Newsletter · 2026-08-10 · Ashutosh Agarwal

Fund Managers Pitch Hims, Intuit and Micron as the AI Top Debate Splits - Weekly Podcast Idea Digest - Week of August 10, 2026

For the week of August 3 to 10, 2026, no idea conference dropped, so the best single-name pitches came from fund-manager interviews: Hims on data rather than GLP-1s, Intuit at 16x after a two-thirds drawdown, Exor as cheap Ferrari, and Micron cross-confirmed by two separate managers, all underneath the Burry versus Eisman fight over whether the AI boom is topping.

Weekly Podcast Idea Digest

Week of August 10, 2026: Fund Managers Pitch Hims, Intuit and Micron as the AI Top Debate Splits

This week's podcasts were unusually full of specific stock ideas. There was no dedicated idea-conference episode (no Sohn, Robin Hood or Delivering Alpha pitch dropped in this window), so what follows comes from fund-manager and analyst interviews, plus a couple of follow-ups on picks first made at investor conferences.

One theme ran underneath almost everything: the argument over whether the AI boom is topping out. Michael Burry says yes and is betting against it; Steve Eisman says not yet. That fight matters here because several of the best bottom-up ideas this week are, in effect, ways to buy good companies that got thrown out with the AI bathwater. A few quick definitions used below: free cash flow is the cash a business has left after running itself and investing; gross margin is what's left of each sales dollar after the direct cost of the product; P/E is the share price divided by a year of earnings; and NAV (net asset value) is what a holding company's investments are worth after subtracting its debts.

Here are the ideas, roughly strongest-conviction first.

1. Hims & Hers ($HIMS), long, Paul Cerro, Cedar Grove Capital

Yet Another Value Podcast, August 7

Paul Cerro has now traded Hims three times (long, then short, then long again) and he used to work for a Hims competitor (Ro), so he knows the business from the inside. The episode title jokes that he "wouldn't trust the CEO to walk his dog," and he really is that skeptical of management. He's still long anyway, and the reason is the interesting part.

The thing the market is excited about is the thing he thinks doesn't matter. Most bulls are focused on how well Hims markets and ships drugs (especially the compounded GLP-1 weight-loss medicines, the class that includes Ozempic and Wegovy). Cerro says that skill is real but is just "table stakes": 'the company has not had an issue getting customers. They have had an issue keeping customers. And that is where they're bleeding.'
The GLP-1 boom was always on a timer. Hims could legally sell copycat versions only while the branded drugs were officially "in shortage." Cerro flagged in early 2025 that the shortage was ending, his team called pharmacies in 32 cities and could get the drug same-day or next-day, and the stock, which had "memed" from $25 to $72, promptly crashed. He covered his short around $14. As he put it, 'You exploit the loophole for as long as you can.'
The real bull case now is data. He thinks the value comes from Hims using lab tests and patient data to keep people subscribed longer, turning a customer who might churn after a year into a multi-year, multi-product relationship (testosterone today, cholesterol or other testing later). That lengthens the customer's lifetime value and shortens the payback on marketing spend, and it can scale internationally. In his words: 'look where the money is going right now. It's all data.'

The honest risk he raises himself: there's very little moat in telehealth, "literally anybody or any group of people can just start up a business, target a certain condition, and boom, there's a new telehealth company," plus lingering questions about management and about the dismissed (but re-fileable) Novo Nordisk patent suit.

2. Intuit ($INTU), long, Shawn O'Malley & Kyle Grieve

The Intrinsic Value Podcast (Investor's Podcast Network), August 5

The setup, in O'Malley's words: 'You've got a company with dominant market share growing revenues at double-digit rates, boasting operating margins north of 25%, and throwing off more than $5 billion of free cash flow... this same stock is the single worst performer in the entire S&P 500 this year, down nearly two-thirds from its high.' Intuit, the owner of TurboTax, QuickBooks, Credit Karma and Mailchimp, has fallen from over $800 to under $300.

