For three years, “AI” has been a magic spell. Say it on an earnings call, and the stock levitates. But markets always sober up the same way: cash flow, not vibes.

That’s why the bear case showing up in 2026 looks different from the usual “AI is fake” rant. The smarter bears aren’t betting that AI won’t work. They’re betting something more brutal:

And that second one matters because hyperscaler capex is now so large that even a “slower growth” trajectory can re-rate the whole complex. Goldman has flagged that hyperscaler AI investment is enormous in 2026 (hundreds of billions) and that “what if the spending slows?” is becoming a real valuation risk.

At the same time, investors have been gravitating toward “HALO” names—heavy assets, low obsolescence—as a hiding place from AI disruption fears.

So if you’re building a short basket, you don’t want a random dartboard of “tech.” You want to short where AI eats economics—and (separately) where AI spend creates a balance-sheet hangover.

Goal: short companies whose economics rely on human time, fees, or information tolls—in categories AI agents can compress.

How to construct it (rules of the road):

Equal-weight (10% each) keeps it simple:

1) Accenture (ACN) – consulting is a labor model; AI threatens billable-hour leverage unless pricing power holds.
2) Cognizant (CTSH) – IT services + outsourcing are prime targets for automation pressure.
3) EPAM Systems (EPAM) – higher-beta IT services exposure; sensitive to “fewer humans needed.”
4) Robert Half (RHI) – staffing/recruiting is a match-making + labor arbitrage business.
5) ManpowerGroup (MAN) – same logic: AI compresses “people as product.”
6) Concentrix (CNXC) – contact center/BPO: voice agents aim straight at the core offering.
7) TTEC (TTEC) – another call-center-heavy model exposed to automated agents.
8) C.H. Robinson (CHRW) – freight brokerage is matchmaking; AI can compress spreads by reducing search costs.
9) LPL Financial (LPLA) – wealth distribution + advisor economics can get fee-compressed as AI/automation improves.
10) Thomson Reuters (TRI) – legal/research workflows are exactly where AI tools are attacking information “tolls.”

Why these work as a basket (instead of a single-name bet):
You’re not claiming “this company dies.” You’re expressing a view that AI shifts bargaining power away from labor/fees and toward the customer. Some names will adapt. Some won’t. The basket captures the theme with less single-name blow-up risk.

These are the “pure play” pain trades—great when the narrative turns, dangerous when it doesn’t:

Goal: identify AI-linked companies where the funding and ROI math is the weak link.

This is where the bear thesis gets spicy: the fear isn’t “AI doesn’t work.” The fear is “AI works, but the economics accrue unevenly—and the spenders eat the bill.”

Oracle has publicly described an enormous funding need to expand its cloud infrastructure: it expects to raise on the order of $45–$50B in calendar 2026 for OCI expansion to meet contracted demand (with a mix of equity and debt).
And the market is aware that Oracle’s AI exposure is tied to major counterparties, including OpenAI—Reuters has reported OpenAI signed a contract to purchase $300B of computing power from Oracle.

When you see “huge capex + huge contracts + big funding plan,” you also see why bears migrate from equity shorts to balance-sheet and credit risk. (Oracle has even faced bondholder lawsuits alleging disclosure issues around debt plans.)

This is not “short these.” This is “these are where risk concentrates if the AI spend path disappoints.”

The spenders (capex + ROI scrutiny):

The “second derivative” suppliers (hit if spending slows):

What makes this group risky:
It’s not that they’re “bad companies.” It’s that the entire bull case assumes a certain slope of spending and monetization. If that slope flattens, the re-pricing can be violent.

If the market keeps rewarding HALO-style safety—heavy assets, low obsolescence—the “winners” list often looks boring on purpose:

HALO-style examples (illustrative):

The core idea: you can’t “prompt” a power grid, a refinery, a mine, or a bulldozer into existence. That scarcity shows up in pricing power.

Shorting is where smart people go broke. If you run this theme, the survivable way to do it is usually:

Last night we had NVDA earnings, pretty much what I told Bloomberg….

