Newsletter · · Ashutosh Agarwal

A Hasbro Short and Power Infrastructure Longs Lead This Week's Podcast Stock Ideas - Weekly Podcast Idea Digest - Week of July 20, 2026

A cross-sector roundup of the specific long and short single-stock ideas fund managers and analysts pitched on podcasts for the week of July 20, 2026, from a detailed Hasbro short to power-infrastructure and AI names.

Weekly Podcast Idea Digest

Week of July 20, 2026: A Hasbro Short and Power Infrastructure Longs Lead This Week's Podcast Stock Ideas

This is our weekly roundup of specific, actionable stock ideas that investors and analysts shared on podcasts over the last seven days (July 20–27). We only include a name when someone made a real case for it, a clear long ("I'm a buyer") or short ("I'm betting it falls") with reasoning behind it. We skip the vague macro chatter.

One housekeeping note on the week's flavor: none of the big-name investor "idea conferences" (Sohn, Robin Hood, Delivering Alpha, Invest for Kids, and the like) actually met this week, so there were no formal stage pitches. The one exception to that spirit was a college stock-pitch competition. Almost everything else this week came from regular fund-manager and analyst interviews. The good news: it was a busy, high-quality week for concrete single-stock ideas. Here are the ones worth your time.

The conviction pitches

Hasbro (HAS), a short from two award-winning students who actually play the game

The most detailed single pitch of the week was a short (a bet the stock falls) on toy-and-games maker Hasbro, delivered by Vito and Glenn Lee, two Georgetown juniors who won both the Fordham and Notre Dame stock-pitch competitions with this exact idea, on the July 20 episode of Pitch The PM. Their target: about $56 a share, roughly 36% below where it trades, over a 12-month window. Notably, the show's host said he is short the stock himself: "I actually have a short in Hasbro. So nice job, guys. You've convinced me that you have an edge."

Here is what makes the pitch interesting. Most people think of Hasbro as a toy company: Nerf, Play-Doh, Monopoly, Transformers. It isn't, really. The students walked through the math: the toy and board-games division ("consumer products") runs at razor-thin margins, and roughly 90% of profit comes from Wizards of the Coast, the gaming arm, and within that, the trading-card game Magic: The Gathering is about 75% of operating profit. As Vito put it, the physical board games every household has in a cupboard are "less than 10% of the value of the business." Hasbro is, effectively, a Magic: The Gathering company.

Their thesis is that Magic's recent boom is unsustainable. The engine has been a strategy called "Universes Beyond", Magic cards built around outside franchises like Spider-Man rather than the game's own fantasy world. It juiced growth (management comped a monster year, up around 60%), but the students argue the core, loyal player base rejects it. They cited player dissatisfaction they pegged near 40%, versus management's claim of about 9%, wallet fatigue from 6–7% annual price increases, and (using alternative data from a firm called CarbonArc) average weekly app users down about 11% quarter-over-quarter. Their view: Magic historically grew low-single-digits (even flat to slightly negative in some years), the 60% year was a one-off, and Wall Street is still modeling high-single to double-digit growth on top of that inflated base. When they built "a quick and dirty five-year DCF" against Street numbers, "we couldn't get to any sort of upside." Honest note from them: this idea has started to work, the stock is already down about 12% since their April pitch, and short interest has roughly doubled year-to-date (from 3–4 million shares to 7 million), which the host flagged means it's no longer the "consensus long" they framed it as.

Light & Wonder (LNW), "a slot-machine oligopoly at half Aristocrat's multiple"

On Yet Another Value Podcast (July 24), Zack Buckley of Buckley Capital laid out a long case for Light & Wonder, a maker of casino slot machines. The pitch in one line: it's one of only three real players in the business (alongside Aristocrat and IGT), it's growing faster than Aristocrat, yet it trades at roughly half the valuation, 7–8x EBITDA versus Aristocrat's ~14x (EBITDA is a rough proxy for cash earnings; a lower multiple means a cheaper stock).

