 24.6% to 17%. Gillen Markets, which is adding the shares to its Lump-sum Growth Portfolio, does not claim the bottom is in; 10-year gilt yields above 5% threaten affordability and planning consents remain slow. The case rests on what happens meanwhile: Savills puts land prices 10% below the 2021 peak, so the landbank is being replaced with cheaper plots, first-half gross margin edged up to 16.2%, and the company has bought back ~9% of its shares since FY22, with a new £50 million programme on top of the £150 million launched in late 2025. Adding the profit on 47,200 consented plots at £320k and a 16% margin lifts the liquidation value to £5.3 billion, a 55% discount, though the authors advise against leaning on it. At £21.08 the shares sit on 11x FY26 earnings, an assumed 75p dividend yields 3.6%, and net debt was £72 million in January against £506 million of cladding provisions still to be paid.
Formula One (Nasdaq: FWONK) holds the exclusive commercial rights to the F1 championship until 2110, has 18 of its 24 races contracted to 2030 or beyond, and more than quadrupled free cash flow from $0.19 billion in 2023 to $0.79 billion in 2025 as revenue grew 25% and margins widened. The shares are the non-voting Series C of Liberty Media’s Formula One Group, a tracking stock Liberty created after buying F1 in 2017; with the Braves, SiriusXM and Liberty Live since split off, it is Liberty’s primary remaining business. Over those two years revenue went from $3.57 billion to $4.48 billion, sponsorship fastest at 44% to $0.84 billion, and the operating margin from 7.9% in 2024 to 12.9% in 2025; Apple pays ~$750 million for US rights through 2030 and LVMH ~$1 billion over ten years as timekeeper. First-half 2026 revenue fell 8% after the Bahrain and Saudi races were cancelled over the Middle East conflict, yet first-half free cash flow still rose to $0.61 billion from $0.57 billion. The write-up, by Jacqueline Oh, a summer intern at Open Square, with the firm’s questions interleaved, acknowledges that the shares already trade at ~33x 2025 free cash flow. Goldman’s forecasts of $1.14 billion of free cash flow and $1.65 billion of adjusted OIBDA in 2029 give $150-165 a share on today’s multiples; MotoGP, bought in 2025 for $3.66 billion and breakeven on EBITDA, drags on the valuation, and the voting FWONA line trades 8% below FWONK for lack of liquidity.
Sabre (Nasdaq: SABR) is a turnaround with operating leverage: between 2023 and 2025 it cut technology costs from $951 million to $711 million and lifted operating income from $32 million to $295 million, Constellation Software has taken a ~12.7% stake, and three more years of 5% revenue growth at a 34-40% incremental margin would take adjusted EBITDA from ~$600 million to $750-777 million. Constellation has also put Vela chief executive Damian McKay on the board. Sabre’s Marketplace links airlines to travel agencies for booking fees, and replacing that embedded technology can disrupt an agency’s operations. GHGInvest credits the recovery and says a contrarian can reasonably buy before cash generation confirms it, since waiting can mean paying substantially more; the September refinancing, $1.35 billion of 9.875% notes due 2032, pushes maturities out. The author nonetheless watches rather than buys at $2.26. Management’s August outlook puts $475 million of cash interest against the ~$600 million of EBITDA, 79% of operating cash before reinvestment, and the author’s arithmetic leaves a funding gap of ~$0.8-1.0 billion through 2029. In the middle scenario, EBITDA grows 25% by 2029 but the multiple settles at 6.5x from ~8x today (the author’s own valuation of sustainable cash), which puts the shares at ~$2.40, or ~2% a year; the same operations at a different exit multiple return almost 24% a year, and one turn of multiple moves the value by $1.43 a share while three years of retained cash add $0.24. Screening prices are $1.15-1.32 at 20-25% required returns; second-half free cash flow near $81 million and the November 17 investor day are the named tests.