Blowout earnings, but not enough to meaningfully move the stock. The print should provide some stability for the overall AI infrastructure, as it’s business as usual. One thing to keep in the back of your head though is OpenAI slashing long term capex ex from $1.4 trillion to $600 billion by 2030. This reminds us that massive capex can’t go on forever, and at some point the payoff matters.

Strange that Bitcoin has rallied so much since the news of the Jane Street lawsuit. Probably a coincidence, but I also don’t believe in coincidences. Bitcoin is in the red today, bulls want to see it reclaim $70K.

The AMD/META warrant structure is a signal that mega-customers are using more risk-sharing / incentive-aligned contracting. That can be bullish (locking demand) and dangerous (dilution, concentration, headline risk).
Implication: This is a new regime: big AI deals may increasingly come with “strings”. The market will reward the revenue, but punish unclear economics and dilution.

HPQ commentary (memory doubled; % of BOM up) reinforces what we’ve been hearing elsewhere: memory is a real constraint and cost headwind.
Implication: This creates a barbell: (a) suppliers/beneficiaries of scarcity do well, (b) OEMs and “assembly” businesses get squeezed unless they can pass through.

The conversation is shifting to optical interconnect, optical switching, silicon photonics / CPO, and the validation stack that has to scale with it.
That’s not hype — it’s the natural next bottleneck once clusters get big enough that copper becomes a power/thermal/bandwidth tax.

Portfolio implication: if you already own “AI compute,” the next incremental asymmetry is in optics + interconnect + test/validation.

Across mentions of the Anthropic/agents narrative, the message is consistent:

Portfolio implication: don’t add “AI software” broadly. Add only where there is (a) distribution + data gravity, and (b) a credible packaging/monetization path.

This week we made two changes to MEMY: 

These moves are about asymmetry: concentrating capital in setups where (a) the market is still discounting the upside, (b) the path to “surprise” is plausible in the next 1–3 years, and (c) the bull case is driven by durable structural forces rather than a single narrow outcome. 

1) VKTX → SNOW: rotating from “single-asset clinical convexity” to “platform re-rating” 

VKTX has been a classic “call option” stock: a single program can dramatically reprice the equity. The issue is that the obesity landscape is no longer a scarcity trade. The market is increasingly demanding clear differentiation (not just “good results”), and it’s starting to price obesity upside with more skepticism because the competitive set is deep and well-capitalized. VKTX can still work, but the distribution has become more “binary,” and the upside is less uniquely mispriced than it was. 

SNOW gives us a different kind of convexity: enterprise AI moving from experiments to production, where the true bottleneck is not models—it’s governed data, security, and integration into real workflows. The market has tended to value Snowflake through the lens of a “data warehouse cycle,” but the right-tail scenario is bigger: Snowflake becomes a control plane for AI agents and AI-native applications operating directly on trusted enterprise data. 

The catalyst path is also improving: Snowflake and OpenAI have announced a partnership to bring OpenAI capabilities directly into Snowflake’s platform for enterprise customers. That strengthens SNOW’s positioning at exactly the moment “agentic AI” shifts from demos to deployment. 

In short: we’re swapping a crowded single-asset convexity for a platform whose upside can compound across many customers, many workloads, and many use cases. 

2) LUNR → BABA: keep space conviction, add a new driver set 

We’re not stepping away from space. We’re consolidating it. 

Space remains a powerful multi-year theme, but in a focused “top 20” portfolio we don’t want multiple positions whose returns are dominated by the same kind of event risk (mission cadence, contract headlines, schedule narratives). We’re keeping our highest-conviction space asymmetry elsewhere and freeing up a slot for a different right-tail regime. 

BABA is that regime shift. 

Alibaba is still priced with a heavy China risk premium, which is exactly why the setup can be asymmetric. You don’t need perfection—you need less bad outcomes: stabilization in sentiment, normalization of the discount rate, and even modest evidence that core businesses are resilient. At the same time, Alibaba has multiple “shots on goal” for upside: platform cash flows, capital returns, and a credible commitment to investing through the cycle—especially in cloud/AI infrastructure. 

