| Account | Shares | Price | Value | % of acct | Cost/sh | Gain $ | Gain % | Target |
|---|---|---|---|---|---|---|---|---|
| 401K | 367 | $339.60 | $124,633 | 5.09% | $155.83 | $67,443 | +117.9% | — |
| ROTH | 5 | $339.60 | $1,698 | 0.66% | $190.00 | $748 | +78.7% | — |
| Total | 372 | $126,331 | 2.81% | $68,191 | +117.3% | — |
In short: Cited as the best of the three 2024 breakout calls — a retrospective, not a fresh rating; it remains a Holds/Trims position rated H on the published table (lot 07/22/2024 @ $168.68, marked $334.93, +98.56%). The original mention sits in the same February-2024 note as MPLX: "we also expressed our fondness for Google (GOOG; forget that stupid 'Alpha' handle), despite, or because of, what we thought was an overreaction to a decent earnings release back in February of 2024. Though we didn't publish a price chart for GOOG back then, we did convey that it, too, had achieved an upside range expansion." The self-criticism is the point: "ironically, because that was the one graphic we didn't run at the time, GOOG's chart was the most bullish of the three. Our hesitance was due to the fact it hadn't clearly broken out yet. As you can see, it did so in the spring of 2024, and after that it was moonshot time, with a couple of brief altitude losses, primarily around Mr. Trump's infamous Liberation Day a year later." Used to argue the general rule rather than the name: "the main reason for this tangent was to once again reinforce to you, precious reader, the importance of these events" — range expansions above multi-year ceilings — against readers who "feel we are either misguided or obsessed, or both, with this subject."
Alphabet appears here as a piece of evidence rather than as a recommendation. In the same February 2024 note that flagged MPLX, Haymaker also said he liked Google — bought, in effect, because he thought the market had over-reacted to a perfectly decent earnings report. The difference was that he published breakout charts for Meta and Microsoft but not for Google, because Google had not quite cleared its old ceiling yet.
With hindsight that hesitation was the mistake: of the three, Google's chart turned out to be the most bullish. It broke out in spring 2024 and, as he puts it, "after that it was moonshot time," interrupted only briefly around Trump's April 2025 tariff announcement. The point of retelling this is not to re-rate the shares — it is to defend the technique. Haymaker's argument is that when a stock pushes decisively above a price ceiling that has held for years (an "upside range expansion"), that is a real signal, and readers who think he is obsessive about it should look at what the three charts did.
For the record, Alphabet is still a Haymaker holding, sitting on the Holds/Trims table rated Hold, with a July 2024 cost of $168.68 against a current $334.93 — roughly a double. No change was made to it this week.
In short: Cited as the earnings-quality exhibit, not as a rated call. Named first among the mega-caps "that have been reporting blockbuster earnings," then used to make the specific point that those earnings are not all operating: "as several recent Haymaker notes have pointed out, in addition to robust organic growth, this is a function of companies such as GOOG reporting massive gains on their holdings in AI entities like Anthropic. In GOOG's case, those have represented more than 50% of its profits in the first two quarters of this year." The characterisation is blunt — an "immense, though unsustainable, earnings booster rocket" — which is what underwrites the "looming mean reversion in the S&P 500's profit margins" and the normalized-margin overvaluation conclusion. Continues the accounting-quality thread of Jul-15 (Minack's "earnings are an opinion; cash is a fact") and the Feb-20 trim, but no rating, target or portfolio action is attached here; the name is evidence, not a recommendation.
Alphabet is Google's parent — search, YouTube, cloud computing and a large stake in the private AI company Anthropic. It appears here purely as evidence, not as a recommendation.
Hay's point is about how its profits are made up. When a private company you own a piece of is revalued higher, accounting rules let you book the increase as a gain, and that gain lands in reported profit alongside money you actually earned selling things. In Alphabet's case those gains "represented more than 50% of its profits in the first two quarters of this year."