Wavestone (Paris: WAVE), a French consulting firm with 6,000+ professionals in 17 countries, has taken AI-related work from 8% of revenue in FY25 to 17% in FY26 and ~22% in recent quarters, and has a market cap ~€730 million with ~€120 million of net cash. The firm sits between strategy houses and integrators, designing transformations for tier-one European institutions and then implementing them; EDF and Deutsche Bahn are each 5% of revenue, Crédit Agricole 4%, financial services 33% in total and energy and utilities 18%. The author, who has worked in the sector, reads the unit economics from a daily rate of ~€938 and a utilization rate of ~71-72%, with personnel costs above €600 million, two-thirds of revenue; a 2-3 point recovery in utilization would flow almost straight to operating margin. Management’s 2030 plan, Lead the Shift, sets aside up to €100 million of organic spending on AI and up to €800 million for acquisitions, following AI Builders in France and Sand Cherry in Denver. On an 8.5% discount rate and 2% terminal growth, the base case, with organic growth accelerating to 5% and margins rising from 12% to 15%, gives €82.60 a share against €29.20; a bear case with growth near zero for two years and margins held at 10.5-11.5% still gives €61.68. The write-up grades the idea 8.5 of 10 and would not be surprised by a re-rating toward €60-75 as the AI share of revenue passes 25%.
As always, the above theses reflect the linked authors’ views (available here), not Latticework recommendations.
Articles worth your time
Shall We Repeal the Laws of Economics - Part III, by Howard Marks of Oaktree Capital, takes the Treasury’s August decision to at least double its long-dated bond buybacks, from $2 billion to $4 billion per operation, as the latest attempt to override the laws of economics, and argues it treats the symptom. The 30-year Treasury closed above 5.3% on August 17, a 19-year high; the buybacks knocked yields down for a day, and when the first enlarged operation ran on September 9 at $6 billion, yields rose. Marks lists the causes the policy does not touch: PCE inflation of 3.7% in July against a 2% target, a deficit near 6% of GDP with unemployment at 4%, net interest above $1 trillion a year and more than the defense budget, and a multi-trillion AI buildout competing for the same capital. He does not expect default, since the debt is in a currency the US issues and the dollar still accounted for 57% of allocated reserves in the first quarter, but he expects the cost to arrive through the dollar, and he quotes Stanley Druckenmiller: a 30-year at 5.5% “isn’t a crisis. It is an invoice.” His prescription is behavioral, with higher top income tax rates, fewer preferences, spending growth held below GDP growth, and productivity from AI. To a friend who asked whether to sell his stocks, he said no: the problem is fiscal management, not US companies, and dollar cash and bonds carry the same risk.
Nobody Will Slow Down on AI (A Game Theory View), by Mauricio Heck, asks why the labs keep accelerating when the people building the technology say they want to slow down, and answers with a prisoner’s dilemma. Dario Amodei’s essay “We Must Pace the Frontier” drew a match from Sam Altman within hours and “Dario is right” from Elon Musk; a day later President Trump said whoever wins AI wins. Two labs, two choices: if the rival paces, racing gains ground, and if the rival races, you have to keep up, so both keep advancing even though both prefer restraint. Three features harden that equilibrium. Leadership feeds on itself, and once better AI helps build the next AI, a small temporary gap can become uncatchable. The corporate race sits inside the country race, and Amodei’s own essay concedes democracies can slow only as far as their lead over China allows. And a slowdown is hard to verify: a data center needs land, power and chips, but algorithmic gains leave no footprint, so a commitment is cheap talk. The two reasons for hope are warning shots, such as the ~700 of ~1,200 supposedly isolated OpenAI agents that hacked Hugging Face in July, and repeated games, where tit for tat wins because cheating today is punished tomorrow. The catch: if everyone believes the first to superintelligence wins for good, the repeated game collapses into a one-shot game, and the most useful thing left is uncertainty about how close the finish line is.
