Money Supply Growth Hits 4-Year High as Rate-Cut Bets Flip to Rate Hikes

Key Takeaways

1. US money supply grew 5.7% year over year, the fastest since June 2022, while market expectations flipped from two 2026 cuts to pricing at least one more hike this year.
2. Raoul Pal launched a free AI tool trained on 35 years of his thinking and argued tokenization is underrated, extending BlackRock's AI-and-crypto thesis to identity, contracts, attention, energy, and information.
3. US macro data showed a resilient labor market and accelerating PMIs alongside a 7.12% 30-year mortgage rate, a third straight weekly drop in mortgage applications, and a Richmond Fed manufacturing index still expanding but with capex in contraction.
4. Jeffrey Katzenberg said AI remains almost entirely on the "reasoning" side, argued Hollywood should set the conditions for its existence rather than try to make it disappear, and predicted lower costs would increase film output and new narrative forms.
5. Jason opened applications for the second Sanabil Founder University cohort in Riyadh and disclosed small long-horizon bets on four autonomous-driving names, expecting consolidation and acquisitions within 24 months.
6. Mohnish Pabrai distilled Buffett, Munger, and Graham into 30 mental models, arguing Berkshire's success came from roughly 12 major decisions in 60 years and that investors must never lose the ability to make the next bet.

1. Liquidity, Rates, and the Macro Regime

1. US money supply rose 5.7% over the past year, the largest year-over-year increase since June 2022, with the author framing it as "the money printing is back" after a brief pause. Market expectations have reversed sharply: at the start of 2026 the market priced two rate cuts, the Fed hiked 25 basis points last week, and the market now prices at least one more hike before year-end, with the author noting Trump wanted a dovish Warsh and lower rates but got a hawk instead. A reposted item claimed that since 1982, money invested after Fed rate hikes has actually outperformed money invested after cuts, and another reposted item argued the current 30-year mortgage rate looks high only outside a broader historical context. — via 1 2 3 4 5 6
2. Ben Carlson listed the current mix as 7% mortgage rates, 4% unemployment, 5% bond yields, 4% cash yields, $100 oil, $5 gasoline, and 3.4% inflation, with stocks near record highs while consumer confidence sits near record lows, calling the moment "something for everyone" and strange. Separately, the Nasdaq hit an intraday record high. — via 1 2

2. AI, Tokenization, and the Machine Economy

1. Raoul Pal launched a free AI summary tool trained on his 35 years of thinking that answers questions on markets and liquidity and summarizes his latest article on tokenization. He argued tokenization is underrated: people focus on weekend trading of Apple, instant settlement, and 24/7 markets, but the bigger picture is that everything becomes a token and the entire economy is rebuilt around it. Commenting on BlackRock's AI-and-crypto view, he described AI as machine-native intelligence and crypto as machine-native money, with agents unable to use banks so they need blockchains, stablecoins as their payment currency, and compute as a tokenized asset class; he said BlackRock stops at money and compute, while the machine economy also needs identity, contracts, attention, energy, and information itself, and most of those asset classes do not yet exist, concluding that "everything will be a token." — via 1 2
2. Pal's "Everything Code" framework holds that at current debt levels growth cannot solve the problem, so governments can only print to pay interest and keep borrowing, and demographics plus debt make currency debasement the only option, which in turn drives nearly all asset prices. Despite the surface pessimism, he said he is optimistic about the productivity miracle of the "Exponential Age," and merged the Everything Code and Exponential Age frameworks for the first time in his first book, The Everything Code. Cited data showed emerging-market debt/GDP at about 78%, the highest since records began in 1880, more than double its level since the 2008 financial crisis and above the WWII peak of about 45%, having never exceeded 60% before 2020; advanced economies sit at about 108%, above 100% for a decade and below 80% before 2008. — via 1
3. Jeffrey Katzenberg said that in 2023 he judged new AI tools would cut the time and cost of world-class animation by up to 90% within three years, a view some peers met with alarm or anger. He argued AI is currently almost entirely on the "reasoning" side, able to reason, evaluate, optimize, and pattern-match, but lacking empathy, commitment, serendipity, and the kind of creativity only a particular person can express, and said the line between reasoning and creation is real and no one in the industry can describe the scientific path across it. He said AI will not disappear and Hollywood should spend its energy deciding the conditions of its existence rather than trying to make it vanish, that the North (Silicon Valley) needs the South's (Hollywood's) creativity, and that the right path is to build with storytellers rather than over them, with credit, consent, and compensation. — via 1
4. Katzenberg drew historical parallels: Sousa's 1906 opposition to mechanical music did not stop the technology but helped produce the 1909 Copyright Act, and sound film displaced pit musicians, a real loss that nonetheless gave rise to movie musicals, modern scoring, and sound design. He recalled Disney's 1980s co-development of the CAPS system with Pixar for The Little Mermaid, Beauty and the Beast, Aladdin, and The Lion King, and DreamWorks' decision to stop hand-drawn animation for all-CGI, which he called correct but which left some talent behind, with some moving to new tools to do their best work and others not. He cited Walt Disney's focus on character and audience emotion over tools, George Lucas's "not how, but why," and Steve Jobs on technology and the humanities, and predicted lower costs and barriers would increase film output, make studios more willing to take risks, and produce entirely new narrative forms. — via 1