Buckley has traded this name well before: he bought around the low $40s in September 2022 and sold in the $80–100 range two years later (all in US dollars). He got back in after Light & Wonder moved its stock listing to Australia, which forced US funds that couldn't hold an Australian listing to dump shares, pushing it from about $100 down to the low $70s "with really no buying from Australia." That flipped, and the stock ran to about $120 before settling back to the high $70s. Why it's cheap again now, in his view, is simple timing: rival Aristocrat launched a wave of new games in the first half of the year while Light & Wonder's launches are back-half loaded, so recent market-share numbers look soft. It's a business with over 70% recurring revenue (casinos mostly lease the machines and Light & Wonder takes a cut of what players wager), which is exactly the kind of predictable stream that deserves a premium, not a discount.

The upside math: he thinks the company hits its $2 billion EBITDA target and around $1 billion of net income by 2028, worth roughly $13–15 of free cash flow per share. At Aristocrat's multiple that's about a $280 stock; even at a more conservative 15x it's around $210, "either way, we're talking about several hundred percent upside." The main risk he flagged is SciPlay, the free-to-play mobile slots app (about 20% of EBITDA), where competition and AI could erode the moat over time, but he argued that's a small headwind, not a thesis-breaker.

Perimeter Solutions (PRM), the fire-retardant near-monopoly run like TransDigm

Two of our value-investing shows converged on the same name this week: Perimeter Solutions, covered in depth by hosts Kyle Grieve and Shawn O'Malley on both The Investor's Podcast / We Study Billionaires (TIP833, July 23) and its sister show The Intrinsic Value Podcast (TIVP086, July 26).

The appeal is the pedigree and the price. Perimeter is run on the playbook of TransDigm, the serial-acquirer that compounded shareholder money for decades, and its architects are TransDigm founder Nick Howley and William Thorndike, author of the investing classic The Outsiders. They screen for the same five traits TransDigm used: recurring revenue, secular tailwinds, high-value-but-low-cost products, high returns on tangible capital, and growth through acquisition, targeting private-equity-like returns of 15%+ a year.

The crown jewel is fire safety, the fire retardant dropped from planes onto wildfires. It's close to a US monopoly. When it was bought in 2021 it did $261 million of revenue and $118 million of EBITDA (a ~17x purchase price); on a trailing basis it now does about $500 million of revenue and $290 million of EBITDA, with margins expanded from the mid-40s to about 60%. Yet today the whole company (roughly a $5.5 billion market cap) trades at around 7x EBITDA, as the hosts put it, "insanely cheap for a business that is a near monopoly." The wrinkle: management has been diversifying via acquisition (a medical-manufacturing business, MMT, bought for $685 million in December 2025), which shifted the revenue mix toward specialty products and makes the story harder to model. And fire-safety earnings are genuinely lumpy, margins have swung from 27% to 65% quarter to quarter because a light wildfire season hits revenue while fixed air-base costs stay put. On the companion episode, the hosts pegged a rough fair value near $46 a share with a margin of safety, while flagging concerns about the founders' advisory fee and some litigation.

Constellation Software (CSU), buying the "software apocalypse"

On The Investor's Podcast's "Richer, Wiser, Happier" episode (RWH070, July 26), investor Christopher Begg of East Coast Asset Management (interviewed by William Green) explained a purchase that would normally be off-limits for a value investor because it's never cheap: Constellation Software, the Canadian acquirer of hundreds of niche vertical-software businesses.

Begg's whole method is buying great businesses when a "cloud" of fear temporarily mis-prices them. He said the biggest cloud of 2026 is what he calls "the software apocalypse", the fear that AI will disrupt software companies wholesale. That fear knocked Constellation down about 50% from its highs in the first quarter of this year, and he used it to buy a business he'd studied for over a decade. To separate real risk from perceived risk, his team built a framework grading eight "layers" of a software moat, from the shallow (a simple user interface AI can copy) to the deep (how embedded the software is in a company's daily operations, institutional memory, security, and trust). The acronym, fittingly, is IMMORTAL. His conclusion: Constellation's businesses sit at the deep, hard-to-disrupt end, and the key thing to watch is customer churn: "if we were to see churn numbers accelerate... we'd start being concerned," but so far "the hypothesis is holding." He noted the same playbook worked when he bought Google at 15x earnings during an earlier AI-fear cloud that has since lifted.