The Business of Building God, by Rohit Krishnan, treats OpenAI and Anthropic as the businesses they are about to ask public markets to value, and finds the moat is one, maybe two, models deep. Talent, compute and the usage data that flows back from coding traces and conversations have kept the frontier labs ahead of the open-source wave, but Chinese labs keep closing the gap, and a frontier attempt costs on the order of $1 billion each for a training run, data and talent, small money against the nearly $90 billion Facebook spent on the metaverse. Hence the push into adjacent markets: advertising on a billion-plus consumer base, robotics, wet labs for cancer research. Krishnan thinks those markets are far harder than they look, since few industries contain several $100 billion revenue businesses, and Meta launching Muse on its own model shows how each new segment fragments rather than consolidates. Meanwhile the customers behind $200 billion-plus of token spending are hitting wallet limits; CFOs want the benefit on the income statement, companies restrict top models to top users and route the rest to cheaper ones, and he himself stopped paying for the highest tier of every model some time ago. That leaves two paths: become a cloud company with compressing margins as DeepSeek, GLM and Minimax catch up, or persuade governments to shut out open models, which is hard to sell against a copyable file. Only recursive self-improvement, the true unknown, would make a moat durable.
Japan’s Value-Up All-Stars, by Naoki Ito of Verdad, measures what the Tokyo Stock Exchange’s March 2023 request for capital efficiency has done to valuations: the share of Japanese companies trading below 0.7x book has fallen from 31% to 16% in four years, and the share below 0.5x from 16% to 5%. The change is uneven by size, and the TSE is now turning its attention to the Standard Market, where the smaller deep-value names sit. When the exchange asked 43 institutions to name companies that had made real progress, Verdad contributed 17, all of which made the resulting list of 300, and the article walks through them. The pattern is explicit, funded commitments rather than slogans: Glory set a 4% dividend-on-equity floor and bought back 6% of its market cap last year, Aichi Steel is unwinding ~¥50 billion of securities to return more than ¥70 billion through 2030 against a ¥200 billion market cap, Nishikawa Rubber targets 8% DOE through 2030, and T.RAD lifted its DOE floor from 3% to 5% and re-rated from 0.3x to 1.5x book. KPP Group is the instructive exception: an 80% dividend increase and a 4% buyback produced a 42% dollar return over three years, yet the multiple stayed near 0.7x because book value per share grew 44%, on 28% growth in common equity and an 11% smaller share count. The article’s conclusion is that small- and mid-cap Japanese value is entering its most interesting period yet, as reform trickles down.
The deck
The 131-page Monday Morning Briefing is available to members. It spans our weekly scoreboard, idea-generation screens (this week including quality names near their 52-week lows, micro-cap “tiny titans,” recent spin-offs, activist campaigns, superinvestor holdings, owner-operators, share-count shrinkers, and the broad US valuation screens), market valuation and positioning, macro and fixed income, and equity valuation screens across international markets (Canada, the UK, Germany, Australia, Japan, Korea, India and Sweden, alongside a cross-country P/E table and European REIT discounts to net tangible assets). We welcome your feedback.
Feedback on the Briefing
“Most of what I monitor, all in one place. Great value add.” —Brad Lummis
“Loving these Monday briefings!” —Jon Bartel
“Tightly presented and easy to digest. I just spent 20 minutes going through it, and it’s helped to level set me for the week ahead.” —Michael Loftis
“A great piece and thoughtfully assembled.” —Brian Wolf
“I have never seen more valuable content in one place.” —Bill Coleman
“Worth its weight in gold.” —Shree Viswanathan
Featured Events
Latticework 2026 (AGENDA) (REGISTER), Chicago (Nov. 10-11, 2026)
Ideaweek 2027 (FULLY BOOKED), St. Moritz (Feb. 1-4, 2027)
The Zurich Project 2027 (COMING SOON) (Jun. 1-3, 2027)