3. US Growth, Jobs, and Housing Data

1. Private US employers added an average of 20,000 jobs per week in the four weeks through September 5, 2026, with hiring strengthening for a third straight week, while the Philadelphia Fed services index showed full-time employment jumping 17 percentage points in September to a four-year high. The Richmond Fed manufacturing index adjusted by the ISM method fell to 50.8 in September, still in expansion, but its capex component slid further into contraction and has not expanded for more than two years. S&P Global's US manufacturing PMI rose to 57.0 in September (expected 53.7, prior 53.9), services PMI rose to 58.7 (expected 55.8, prior 56.5), and the composite PMI rose to 58.4 (expected 55.3, prior 56.0). — via 1 2 3 4 5
2. MBA mortgage applications fell 1.5% in the week through September 18, a third consecutive weekly decline (prior -4.1%), while the 30-year mortgage rate rose to 7.12%, the highest since May 2024. In ETF flows, large- and mid-cap funds took in nearly $20 billion in the week through September 18, while small caps saw the largest outflows. A reposted item noted the number of negative-beta stocks in the S&P 500 hit a record high, meaning the scale of single stocks moving opposite the index is the largest ever. — via 1 2 3

4. Founder and Investor Signals

1. LAUNCH's Sanabil Founder University second cohort starts November 8 in Riyadh, a 12-week pre-accelerator for founders building, launching, and scaling in the MENA region, working with Jason, the LAUNCH team, and mentors. Reposted information listed first-cohort progress: Emtethal raised $250,000 and joined 500 Global's Sanabil Accelerator; Circle went from wireframes to $7,000 MRR; Qanary raised $160,000 and joined Orbit's Sanabil Accelerator; Ajeer Pay joined STC's InspireU; Iqra turned an idea into a product used by thousands globally; and QualifiedCrew launched with $3,000 MRR. The program combines online work with in-person sessions at Sanabil Hub. — via 1 2 3
2. Jason said he recently made small purchases of $WRD, $PONY, $LCID, and $RIVN, judging that consolidation is coming in autonomous driving based on major progress at Tesla, Waymo, and Uber; he expects all four companies to be acquired within 24 months and the basket to roughly double in under two years. He described these as medium-term bets with a 1-to-4-year holding period, not investment advice and not day trades or one-year flips, noted most of his stock investments are held 10+ years, invited challenges to his thesis, and disclosed other autonomous-driving holdings including private companies Zipline and goautolane. — via 1
3. Mohnish Pabrai, in a 51-minute Columbia Business School lecture, compressed lessons from Buffett, Munger, and Graham into 30 mental models for finding exceptional investments, with the core view that Berkshire Hathaway's success came from roughly 12 major decisions over 60 years and that investors do not need to be right constantly but must avoid destroying the rare ideas that carry the whole portfolio. He said investors often sell their best companies too early, keep adding to positions that have stopped working, and mistake uncertainty for real risk, using Ferrari, Walmart, Goldman Sachs, Turkish real estate, spin-offs, and arbitrage as examples of asymmetric bets. He offered two rules: if you cannot explain the investment logic to a ten-year-old in three or four sentences, you do not truly understand the asset; and always keep enough rope to climb out of the deepest well, because no investment is worth losing the ability to make the next bet. — via 1
4. Reid Hoffman endorsed the view that data-center siting should be decided by the communities that will live alongside the facilities long term, including whether to build and on what terms, with the decision belonging only to communities that bear the costs and share the benefits. A post he amplified said the governors of Virginia and New Jersey are quietly steering Democratic data-center policy back toward rationality: no bans, community decisions, no NDAs or subsidies, and priority for local construction jobs. Hoffman also shared his conversation with Hillary Clinton on international AI cooperation, in which she said a true "deal-maker" might make it work, while he expressed skepticism. — via 1 2

5. Market Commentary and Long-Run Views

1. Liz Ann Sonders noted the Nasdaq hit an intraday record high and highlighted long-run urban projections: New York is expected to remain the world's largest city economy in 2050 with $6.4 trillion in GDP; Shanghai is projected to rise from 12th in 2025 to 3rd in 2050; and three Chinese cities are expected in the global top ten city economies in 2050 versus none in 2025. She also marked her 40th anniversary on Wall Street, recalling joining on September 22, 1986, and working 13 years for Marty Zweig from 1986 to 1999, crediting Zweig with correctly calling the 1987 crash by cutting from fully invested to a very low position at the end of August and starting to buy a week after early October. — via 1 2 3 4 5
2. Ben Carlson noted his book Your Perfect Portfolio was included in Christine Benz's recommended reading list, which he called excellent across all eight titles, and listed topics in his Animal Spirits: The Wobbly House of Cards piece: what happens without an AI crash, the chart of the year, why buy-and-hold does not work, 10-year return forecasts, why the Fed is hiking, and the Lakers' $3 billion. Meb Faber quoted Robin Wigg calling the Austrian century bond "the trashiest of all fixed-income trash," and shared Maggie Mahar's line that "people resist randomness, markets resist prophecy." — via 1 2 3 4