That matters for the market as a whole, because the S&P 500's impressive earnings growth — the thing that justified its 50% return since March 2024 — is partly built from this. It is a real accounting number and an unreal economic one: it stops the moment private AI valuations stop climbing, and it reverses if they fall. Hay calls it an "immense, though unsustainable, earnings booster rocket," which is why he expects the index's profit margins to revert toward normal, and why he prefers medium-sized companies that never had the rocket in the first place. No view is taken on Alphabet's shares here.
In short: Bought at the value end of the growth-to-value journey. "The next step is a value stock, like Google ($GOOG) was when everyone thought AI would kill it. Because I follow AI closely, I bought a position in Google at that point in time." The edge claimed is domain knowledge rather than valuation work — following the technology closely enough to judge that the disruption narrative was wrong. No entry price or current view is given.
Alphabet owns Google search, YouTube, Android and the Google Cloud business. It appears here as the final stage of the guest's "growth to value" idea: a company once considered a growth stock that had become cheap enough to count as a value one, because the market had decided AI chatbots would destroy the search business.
He bought it at exactly that point, and the reason he gives is worth separating from the outcome: not that the shares were statistically cheap, but that he follows AI closely enough to judge that the disruption story was wrong. That is a claim to domain knowledge rather than to valuation skill — a different kind of edge from the one this archive normally argues for, and one that is much harder to check from the outside.
In short: "Google reports its first quarter of negative cash flow. Is this still a company that deserves a premium valuation?" The buildout is now being funded with borrowed money — "the companies are shifting to debt to continue their buildouts… not sure this will end well. When has piling on debt to invest in a commodity-type investment worked out long term? I guess they will continue to borrow and build until the money runs out. We have seen this movie before: fiber, shale, housing." The value, in his frame, is transferring away from the hyperscalers to the semiconductor makers.
"Free cash flow" is the cash a business has left after paying for everything it needs to keep running, including new buildings and equipment. For twenty years Alphabet was the archetype of a company that generated enormous amounts of it, which is why the market paid a premium price for the shares. This quarter it generated none — the AI data-center build swallowed more cash than the search business threw off — and Polomny asks the obvious follow-on: "Is this still a company that deserves a premium valuation?"
His deeper objection is about what the money is buying. Data-center compute, in his view, is a commodity: undifferentiated capacity that anyone with capital can add, so the extra supply eventually competes the price down. Funding that kind of asset with borrowed money is the specific pattern he distrusts — "when has piling on debt to invest in a commodity-type investment worked out long term?" — and he names three precedents where exactly this happened: the late-1990s fiber build, the US shale boom, and the housing bubble. Each ended with the capacity still standing and the lenders and shareholders taking the loss.
The corollary is where the value goes instead: a "wealth transfer from hyperscalers to semiconductor makers." The people selling the picks and shovels book the spending as revenue while the buyers book it as depreciation and debt. He does not name a chip stock to buy here, and he is not calling a top on a date — his expectation is that "they will continue to borrow and build until the money runs out," which is the same "it goes on until it can't" framing he has used since calling data centers "the new shale oil."
In short: "Google's business without an AI division will be obsolete in 6, 18 months" — so it's "hamstrung," forced to spend the capex. That will show up as negative cash flow; "they can only use creative accounting for so long" before a multiple rerating. Part of a Mag 7 that could see a "70, 80% correction."
Polomny thinks Google is trapped: it has to pour enormous sums into AI data centers or "be obsolete in 6 to 18 months," but that spending will start showing up as negative cash flow. For a while accounting choices can hide it, but eventually the market sees these are no longer the fat-margin, capital-light businesses they were — and the stock's valuation has to reset lower. He lumps it with the rest of the Mag 7 as a candidate for a big (70–80%) drawdown, like the dot-coms in 2000.
50:04This is not the same thing. The infrastructure needs to be built. And if you're looking at the quarterly filings of these, call it the Mag 7 or 8 or whatever, that the Googles, the Microsofts, all these guys — that by force they have to spend the capex on this thing or they will be left behind. Google's business without an AI division will be obsolete in 6, 18 months, something like that.