Crown Castle (CCI), betting 2026 is the bottom

On Hedge Fund Tips with Tom Hayes (July 23), Tom Hayes of Great Hill Capital made the income-and-recovery case for Crown Castle, now the only pure-play US cell-tower landlord after selling off its fiber and small-cell business. The pitch is a stock that "becomes increasingly hard to ignore as rates come down over time and capital rotates back into yield."

The just-reported quarter did the heavy lifting: site-rental revenue of $967 million beat the $942 million consensus, and AFFO (a REIT's version of cash earnings) came in at $1.13 a share, up 11% year-over-year. Crown Castle closed the $8.5 billion sale of its small-cell and fiber unit on May 1, two months early, and used the proceeds exactly as promised: over $7 billion of debt paydown and $1 billion of buybacks at an average of $88.66. The overhang is DISH: its bankruptcy created a revenue headwind, but Crown Castle is pursuing a $3.5 billion contractual claim through the bankruptcy court and sits on the creditors committee, and full-year guidance assumes zero from DISH. Hayes's core argument is that management calls 2026 the trough year for organic growth, with acceleration after Sprint-related churn rolls off. He also pushed back hard on the fashionable fear that satellite phones make towers obsolete: 90% of mobile use happens indoors or in vehicles where satellite signals struggle, satellite covers vast areas with far less capacity, and, tellingly, carriers haven't changed a single rural build in response. Longer term, he sees a path to closing an ~11% cost gap versus rivals American Tower and SBA (Crown Castle owns the land under only 43% of its towers) for 200+ basis points of margin upside.

Power and AI infrastructure: where the money is rotating

Two shows this week made essentially the same argument, that the AI trade is quietly rotating out of memory chips and into the companies that supply electricity and power equipment.

GE Vernova (GEV), "the TSMC of power"

The clearest version came on Limitless (July 22), where the hosts called GE Vernova (the power-turbine and grid-equipment maker spun out of GE) "the TSMC of power." Their case: as data centers scramble for electricity, GE Vernova sells the turbines and grid gear everyone needs, and its order book is booked out to 2031, making revenue unusually predictable. They cited 2025 orders that doubled year-over-year to $7.1 billion, roughly 30% annual growth (which they think could go "exponential" over the next 6–12 months), a stock up about 300% in three years, and marquee customers including a $7 billion Microsoft relationship and OpenAI. The backdrop stat they leaned on: US data-center power demand is on track to roughly double in 24 months, from about 31 to 66 gigawatts.

A more sober, analyst-style read came the day after on Wall Street Unplugged (July 23), where the hosts noted GE Vernova actually missed earnings by 70 cents ($2.47 vs. $3.17 expected) and didn't raise its profit guidance (which knocked the stock) even as revenue rose about 22% and gas-turbine output is set to climb from 20 gigawatts this year to 24 by 2028. Their takeaway was to buy it on pullbacks rather than chase it, but the demand story is intact.

Alphabet (GOOGL), "in a league of its own among hyperscalers"

Also on Wall Street Unplugged (July 23), the hosts argued you shouldn't lump Alphabet in with the other big AI spenders. Yes, Alphabet's massive data-center build turned its free cash flow negative in Q2 (about -$5.9 billion) as capital spending surged to $44.9 billion. But, they countered, that's the lazy read: Alphabet still generated $53.3 billion of free cash flow over the last 12 months and sits on $242.5 billion in cash and securities against $98.2 billion of long-term debt. Their clinching point was the shareholder register: Alphabet raised $80 billion in a rare equity offering, and Berkshire Hathaway bought $10 billion of it in a private placement: "if Berkshire Hathaway is one of your largest shareholders and they are committed to you long term... do you think they'd be deploying billions if they thought this AI buildout was just farce?" They've taken profits but wouldn't write the stock off on a pullback. (Christopher Begg, in the Constellation discussion above, separately confirmed Google is still one of his largest holdings.)

The short sellers and skeptics

Supermicro (SMCI), fading the pop

On Schwab Network (July 22), options trader Don Kaufman of TheoTrade said he was actively shorting Super Micro Computer even as it popped 24.5% on strong preliminary results, a record $60 billion backlog and nearly doubled margin guidance. His reasoning is reputational, not fundamental: the company is "just ripe with controversy, scandals... a number of discussions of even fraud," and "any opportunity I can get to fade this marketplace, I'm going to take": the more it rallies, the better his entry. His mechanic was a defined-risk put spread expiring September 18 (buy the $26 puts, sell the $21 puts) for about a $2 debit, so his downside is capped.