What broke was the valuation, not the business. For a decade Intuit traded at a median of roughly 50x earnings. It now trades around 16x, a 50% discount to the S&P 500. As O'Malley notes, that implies the market thinks Intuit is lower quality than the average big company, which he finds hard to swallow for a business with 80% gross margins and ~30% operating margins.
The bear case is AI. The fear is that cheap AI tools make TurboTax (perhaps two-thirds of profit) obsolete, and hollow out QuickBooks too.
His rebuttal: the market is 'confusing the idea that AI can do a piece of a task with the notion that AI therefore destroys the entire business.' Intuit's moat, he argues, is a tangle of switching costs, proprietary data (bank feeds, millions of small-business ledgers you can't legally copy), distribution, trust and, crucially, accountability. If your taxes get audited, you can blame Intuit; you won't blame a chatbot. His line: 'a brilliant AI model with no data, no distribution and no trust is nothing.' Management is guiding to re-accelerate growth toward 20% by 2030, recently cut headcount 17% (revenue per employee up from ~$633,000 to an estimated ~$1.2 million), and is buying back ~$8 billion of stock.
Valuation: he pegs fair value near $400, wants a margin of safety, and would buy around $320 or lower, sizing it as a ~5% position, a "mean reversion" bet where the stock could simply double just by returning to a market-average multiple.

His own caveat: if TurboTax really does "go the way of Blockbuster" over the next decade, roughly half the company's earnings are at risk and today's price wouldn't look so exaggerated.

3. Exor ($EXO), long, as a cheap way to own Ferrari, Shawn O'Malley & Kyle Grieve

The Investor's Podcast, August 6 (re-aired on The Intrinsic Value Podcast, August 9)

This was a "hold ourselves accountable" update on an earlier pitch. Exor is an Italian holding company whose biggest asset is a large stake in Ferrari.

The idea: Exor's stock trades at roughly a 60% discount to the value of what it owns. So you can 'effectively acquire exposure to Ferrari's business at a substantial discount... more than 50%.' They bought at about $86 a share when NAV was about $193; the stock is now around $79. If that discount simply narrows from 60% to 30%, through buybacks or better sentiment, that alone 'would actually just be a double.'
Why they still like Ferrari underneath it: average revenue per car has climbed from about $239,000 in 2017 to about $446,000 today, a ~5% annual rise purely from pricing power, so Ferrari doesn't even need to sell more cars to grow. Management targets ~30% operating margins by 2029 (already guiding 29.5%), and R&D is tracking their ~13%-of-sales assumption.
The wrinkle: the discount has widened, not closed, since they bought. One reason is that Exor trimmed its Ferrari stake (well-timed, near Ferrari's peak) to free up cash for other deals, which annoyed shareholders who only wanted the "cheap Ferrari" story. Their defense: even if Ferrari just grows "decently," you're getting Exor's other investments as free call options. Timing the discount closing, they admit, "just comes down to making an educated guess"; this is a five-year-plus bet.

4. AST SpaceMobile ($ASTS), long, high conviction, "the Kook"

AST SpaceMobile Podcast ("The Kook Report, the $ASTS DD"), August 8

A read-through of the full investment thesis on a company building satellites that connect directly to an ordinary phone, no special hardware. The host has serious skin in the game: he bought call options (a leveraged bet on the stock) when ASTS was around $12; it later hit $40.

The one-sentence pitch: 'It allows you to connect your phone anywhere.' The bigger prize is low-cost satellite broadband with use cases far beyond "I'm hiking in Yosemite."
Why the business model works where earlier satellite phones failed: ASTS uses a "super-wholesale" model, it sells through carriers like AT&T and Verizon rather than selling you a second phone and plan. When AT&T pays for an ad campaign, that is ASTS's customer-acquisition cost being covered by someone else. He frames it as a potential "winner-take-most" business with commercial, patent and regulatory barriers to entry.
Recent de-risking: the "Block 1" satellites launched and deployed successfully in mid-June, and, the number that clearly excited him, spectral efficiency (how much data a slice of airwaves carries) has run well past his original assumption of ~2 bits per hertz to roughly 10 at peak: 'they've clearly advanced this technology far beyond any comprehension I previously had.'
The math he'd defend loosely, not precisely: at a $10 monthly revenue-per-user assumption and a range of adoption rates, he sees a path to $10 to $40 billion of revenue. He thinks the company does at least $10 billion, which at today's price is roughly 1 to 3x sales. He can "see a path for this thing to be $250 to $500 pretty soon," but adds that if it were already valued at $1 trillion he wouldn't own it, because that would price in more than he can see.

His candor is the best part: 'I'd rather own it and eat dirt than miss it because I was too cute.' He's transparent that he bought too early, that the business is capital-intensive and needs cheap financing, and that past dilution "sucked at the time."

5. Rocket Lab ($RKLB), a bullish "watchlist," not a buy-here call, Nicholas & Kasey Rossolillo

Chip Stock Investor Podcast, August 7

Their reaction to Rocket Lab's proposed $8 billion takeover of Iridium Communications. They like the strategy but are wary of the price.