In short: The asset-light-turned-heavy-industry tell: Google's once-90%-margin software business is now "building power plants, substations, talking about nuclear reactors, buying gas turbines — this is not their business… low margin." And "Google is going out and borrowing money now to sustain this" via a bond auction — confirming the negative-ROIC capex theme. (See 6.17 for the GOOG equity-issuance angle.)
The warning is about what Google is becoming. Its core business was software: build it once, and each extra user costs almost nothing — that's why margins were ~90%. But to win the AI race it's now spending like a heavy-industry company — building power plants, substations, even talking about nuclear reactors and buying gas turbines. Those are low-margin physical businesses, not 90%-margin software. The tell he points to here: Google had to borrow $10 billion in a bond sale to fund the buildout — a sign the spending has outrun even its enormous cash flow. (His related 6/17 note adds Google's first big stock issuance since 2004.)
19:43As Zuckerberg said, it's an arms race. Google — these were high margin businesses before this. Asset light, high margin. The incremental cost to — once you set up your software and your server so somebody can download Adobe or whatever is almost zero after you do the initial investment, create the software, create a service so people can download it. Adding additional customers at $99 a year or whatever they charge is zero.
In short: AI datacenters are "deeply negative ROIC investments," yet hyperscalers build them anyway — used up annual cash flow, took on debt, and now "Alphabet (GOOG) has even resorted to its first major equity issuance since 2004" to fund them. Datacenters = "the new shale oil" (the negative-ROIC drilling analog); MAG7 "no longer the high-margin, asset-light businesses they once were." Profitless growth in a revolutionary industry "goes on until it can't," then "the crash and the reallocation of capital."
Polomny isn't shorting Alphabet — he's flagging a warning sign. "ROIC" (return on invested capital) is just: for every dollar a company sinks into a project, how many cents of profit does it earn back? His claim is that AI datacenters earn back less than they cost — a money-losing use of capital — yet Google keeps building them. The tell he points to: Google has paid for this buildout first out of its own cash flow, then by borrowing, and now by issuing new stock for the first major time since 2004. Selling new shares dilutes existing owners, and a famously cash-rich company only reaches for it when the spending has outrun what cash and debt can cover.
His bigger frame: the "MAG7" mega-caps used to be lean, high-margin software businesses; now they're pouring money into datacenters and power plants like a heavy-industry company, but the profits from all that spending aren't showing up. He compares it to last decade's shale-oil boom, where drillers kept sinking money into wells that didn't earn their cost back — growth for its own sake that "goes on until it can't," then ends in a bust and a rotation of money into cheaper, out-of-favor sectors (which is where he wants to be positioned).
In short: #6 on the Q4 2025 aggregate Top Buys table (the class C line, listed separately from GOOGL at #9 on the two-quarter table). Data point only.
In short: Trim — a past Haymaker favorite (written up at ~$140 in Feb-2024 alongside TSM), now ~$315, "on the very spendy side" on its 5-yr P/S and P/E charts. Already gain-harvested on the way up; cited (with TSM/XOM) as why "so many valuation charts look like this" warrants raising 20–25% cash. A take-some-off, not a sell-out.
In short: A news data-point reinforcing the nuclear-demand thesis, not a stock view: "Google CEO eyeing electricity from nuclear plants for its data centers, just like everybody else." Cited alongside the $14T bank coalition and Japan as evidence that "the entire zeitgeist has changed" toward nuclear power.
12:18And we're starting to see maybe the initial idea around that because there's a lot of money involved here and a lot of big players. So, it says here, Google CEO eyeing electricity from nuclear plants for its data centers, just like everybody else. And then we had this, Mark Nelson put this on X, breaking in New York City, world's biggest banks pledge support for nuclear.
Nothing matches this filter.
Verbatim excerpts from the public transcripts (auto-pulled at each mention's timestamp, lightly cleaned). Timestamps link into the video; "source page" opens that commentator's full analysis at this row. Click a mention's header line to expand it (one open at a time). For personal study — not investment advice.