SpaceX, a fresh short target, and a way to buy it

The just-public SpaceX drew opposite views. On The Julia La Roche Show (July 21), George Noble (CIO of Noble Capital Advisors) called it a short, arguing its ~90x revenue valuation will crack as the tradeable share count expands "from 5% to 100% over the next 6–12 months", in his words, the decline is "a feature, not a bug." On the flip side, Don Kaufman (same Schwab Network segment as above) said he wants to own it on weakness (the stock has fallen from a $150 open to about $124) and would rather get there by selling an August 21 $100 put for ~$4.60, effectively agreeing to buy near $95 if it keeps dropping. (Reminder: SpaceX is a newly listed, thinly-traded name with barely a month of price history, treat all of this as unusually speculative.)

General Motors (GM), a genuine bull/bear split

Two shows landed on opposite sides of General Motors within a day. On Schwab Network's Ca$htag$ segment (July 20), Landon Swan of LikeFolio leaned bearish into earnings on a data divergence: his consumer-demand signal for GM is down about 27% year-over-year while the stock holds around $76 (near analyst targets of ~$96) at a price-to-earnings ratio of roughly 6. His worry is that all the good news (dividends, buybacks, cheap multiple) is well known, while the falling-demand signal is not, and "you don't get a lot of these crossing patterns where demand is falling off and the stock kind of raised." Then GM reported, and on Brew Markets (July 21) the read flipped: the company beat, raised full-year profit guidance for the second time, and grew North America profit 43% year-over-year by leaning into higher-margin trucks and SUVs, with the CFO calling the stock a bargain. A clean example of a fundamentals-vs-alternative-data debate playing out in real time.

Also heard this week

  • MicroStrategy (MSTR), long. On The Wolf of All Streets (July 20), multiple guests argued MicroStrategy trades below intrinsic value at roughly 1x its Bitcoin net asset value, with Bitcoin-per-share compounding ~65% a year since 2020; a TD Cowen analyst valued its digital-credit business alone at about $8 billion (4x revenue), a quarter of the company's value.
  • Harrow (HROW), long. On Telltales (July 22), value investors made the case that this small ophthalmology drugmaker has big free-cash-flow growth ahead as it commercializes drugs and cross-sells through its eye-care sales force, buying rights to sub-blockbuster products that big pharma won't chase, a niche with little competition.
  • Merck (MRK) and AbbVie (ABBV), long (analyst ratings). On Power Lunch (July 23), BMO's Evan Seigerman put a buy on Merck ($142 target), crediting management's plan to offset Keytruda's looming patent cliff via acquisitions, and rated AbbVie outperform ($300 target) on its inflammation franchise and the Apogee deal.
  • Intel (INTC), long. On the Chip Stock Investor Podcast (July 21), hosts argued Intel's EMIB and Foveros chip-packaging technology is a real edge over TSMC's approach (simpler and cheaper), positioning it to win AI-inference work as customers like Google look for alternatives to Nvidia, provided Intel can finally execute on the financials.
  • Gold miners and psychedelics, long. On The Julia La Roche Show (July 23), Chris Irons made a contrarian long case for beaten-down gold miners (the GDX ETF, down from ~$120 to ~$75) alongside a longer-term view of gold reaching $7,500+, and named psychedelic-medicine stocks Compass Pathways (CMPS) and MindMed (MNMD, up ~240% year-to-date) on positive trial data and FDA fast-tracking.
  • Junior gold miners, long. On Company Interviews (July 20), portfolio manager Michael Gentile discussed buying while the sector is hated, highlighting Radisson Gold (RAD) trading near $100/oz in the ground on a 4-million-plus-ounce, high-grade resource sited next to existing mills.
  • What a great investor just sold. On The Acquirers Podcast (July 23), Artisan's David Samra explained why he exited Samsung Electronics after a multi-year hold, he bought it in down-cycles when the market ignored its leading-edge memory capacity, but recently sold because today's valuation implies gross margins above historical norms in the AI boom, tilting risk-reward against it. A useful reminder that a great buy eventually becomes a sale.