Chaikin's '100X Starburst Opportunity' — The Three Stocks Behind the Frontier AI Pitch

The Hook

Another “Frontier AI” pitch landed in inboxes recently, this time from the Chaikin Power Gauge Report. The framing is ambitious: we’re told that “Frontier AI”, centered on Anthropic’s Claude and Mythos models and the Project Glasswing cybersecurity initiative, represents a leap “from horse-and-buggy directly to time travel.”

The sales copy is standard newsletter marketing, but the underlying idea is worth examining: that Chaikin’s proprietary stock-rating system, built around tracking institutional money flows, can separate winners from losers in the AI transition. The presentation presents three “pair trades”, sell one stock, buy another, culminating in a “highest conviction” pick that could supposedly deliver a “100X” return through a “starburst” breakup.

Marc Chaikin spent decades on Wall Street and built the Chaikin Money Flow indicator used by professional traders. His Power Gauge system rates stocks on a scale from Very Bearish to Very Bullish. Let’s walk through each of the three picks and see how they hold up.

The Big Claim

The presentation claims that “Frontier AI”, AI that “can think, plan, reason, and act across any domain, at a level that meets or exceeds human capability”, is about to reshape the economy. The argument is that tracking institutional money flows will let you get ahead of the winners.

The specific “100X” claim is attached to the final pick: a conglomerate supposedly hiding three world-class businesses inside, a streaming service bigger than Netflix (YouTube), an autonomous car company ahead of Tesla (Waymo), and a cloud division growing faster than AWS (Google Cloud). Plus “built-in exposure to two of the biggest anticipated IPOs in recorded history” (SpaceX and Anthropic). If this company “starbursts”, that is, breaks itself up, holders could supposedly “double or even triple the number of shares in your possession, in a single day.”

The Mechanism

Pair Trade 1: Sell Tesla, Buy Magna (MGA)

The pitch: NVIDIA built the brain of the autonomous car and handed deployment of that brain to Magna, the auto parts giant that “gets paid every single time they install NVIDIA’s AI into a car, truck, or commercial vehicle.” Meanwhile Tesla “spent billions of dollars trying to build its own self-driving AI from scratch” and sits at “1/20th of the customers it needs to justify that huge outlay.”

Magna has real revenue, roughly $42 billion per year, real earnings, and trades at a modest valuation. For an auto supplier, it’s priced attractively, largely because investors are cautious on auto companies right now. The NVIDIA partnership is genuine. But the idea that this is an AI play rather than an auto parts company benefiting from a secular trend is a stretch. Magna assembles complete vehicles for multiple brands and faces the same cyclical pressures as any auto supplier.

At the June tease price of $68.15, MGA has held roughly steady, currently around $70. A “Very Bullish” Power Gauge rating hasn’t translated into dramatic outperformance, but at least it hasn’t lost money.

Pair Trade 2: Sell Oracle, Buy Fabrinet (FN)

The pitch: Oracle is “digging itself into debt to build the infrastructure for the world before frontier AI, not the world after it.” Fabrinet, by contrast, builds the “physical core that makes AI run”, fiber optics, transceivers, and photonic systems. The Power Gauge rates FN “Bullish.”

Fabrinet is a real company with a real role in AI infrastructure. Based in Thailand, it manufactures precision optical components for data centers. Revenue is growing 20-25% annually, and earnings are following. At the tease price of $621, it traded at roughly 40X forward earnings. That was too rich — you needed flawless execution quarter after quarter.

The problem is it didn’t hold. FN shares dropped sharply from a May 2026 high of $746, falling below $434 at the worst point. They’ve since partially recovered to around $527, still down about 15% from the tease price. The AI infrastructure rotation of summer 2026 hit optical component names hard, and the Power Gauge’s “Bullish” rating didn’t provide much shelter.

The Highest Conviction Pick: Alphabet (GOOGL)

This is where the math gets interesting. Chaikin’s “dark horse poised to be THE dominant winner in the age of frontier AI” is Alphabet. The third-largest publicly traded company in the world, with a market cap of about $4.3 trillion.

The “starburst” thesis: Alphabet is a conglomerate hiding three world-class companies inside it, YouTube (streaming), Waymo (autonomous driving), and Google Cloud (cloud computing). Plus, Alphabet owns stakes in SpaceX and Anthropic, two hot pre-IPO companies. If Alphabet breaks up, these hidden assets get repriced at standalone valuations.

StockGumshoe’s Travis Johnson, who owns GOOGL as his largest holding, puts this in perspective: Waymo’s last estimated valuation was about $126 billion. That’s roughly 2.9% of Alphabet’s market cap. Even at $200 billion, it’s 4.7%. YouTube and Google Cloud are already meaningful, visible drivers of Alphabet’s revenue and earnings. They’re not “hidden” at all.