Alibaba has publicly outlined a large multi-year investment plan into cloud computing and AI infrastructure and has also disclosed meaningful remaining authorization under its share repurchase program. That combination matters for asymmetry: it creates several routes to upside while the downside is already widely debated and heavily priced. 

In short: we’re rotating from a second space exposure into a large, under-owned platform where a change in risk perception alone can drive meaningful re-rating. 

What these swaps do for MEMY 

As always, MEMY is designed to own the highest-quality right tails—not the loudest stories. 

Not investment advice; this is portfolio commentary. 

Important Risk Information 

Investing involves risk, including the possible loss of principal. The Fund’s investments in equity securities are subject to the risk that the value of the underlying stocks may decline. 

Because MEMY focuses on "Meme Stocks," it is subject to unique risks. Meme Stocks often have trading volume that increases not necessarily because of a company's performance, but because of social media attention. As a result, they are prone to extreme volatility, experiencing spikes of rapid growth followed by dramatic drops. The Fund’s strategy relies on social media analytics, which are relatively new and untested, and sentiment analysis may prove inaccurate in predicting stock performance. 

The Fund’s use of options, including put spreads and FLEX Options, involves additional risks, such as the possibility that options may expire worthless, the impact of changes in implied volatility, and the potential for increased losses due to leverage embedded in options positions. The Fund is non-diversified and may be more volatile than a diversified fund. 

There is no guarantee that the Fund will achieve its investment objective or that its options strategy will be successful. Shares of the Fund are bought and sold on an exchange at market price (not NAV) and are not individually redeemed from the Fund. Brokerage commissions will reduce returns. 

Investors should carefully consider the investment objectives, risks, charges, and expenses of the Tuttle Capital Meme Stock Income Blast ETF before investing. For a prospectus with this and other information about the fund, please call (833) 759-6110 or visit the Fund’s website. Please read the prospectus carefully before investing. 

Distributor: Foreside Fund Services 

Lumentum (LITE) — (We own LITE in UFOD) Photonics is quickly becoming the “next bottleneck trade” inside AI: as GPU clusters scale, performance (and power efficiency) increasingly comes down to how fast and reliably you can move data between racks, pods, and campuses—and that means more optical content. Lumentum is positioned right at that intersection (optical components + systems into cloud/AI networks), and its latest quarter read like a real inflection, not just hype: revenue was $665.5M (+65% YoY) with non-GAAP EPS of $1.67, and management guided the next quarter to $780–$830M with 30–31% non-GAAP operating margin and $2.15–$2.35 non-GAAP EPS. Even more important than the beat was what they flagged as the next legs: optical circuit switches (OCS) (with backlog already > $400M) and co-packaged optics (CPO) (including a new multi-hundred-million-dollar order deliverable in 1H 2027). That’s real demand visibility into where AI infrastructure is headed. The main risk is classic photonics whiplash—customer concentration and “build/digest” cycles—but if interconnect becomes the constraint, LITE is one of the cleaner ways to express that shift.

The H.E.A.T. (Hedge, Edge, Asymmetry and Theme) Formula is designed to empower investors to spot opportunities, think independently, make smarter (often contrarian) moves, and build real wealth.

The views and opinions expressed herein are those of the Chief Executive Officer and Portfolio Manager for Tuttle Capital Management (TCM) and are subject to change without notice. The data and information provided is derived from sources deemed to be reliable but we cannot guarantee its accuracy. Investing in securities is subject to risk including the possible loss of principal. Trade notifications are for informational purposes only. TCM offers fully transparent ETFs and provides trade information for all actively managed ETFs. TCM's statements are not an endorsement of any company or a recommendation to buy, sell or hold any security. Trade notification files are not provided until full trade execution at the end of a trading day. The time stamp of the email is the time of file upload and not necessarily the exact time of the trades. TCM is not a commodity trading advisor and content provided regarding commodity interests is for informational purposes only and should not be construed as a recommendation. Investment recommendations for any securities or product may be made only after a comprehensive suitability review of the investor’s financial situation.© 2026 Tuttle Capital Management, LLC (TCM). TCM is a SEC-Registered Investment Adviser. All rights reserved.