By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panel consensus is that the fund, MISEX, is dangerously concentrated with 40% in Alphabet, and while the management quality thesis is sound, the high concentration and lack of disclosure on fees and potential tax advantages pose significant risks.
Risk: Concentration risk, specifically the 40% exposure to Alphabet in a $27M fund, and lack of fee disclosure.
Opportunity: None explicitly stated.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
While Alphabet is right in the middle of the artificial-intelligence build-out by offering a chatbot, cloud services and microchips, fund manager Thomas Winmill points to another business in the tech giant’s toolbox that gets less credit: its millions of YouTube videos.
“I just don’t think there’s going to be any other business that’s going to be able to do anything about that. People I know, thanks to YouTube, can learn how to re-side their house and fix their own car. And [Alphabet makes] so much money off that particular business alone,” the manager of the $27 million Midas Special Opportunities fund MISEX told MarketWatch in an interview.
Of course, there’s far more to the fund’s 40% stake in Alphabet GOOGL GOOG than DIY videos. Winmill, a former lawyer, took over the fund in 2012 from his father, who bought the tech giant, then called Google, and the second-largest holding, Mastercard MA, early in the 2000s and never sold. The average cost basis of those stocks for the fund is a mere $9.56 for Alphabet and $9.37 for Mastercard.
“What we like to see in the numbers are a lot of free cash flow and then the ability of that company to reinvest free cash flow in order to get the compounding effect, and Alphabet has the ability to do so,” he said, adding it has an “unassailable advantage in growth.” He said he favors Mastercard for many of the same reasons.
Longtime holding AutoZone AZO was cut a couple of months ago after the auto-parts chain failed to generate revenue growth. “They were basically investing in a business that was slowing down by buying back their stock, taking on debt. Those are bad signs,” he said. “And then they’re going to lower-margin markets like expanding into Central and South America.”
AutoZone was replaced with “terrific names,” like Microsoft MSFT, which had historically been too expensive for the fund. The fund bought Microsoft in early April when shares dropped to around $370 on fears AI was going to dislodge it, buying it at around 23 times earnings.
The fund also added Constellation Software CA:CSU, another stock that Winmill viewed as unfairly sold off this year, buying it for 17 or 18 times earnings.
Lam Research LRCX, J.P. Morgan Chase & Co. JPM, Williams-Sonoma WSM, Berkshire Hathaway BRK.B and Progressive Corp. PGR are also in the portfolio, with that last name a “cheap” favorite of his.
“They’re growing, and they have a dynamic approach to pricing in the market,” along with a solid history and focused management, he said of the insurer.
As for his Berkshire holding, he noted the results haven’t been “as dynamic as we were hoping,” with a “subpar” return on assets. However, given Berkshire’s vast cash flow, he said the stock will probably “perform in the next financial crisis.”
Winmill said his bets rely on smarts at the tops of companies. “One thing that we can be pretty sure of, is that the management of Alphabet, Mastercard, Lam, JPMorgan, Williams-Sonoma, Constellation, Progressive and Microsoft have the resources to attract probably some of the smartest people in the world.”
“I feel pretty comfortable that, when my day ends, they’re probably still having late-night meetings and working out how to increase their competitiveness and what they have to offer,” he said.
As for the AI narrative and the excitement it’s created, he is concerned that revenue for some of those companies will end up banking on short-term product cycles.
“I’ll miss the moonshots. Absolutely. I’m not worried about that. What I’m trying to do is invest my money, and this fund, in companies where I know they have one of the best management teams in the world,” he said.
The National Federation of Independent Business reported a further fall in business optimism. Data on the U.S. trade balance for April was $55.9 billion in April, down from a revised $56.6 billion in March. Existing-home sales and wholesale inventories at 10 a.m.
Mark Newton, head of technical strategy at Fundstrat, has warned that the bounce in stocks following Friday’s rout is likely a dead-cat bounce because more stocks fell than rose on Monday, with reduced volume. He also points to a downturn via his chart of the S&P 500 Cycle Composite — a predictive market model combining historical seasonal and calendar trends. He said that cycle forecast a peak around current levels but has now “rolled over, turning negative between now and at least late July with the risk of lower prices into October.”
Top tickers
These were the most-searched ticker symbols on MarketWatch as of 6 a.m.:
Four leading AI models discuss this article
"Concentration risk in Alphabet makes the fund highly sensitive to one-name performance, risking outsized drawdowns if the ad business or regulatory backdrop worsens."
On balance, the piece promotes a best-management teams theme with a heavy tilt to Alphabet and a slate of quality compounders. The strongest risk is concentration: about 40% of the fund in Alphabet, plus exposure to MSFT, MA, CSU, JPM, LRCX, WSM, BRK.B, PGR. In a downturn or regulatory/ad-cycle disruption, relying on a small handful of names could erase years of outperformance. The piece understates the potential downside of AI hype cooling, valuation compression, and how macro cycles could dent free cash flow reinvestment in the biggest holdings.
The strongest counter is that Alphabet alone could drive the whole fund's fate; a material ad-revenue or regulatory shock would wipe out performance, making the strategy dependent on one mega-cap rather than skill.
"The fund’s 40% concentration in Alphabet creates a single-point-of-failure risk that undermines the manager’s stated goal of relying on a broad 'best-in-class' management cohort."