The Real Picks

📊 Market Data (Aug 10, 2026 close): GOOGL $354.30 | FN $527.15 | MGA $70.36. All three tickers verified active. FN has partially recovered from lows near $434 but remains well below its $746 May high. The AI infrastructure selloff hit optical names particularly hard. Earnings for FN are expected August 17, which will provide an important update on the trajectory.

Does the Math Check Out?

The “100X” claim for a $4.3 trillion company is mathematically extraordinary. A 100X return from Alphabet’s current price implies a $430 trillion market cap, roughly 15 times total US GDP. Even 10X implies $43 trillion, or 1.5 times US GDP. The advertising copy says “100X” to grab attention, but the actual investment thesis, even if you buy it completely, is likely a double or triple over a multi-year period.

Fabrinet’s valuation math has shifted. At $621, FN needed to deliver exceptional growth every single quarter. At $527, the multiple has compressed to about 28X forward earnings, still not cheap, but less extreme. The question is whether earnings estimates hold up as AI capital spending fears grow.

Magna is actually reasonably priced. At about 10X forward earnings with solid growth, the valuation isn’t demanding. But calling it an “AI play” is mostly creative framing. It’s an auto parts supplier with a NVIDIA partnership, the same way it partners with dozens of other tech and auto companies.

The “conglomerate discount” argument runs backwards for Alphabet. The company commands a premium valuation precisely because investors value the portfolio of businesses. If YouTube, Waymo, and Google Cloud were spun out, the remaining Google Search business, facing antitrust risk, AI disruption to its ad model, and slowing growth, would trade at a much lower multiple. The parts might actually be worth less than the whole.

What They Got Right

1. The Frontier AI thesis is a real technological development. Anthropic’s Claude and Mythos models, cybersecurity AI, and enterprise AI agents are genuine advances that will create winners and losers in the market.

2. Magna’s NVIDIA partnership is a genuine competitive advantage. Being the preferred integration partner for NVIDIA’s autonomous driving platform gives MGA a real revenue stream that Tesla’s do-everything-ourselves approach doesn’t replicate.

3. Fabrinet occupies a critical position in AI infrastructure. Optical components are genuinely essential for data centers, and FN is a leading manufacturer. The thesis was directionally correct even if the timing hurt anyone who bought at the tease price.

4. The “sell overhyped, buy underappreciated” framework has analytical merit. Looking for underappreciated beneficiaries of big trends, rather than chasing the obvious names, is a sound investing approach.

5. Alphabet does have genuinely valuable assets. Waymo’s autonomous driving lead, YouTube’s dominance in streaming, Google Cloud’s AI positioning, these are real advantages the conglomerate structure can obscure.

What They Got Wrong

1. “100X” for a $4.3 trillion company doesn’t withstand scrutiny. This isn’t an exaggeration, it’s a claim that can’t be squared with basic arithmetic at any reasonable timeframe.

2. Fabrinet dropped 15% from the tease price. Recommending a stock at 40X earnings that then declines double digits in two months is painful, regardless of the long-term thesis. The Power Gauge’s “Bullish” rating didn’t anticipate the AI infrastructure rotation.

3. The “starburst” breakup scenario is speculative at best. Alphabet has successfully defended against antitrust breakup attempts and has no incentive to voluntarily split up. The synergies between Search, YouTube, Cloud, and AI research are central to why Alphabet works as a business.

4. The pair trade framing oversimplifies. Presenting complex investment decisions as simple “sell X, buy Y” choices ignores that most investors aren’t choosing between Tesla and Magna, they’re evaluating whether to own either at all.

5. Most of the “hidden assets” aren’t actually hidden. YouTube and Google Cloud are prominently discussed in every Alphabet earnings call. Waymo’s progress is covered extensively in the financial press. The idea that these are undiscovered gems inside the third-largest company in the world doesn’t hold up.

The Verdict

GOOGL: A fine company with an inflated pitch. Alphabet is a perfectly reasonable investment if you believe in AI and are willing to hold for years. But the “100X” framing is marketing, and the “starburst” breakup is speculative. Buy GOOGL because you want to own Google, YouTube, and Waymo, not because a newsletter promises 100X.

FN: Wait and watch. Fabrinet dropped because the AI infrastructure trade soured. If you believe AI capital spending will reaccelerate, there could be an entry point, but you’d be catching a volatile name. Wait for earnings on August 17 to see if the narrative stabilizes.

MGA: The cheapest of the three on a risk/reward basis. A well-run business at a reasonable valuation with genuine exposure to autonomous driving trends. Not an “AI stock,” but solid.

For more on how newsletter promos hype AI-related investments, see our analysis of how AI infrastructure plays get pitched and the difference between real photonics breakthroughs and newsletter marketing.

This is not financial advice. NewsletterVetter has no position in any stock mentioned. All investment decisions involve risk, and past newsletter recommendations should not be relied upon as indicators of future results. The Chaikin Power Gauge Report’s advertising claims about “100X” returns are marketing language and should not be treated as financial projections. Always do your own research before investing.