Strategically it makes sense. Rocket Lab wants to be a vertically integrated space company, building, launching and operating satellites. Iridium gives it an existing 66-satellite constellation and, importantly, cash flow. Note: Iridium's low-frequency "L-band" network is built for reliability (voice, text, basic data anywhere on Earth), not the high-bandwidth broadband that Starlink or ASTS chase, so this is a complement, not a Starlink clone.
The deal fixes the finances. Rocket Lab flips from net cash to net debt to fund it, but Iridium generated roughly +$288 million of free cash flow over the trailing year versus Rocket Lab's -$316 million, taking the combined company to roughly break-even (about -$28 million), and to positive EBITDA of about $250 million. Rocket Lab's own revenue grew 63% year-over-year in Q1 2026.
The catch is valuation. Combined market value is ~$45 billion (enterprise value ~$50 billion including debt), which is under 30x sales and about 200x trailing EBITDA. Their reverse-DCF (working backwards from the price to the growth it implies) says you need ~35% annual EBITDA growth for a decade to justify today's value. Their verdict: 'This is not a value stock. This is very much a high growth stock', they missed the earlier run, it's on the watchlist, and they think there "could be some more downside."

6. The big AI-top debate: Michael Burry (short) vs. Steve Eisman (long)

Prof G Markets, August 6, and Eurodollar University, August 7

Not a single-name pitch so much as the week's central argument, and both men are famous for calling the 2008 crash.

Burry's bet: he's staying short NVIDIA, Micron, Tesla, Palantir and the semiconductor index, warning of a possible "1987-style" top.
Eisman's pushback (on Prof G Markets): the economy is "very strong," bank credit is "as benign as it possibly could be," and there's no crash without a trigger. The trigger he's watching is a possible price war among AI model makers: he cites a report that ~70% of Microsoft's, Google's and Amazon's AI revenue comes from just OpenAI and Anthropic, both losing enormous sums. 'If a massive price war broke out because of the Chinese models and OpenAI and Anthropic got in big trouble, that would unwind a lot of the AI trade.' Until then, with NVIDIA reporting August 26 and Amazon spending ~$225 billion on capex, 'I'm not on Burry's side.' His one real worry was Meta's last quarter: 28% revenue growth but 55% expense growth and just $785 million of free cash flow, which he called 'basically nothing.'
The bear's structural case (Eurodollar University, Jeff Snider): what's different this year is the credit cycle. He points to a string of alleged fraud and collateral blow-ups he calls "cockroaches" (Tricolor, First Brands, and now Radiant, with Jefferies exposed for a third time) plus rising junk spreads and investment-grade names like Oracle and Stellantis "trading more and more like junk bonds." His parallel to 1987 isn't the computer-trading crash itself but the credit deterioration underneath it. His verdict is measured: 'While we aren't there yet, more and more of the system continues to align in that direction.'

The useful takeaway for a stock-picker: the bulls and bears agree on where the fault line is (OpenAI and Anthropic economics, and private credit), and disagree only on timing.

7. Micron ($MU), long, "double from here," Adam Parker, Trivariate Research

Squawk on the Street, August 7

Parker was blunt: 'I think Micron's going to double from here... by the end of the cycle.'

His argument is about the balance sheet, not just earnings. The price today reflects roughly "four times peak and 10 to 11 times normalized earnings," and it assumes a sharp drop-off after the peak. He thinks the down-cycle could be 'an erosion, not an implosion,' and, his memorable framing, investors ignore that even a "paper boy" with a bad income statement is rich if he "wins the lottery" on the balance sheet. Micron guided to roughly $50 billion of revenue at an 86% gross margin, on track to generate close to $300 billion of free cash flow in coming years.
A caution he added: the market plumbing has gotten frothy, there are now ~540 leveraged or inverted ETFs (up from 75 in 2023), and about 64% of all options traded are same-day expiry. That fuels violent "$100 billion-plus moves in stocks," so he'd rather Micron "grind higher" than "rip higher."

Notably, Micron came up as a top pick in two separate podcasts this week (see the "Rules of Investing" item below), which is a rare cross-confirmation.

8. Ouster ($OUST), long, pitched by "Eric," discussed on Stock Club

Stock Club (Episode 323), August 6

A one-year check-in on a pick first made at the "Investicon 2025" conference. Ouster makes LiDAR, laser sensors that build a precise 3D map of a machine's surroundings, the "eyes" for robots, drones, self-driving vehicles and smart infrastructure.

