Trade Alert: FUD (Fear, Uncertainty, And Doubt)
Why we're not buying the FUD, and why we're placing bullish options bets on four names that passed our screens instead.
FUD, Again
Markets have been hit with another wave of Fear, Uncertainty, and Doubt since Friday.
Some of it is the usual “AI bubble” talk. The comparison being pushed again is that this is 1999, and the recent weakness is the start of a Dot-Com-style collapse.
We addressed that argument directly a couple of weeks ago in “Still Comparing This To The Dot-Com Bubble?”.
The short version is that the earnings, cash flows, and valuations of the leading AI companies today don’t look much like the profitless dot-com names that defined the late 1990s.
The AI buildout is already showing up in real revenue across semiconductors, memory, servers, optics, networking, power, cooling, and data-center infrastructure.
Still Not 1999
The stronger analogy is not “the bubble is popping.”
It is that markets are starting to sort the winners from the pretenders.
Another version of the FUD says the coming IPOs of SpaceX, Anthropic, and later OpenAI will suck the air out of the rest of the market. That sounds plausible at first, but the Dot-Com analogy breaks down there too.
Some of the most iconic Internet IPOs of that era happened years before the market top: Netscape in 1995, Amazon in 1997, and eBay in 1998. As Eliant Capital noted on X, those listings didn’t mark the end of the bull market.
They were part of the process by which the market started separating durable companies from promotional stories.
IPOs Aren’t The Top
The IPO wave will matter.
It will test demand. It will reprice some private-market assumptions. It will give public investors a chance to compare AI leaders directly instead of valuing them through second- and third-derivative plays.
But treating those IPOs as an automatic liquidity vacuum assumes the conclusion.
If anything, the better historical lesson is that important IPOs can broaden a bull market before they eventually help discipline it.
The AI Roadmap Fight
Then there is the AI-infrastructure-specific FUD.
A summary of a SemiAnalysis report this week suggested that NVIDIA’s 800V DC architecture and co-packaged optics timelines may be pushed out, with 800V DC large-scale shipments slipping toward 2028 and broader CPO mass production delayed into 2028–2029. That hit a number of AI-infrastructure names.
It’s a serious claim, but it isn’t the only data point.
The Companies Say Otherwise
NVIDIA has publicly positioned co-packaged optics as central to its AI networking roadmap. Its Spectrum-X Photonics release said Quantum-X Photonics InfiniBand switches were expected later that year, with Spectrum-X Photonics Ethernet switches coming in 2026. NVIDIA’s own silicon photonics page says Spectrum-X Ethernet Photonics is available in the second half of 2026.
Lumentum has also pointed to optical timelines that don’t match the most bearish interpretation. A conference transcript summary described optical scale-up products shipping in the second half of 2027, with a larger ramp after that.
Coherent has been pointing in the same direction. Its OFC 2026 investor presentation highlights VCSEL arrays as ideal for NPO/CPO transceivers for scale-up, with ongoing hyperscaler engagements.
That doesn’t settle the debate. Roadmaps can shift. Yields can disappoint. Bridge technologies can last longer than expected.
But when one report causes a forced selloff in names where the companies themselves are still pointing to earlier AI-infrastructure and optical-networking timelines, it’s not a reason to panic. It’s a reason to compare the report against the primary-source evidence.
Incentives Matter
Markets aren’t populated by disinterested monks.
Plenty of institutions missed parts of the AI-infrastructure rally, and some of them would welcome better entry points. That doesn’t prove any specific bearish report is wrong, and it doesn’t require assuming bad faith. It just means we should separate the claim from the market reaction.
Temporary bridge technologies can matter for a while. Roadmaps can shift. But when the market treats any uncertainty as proof that the whole AI-infrastructure thesis is breaking, it creates exactly the kind of volatility we can sometimes use.
Hormuz: The Question Now
Finally, there is the Strait of Hormuz.
That risk is real. Iran has already disrupted the Strait. The question now is whether that disruption is starting to be lifted, and whether the worst-case version of the story is losing force.
Iran’s limited responses so far to the latest limited strikes by the U.S. and Israel suggest that, despite its bluster, it may be unable or unwilling to resume full hostilities.
There is also a theory that the U.S. may be quietly reopening the Strait of Hormuz, as laid out in this thread by Guy Laron. Recent Kuwaiti oil-export data suggest there may be some truth to that.
This morning’s CPI report fits the same pattern. The headline number was hot because energy was hot, and energy was hot in part because the Strait of Hormuz disruption helped push oil and gasoline prices higher. If that disruption is now being gradually lifted, it should put downward pressure on the energy component of future CPI reports. Markets don’t need perfect news to stabilize after a FUD-driven selloff. They just need the worst-case version to stop getting worse.
Back To Process
That’s the backdrop for today’s trades.
We’re not trying to call a bottom in every stock that sold off. We’re looking for names that pass our technical screens and have option chains where volatility gives us attractive defined-risk structures.
Today’s alert has exposure to four different themes:
AI hardware / memory and storage infrastructure,
semiconductor-capex / advanced-packaging equipment,
a special-situation biotech trading below its cash and financial investments,
resilient-positioning / GPS-alternative infrastructure.
All four names pass the relevant screens we use to avoid falling knives. For newer readers, RSI stands for Relative Strength Index, and our preferred range for fresh entries is generally 40–60. We also look at Chartmill Setup scores, technical scores, and directional-indicator behavior.
Those screens matter because they help us avoid simply buying weakness because a story sounds interesting.
The structures below use the same basic approach we used in yesterday’s alert: harvest elevated volatility where it’s available, define the downside, and keep meaningful upside exposure.
Today’s First Market Watchers Trade
AI hardware / memory & storage infrastructure theme