Winmill’s strategy is a classic 'quality at a reasonable price' (GARP) approach, but the portfolio construction is dangerously concentrated. With 40% of MISEX in Alphabet, the fund is effectively a proxy for Google’s regulatory and AI-execution risk rather than a diversified mutual fund. While his focus on free cash flow and management pedigree is sound, his dismissal of 'moonshots' ignores that Alphabet’s valuation is heavily predicated on the success of its own high-risk AI bets. He’s essentially betting on a legacy cash-cow moat—YouTube and Search—to fund a transition that remains unproven. This is a high-conviction bet on institutional inertia over disruptive innovation.
If Alphabet’s search dominance and YouTube moat are as 'unassailable' as Winmill claims, the fund’s concentration is actually a rational hedge against the volatility of speculative AI pure-plays.
"The fund's track record is real, but it's a product of inherited cheap basis and 20-year compounding, not replicable stock-picking skill—and current entry prices (MSFT at 23x, CSU at 17-18x) suggest Winmill is now paying market rates for the same quality thesis."
This is a profile of a concentrated, low-cost-basis fund that works—but only because of path dependency and survivorship bias. Winmill inherited Alphabet at $9.56 and Mastercard at $9.37; those positions are now 40% of a $27M fund. The article frames this as brilliant management-team betting, but it's really a 20-year compounding story that couldn't be replicated today. His recent moves—buying Microsoft at 23x earnings and Constellation at 17-18x—suggest he's now paying full price for quality after decades of getting it cheap. The technical warning buried at the end (S&P 500 cycle 'rolled over' into late July with October downside risk) contradicts the bullish framing entirely.
Winmill's thesis—that best-in-class management compounds free cash flow into durable moats—is empirically sound and his holdings (Alphabet, Mastercard, JPM, Progressive) have actually delivered. The real risk isn't his stock picks; it's that a $27M fund with concentrated positions and no disclosed fee structure has no incentive to grow, meaning this 'outperformance' may simply reflect that he can hold illiquid conviction bets that larger funds cannot.
"MISEX's edge is largely an artifact of two-decade-old cost bases rather than a repeatable process that can be scaled or sustained at current valuations."
MISEX's 40% Alphabet stake and Mastercard holding, acquired at $9.56 and $9.37 respectively, deliver compounding via high free-cash-flow reinvestment, yet the $27 million fund's low turnover and recent swaps (AZO out, MSFT at 23x and CSU at 17-18x in) expose concentration risk. Winmill's management-quality thesis ignores how Alphabet's YouTube margins could face ad-cycle compression or antitrust remedies, while the dead-cat bounce warning from Fundstrat suggests near-term equity weakness that could pressure even these compounders. Small AUM also limits scalability of the approach.
The low historical cost basis lets the fund hold through drawdowns that would force sales or redemptions at larger or newer vehicles, so outperformance may not survive once legacy positions are exhausted.
"A 40% Alphabet weight exposes the fund to outsized downside from regulatory/ad-market shifts, amplified by liquidity constraints that could force forced selling."
Claude, survivorship bias aside, the real overlooked risk is liquidity and fragility of a 40% Alphabet tilt in a $27M fund. A sudden redemption spike or antitrust/regulatory shock could force outsized losses in a single name and force rapid rebalancing, something a tiny fund struggles with. Also, undisclosed fees may erode risk management incentives. The thesis assumes a durable, multi-decade moat; a policy or ad-cycle reversal could reverse quickly.
"The fund's performance is largely an artifact of extreme tax-deferred cost basis rather than superior active management."
Claude, your point on cost basis is the pivot. The fund isn't a 'strategy'—it's a tax-advantaged legacy holding vehicle. By holding these winners for decades, Winmill avoids the capital gains drag that forces larger funds to rebalance, artificially inflating his 'outperformance' metrics. This isn't active management; it's a passive tax-shielded bet on two mega-caps. If he were forced to rotate into current market valuations, the 'quality' thesis would evaporate under the weight of tax friction and lower entry yields.
"Tax efficiency explains *some* outperformance, but not all—and we don't know the fee structure or whether returns are net or gross."
Gemini's tax-shelter framing is sharp, but it conflates two things: Winmill's *personal* tax advantage (real) versus the fund's *reported* outperformance (which should be net of fees and taxes to investors). If MISEX's returns are already after-fee, after-tax, then the 'artificial inflation' claim collapses. The real question: what's the fee structure, and are returns gross or net? Nobody's asked. That's the missing fact that either validates or demolishes this entire thesis.
"Fee opacity prevents separating genuine skill from structural advantages in this illiquid, concentrated vehicle."
Claude correctly identifies the unreported fees as pivotal, but this gap also exposes Gemini's tax argument. Net-of-fee returns would validate skill only if they account for the illiquidity premium in a concentrated $27M portfolio. Absent that disclosure, the outperformance could simply reflect an inability to scale or rebalance without destroying value during volatility spikes like the Fundstrat-predicted October dip.
The panel consensus is that the fund, MISEX, is dangerously concentrated with 40% in Alphabet, and while the management quality thesis is sound, the high concentration and lack of disclosure on fees and potential tax advantages pose significant risks.
None explicitly stated.
Concentration risk, specifically the 40% exposure to Alphabet in a $27M fund, and lack of fee disclosure.