The stock has roughly doubled since the pitch, from about $28 to about $45 (it briefly touched $64). Market value ~$3 billion.
The core of the thesis is a fast-improving business. Gross profit went from about $10 to about $43 per $100 of sensors sold; Q1 2026 was a record 12,600 units and the 13th straight quarter of growth. Its OS1 sensor earned a U.S. Department of Defense certification, a genuine competitive moat, not just another customer. 2026 revenue guidance was raised from ~$200 million to ~$220 million, and losses are narrowing (EPS guided from about -$1.37 to -$0.95).
Risks, stated plainly: still unprofitable, a crowded field (Luminar, Innoviz), and a lot of the recent move looks like momentum. The key catalyst was Q2 earnings, due the day after the episode aired.

9. Arminino Foods (frozen pesto, OTC), long, Ralph Folsom

The Acquirers Podcast, August 6

A wonderfully niche micro-cap story. Folsom put 100% of his net worth into Arminino Foods in 2021 (at age ~23 to 24, invoking Mohnish Pabrai's "bet big when you're young"), and later became a consultant to the company.

What it is: a Bay Area company selling frozen pesto to food-service customers (restaurants, cruises, hotels, universities). It trades over-the-counter and had famously not held a shareholder meeting in ~20 years.
Why he loves it: roughly 100% return on invested capital (excluding cash), a 50%+ dividend payout, ~6 to 8% annual top-line growth, net cash, no goodwill, and it has never issued stock. It just signed a lease for a 91,000-square-foot facility (up from 25,000) and has stated a desire to up-list to the Nasdaq.
The transferable idea: he screens first on capital allocation (net cash, no goodwill, never issued shares) because 'it's very hard to have good, stable, predictable, accretive capital allocation when you're in a bad business.' A colorful footnote: the company's old ticker was BLUE, from a failed Frank Sinatra pasta-sauce line called "Artanis" (Sinatra spelled backwards).

10. AI's "picks and shovels": Howmet, a Taiwanese cleanroom builder, and Micron, an Asia-focused fund manager

The Rules of Investing, August 7

A great example of following a supply chain until you find the bottleneck.

Howmet Aerospace ($HWM), the guest kept calling it "Helmet"; it's the Alcoa spinoff that makes turbine blades. He traced it by asking why Airbus can't deliver planes (not enough engines), then why engine makers (GE Aerospace, Pratt/RTX, Rolls-Royce) can't deliver (not enough blades), down to Howmet. Three long-term demand drivers hit at once: jet engines, gas turbines for data-center power, and fighter jets. Because customers now prepay to fund new capacity, he gets 'that immediate economic benefit... but moreover a company with an entirely different financial risk profile.' He bought under 30x earnings; it has since re-rated higher, so the debate now is when it gets too expensive.
A little-known Taiwanese cleanroom contractor (he called it "Acta"), under $5 billion market value, ~12x earnings, and doubling EPS. The insight: the number-one constraint every chipmaker cites is cleanrooms (essentially football-field-sized precision machines), and only a handful of Taiwanese firms know how to build them. Now that TSMC, Micron and others are building fabs worldwide, those contractors have "endless work coming after them," at margins roughly twice what they earn at home.
Micron was his "desert island, markets closed for five years" pick: 'they will be sitting on a pile of cash roughly 50 to 100% bigger than the current market cap', and you'd still own all the factories and expertise "for free."

11. Beaten-up brands and China internet, a veteran stock-picker

The Disciplined Investor (Episode 985), August 9

An "old-school stock-picking" set of longs, mostly buying things down hard with catalysts ahead:

Exxon ($XOM) and Chevron ($CVX): bought "when things were pretty calm in Iran," both up ~20%.
Nike ($NKE): down ~75% from its high, in the mid-$40s. New CEO, an incoming CFO who is "a turnaround specialist who turned around Levi's," easier upcoming comparisons, and insider buying (the CEO bought twice, "so did Tim Cook").
Lululemon ($LULU): similar story, down ~75%, new management starting August and September, easier comps later in the year.
Alibaba ($BABA) and China internet: bought since June, up 15 to 20% off the lows via names like Alibaba, JD, Baidu and the KWeb/FXI baskets. He argues the Chinese tech giants quietly own stakes in the open-source AI models, Alibaba owns 35% of Moonshot AI (recently valued at $35 billion) and its Qwen model is going into iPhones sold in China. He thinks Alibaba (~$100) is "a $200 stock."

He paired this with a warning about crowded trades, citing Korea's leveraged-ETF blow-up as a cautionary tale, and the reminder that the Bank of America survey has shown semiconductors as the most crowded "long" three months running.

12. Biotech long ideas, "Lou," a biotech specialist

Full Signal ("6 Stocks to Buy Now as Big Tech Burns Billions"), August 4

Halozyme ($HALO): ~11 to 12x forward earnings. It doesn't discover drugs; it licenses a technology that lets other companies' blockbuster drugs be injected under the skin in five minutes instead of via a two-hour IV, and collects royalties. He calls it 'the ARM Holdings of biotech.'
Eli Lilly ($LLY): long. Beyond weight-loss drugs, Lilly has made eight acquisitions this year to broaden its pipeline; he thinks it goes "from the first $1 trillion big pharma to the first $2 trillion." He views Novo Nordisk as a "value trap."
Viking Therapeutics ($VKTX): an explicitly speculative long, an oral-and-injectable weight-loss drug in late-stage trials that he thinks gets acquired by Lilly or Novo. He stresses position-sizing and suggests generalists use the XBI biotech ETF instead.
Broader view: he flipped positive on Amazon as the top Magnificent 7 name (best return year-to-date), and thinks Apple is a touch expensive after re-rating to ~36x earnings.

13. Three defined-risk "fades," Don Kaufman, TheoTrade

Schwab Network ("The Big 3: MRVL, NFLX, META"), August 5

These are short-term options trades with a capped, known maximum loss (a "put spread" profits if a stock falls; a "call spread" if it rises), tactics, not long-term theses:

Marvell ($MRVL), bearish. He calls the recent bounce a "bear market rally" after a ~50% decline and wants to fade it, using an October 175/165 put spread for a ~$3.10 debit.
Netflix ($NFLX), mildly bullish, short-term only. A September 77/80 call spread for ~$0.90, targeting a move to ~$80, after which "all bets are off."
Meta ($META), bearish. An October 550/540 put spread for ~$3.25, targeting a retest of ~$520. His fundamental gripe: Meta hired an AI "dream team" a year ago at pay rivaling "NFL starting quarterbacks" and has "heard absolutely nothing" in new products since, while capex keeps climbing.

14. Morningstar's "still like after earnings" list

The Morning Filter, August 3

Five quality names Morningstar's analysts flag as trading below their fair-value estimates:

Microsoft ($MSFT): $600 fair value, ~23% undervalued; modeled ~16% five-year revenue growth led by Azure.
Alphabet ($GOOGL): ~18% undervalued.
S&P Global ($SPGI): $505 fair value, ~17% undervalued; AI threat seen as overstated, >$7 billion of buybacks planned in 2026.
Charles Schwab ($SCHW): $124 fair value, ~15% undervalued; "a compelling growth at scale story."
Northrop Grumman ($NOC): $630 fair value, ~14% undervalued; early-stage programs (Sentinel missile, B-21 bomber) driving growth.

They also raised Amazon's fair value to $300 after AWS grew 37% (its fastest in 18 quarters), trimmed Apple's to $285, and lifted Baker Hughes to $73 on data-center and LNG orders.

Also worth a mention (lighter touches from the week)

Palantir ($PLTR), a cautious view. On Bloomberg Intelligence (Aug 4), Mandeep Singh acknowledged 93% growth but flagged that Palantir serves only ~1,000 high-value customers versus rivals' tens of thousands, leaving it exposed as OpenAI, Anthropic and the hyperscalers deploy similar tools.
Gold and mining juniors. The KE Report ran two idea-heavy interviews: Erik Wetterling (the "Hedgeless Horseman") laid out cheap gold developers, Revival Gold (RVG/RVLGF) at ~0.11x NAV at $3,000 gold, plus higher-risk exploration bets Westward Gold, Miata Metals (backed by a fresh 19.9% stake from La Mancha) and Red Canyon, while Nick Hodge discussed rare-earth recycler CoTec and Quartz Mountain. Mining Stock Daily and Company Interviews added CEO-pitched names (Resolute Mining, Summit Royalties, Adyton Resources). These are mostly small, speculative and, in the company-interview cases, company-sponsored, so treat accordingly.
Consumer "explainer" longs on The Best One Yet (Aug 3 to 7): Amazon ("the Berkshire Hathaway for tech"), Garmin (fitness division up 25%), Vita Coco ($COCO, up ~5x since its 2021 IPO), and Steve Madden ($SHOE), pitched on a "seasonless" strategy that avoids end-of-season discounting.
Old reliables: Money Tree Investing (Aug 5) liked Microsoft and Amazon as data-center winners and was negative on Apple; The Investing for Beginners Podcast (Aug 6) featured Thomas Chua long TSMC and Alphabet and cautious on debt-heavy Oracle; and on The Gwart Show (Aug 9), the discussion turned constructive on MicroStrategy ($MSTR) as it rebuilds cash reserves and buys back stock.