Can CeylanVienna-based, globally curious.

New this weekRead when you can. Listen when you move.

Articles/Finance & Investing

The Stock Market Feels Like Crypto Now. Here Is What I Am Actually Doing About It.

AI is rotating money through the market the way crypto did in 2017, and if you are still holding the same allocation you built three years ago, it might be time to think again.

22.5.2026·6 min read

Listen to this article

Opens the site-wide player at the bottom, so you can keep browsing.

When the market starts behaving like crypto, the question is not whether to panic. It is whether your allocation was built for a world that no longer exists.

I have been avoiding this article for a few weeks.

Not because I do not have thoughts on it. Because after my burnout, I made a deliberate choice to stop obsessing over my portfolio. Long-term, slow, boring investing had always worked better for me than the anxious tab-switching of short-term trading. Letting it go a little felt healthy. It probably was.

But Q1 numbers are in. And something is shifting in a way that is hard to ignore.

Continue reading?

The next essay lands in your inbox first. The thinking behind the decisions.

Why this matters right now

Here is the thing about Q1 2026 that does not quite add up on paper. GDP growth came in soft. Inflation stayed sticky in both the US and Europe. Consumer confidence was weak. Tariff noise from renewed escalations under the Trump administration rattled markets through January and February.

And yet. Major indices hit all-time highs.

That gap between the economic mood music and where markets actually landed is what I keep coming back to. Money is not sitting in broad indices waiting for macro conditions to improve. It is moving. Decisively. Into very specific corners of the market.

The AI investment wave is doing something structurally similar to what crypto did to speculative appetite in 2017 and again in 2021. Except this time it is not retail traders buying tokens at 2am. It is sovereign wealth funds, pension allocators, and the largest infrastructure investors in the world redirecting capital at a speed I genuinely did not expect to see this decade.

So I stopped asking "what is the market doing right now" and started asking "what are the structural shifts already in motion that will not reverse regardless of a short-term recession or rate cycle." That reframe helped me think more clearly.

Here is where I landed. None of this is investment advice. These are my personal observations and the lens through which I am thinking about my own allocation. Do your own research, and talk to someone qualified before making any decisions.

Where the money is actually going: three macro themes I am tracking

1. Energy infrastructure

This one feels almost unsexy compared to the AI hype. But the numbers are staggering. AI data centres and automation systems require orders of magnitude more energy than the digital infrastructure we built in the 2010s. We are not ready for that demand curve.

Hydrogen, solar, wind, and grid infrastructure are seeing renewed interest from a completely different demand source than before. It is not just climate policy. It is raw computational appetite. And energy infrastructure does not go to zero when an AI model gets outcompeted. The electrons still need to flow.

2. The AI infrastructure stack

Most of the market conversation is here, but there are really three distinct sub-themes worth separating.

Rare earths and critical materials. The hardware required to build and run AI, chips, cooling systems, sensors, storage, depends on materials with genuinely constrained supply chains. Ongoing US-China trade tensions are creating strategic urgency around sourcing. Slow burn, but the structural case is real.

Hardware and data centre infrastructure. Chip production, rack density, power distribution, cooling. This is where the arms race is most visible. The infrastructure layer tends to be less volatile than the application layer because it services everyone, regardless of which AI model wins.

A shifting top of the market. OpenAI is reportedly exploring a public market path. Anthropic is a legitimate enterprise alternative with differentiated positioning. The composition of the most valuable AI companies in five years will probably look different from today, and that has real implications for where capital flows.

3. Robotics and autonomous systems

The longest time horizon of the three, but the investment infrastructure is being built right now.

The hardware stack is already in a full capital cycle: sensors and LiDAR, radar systems, precision localisation, connectivity infrastructure (V2X, 5G private networks), and edge compute. Tesla remains interesting because it is operating across the full stack. But the more durable plays might be component suppliers who benefit regardless of which platform wins the consumer-facing race.

One thing worth watching: SpaceX IPO speculation has been circulating seriously. If and when that happens, it would be one of the largest public listings in years. Worth having a plan for in advance rather than reacting in real time.

What most people get wrong

They treat it as a short-term trade.

The AI infrastructure buildout is a multi-decade capital cycle. The companies that win in 2026 are not necessarily the ones that win in 2032. If you are buying into these themes expecting a clean 18-month return, you will probably sell at the wrong moment.

A genuine risk worth naming: a short-term recession or credit event is still entirely possible. The macro picture in Q1 2026 is genuinely uncertain, tariff volatility is not resolved, and consumer balance sheets in the US are more stretched than the headline numbers suggest. A long-term structural thesis does not immunise you from a painful drawdown along the way.

When the market starts behaving like crypto, the question is not whether to panic. It is whether your allocation was built for a world that no longer exists.

What to actually do

  • Audit your current holdings for real exposure. Pull up the top 10 underlying positions in any fund you hold. You may be less diversified than the fund name implies. Adjust from the actual holdings, not the label.
  • Think in layers, not bets. Energy infrastructure, AI hardware, and robotics components carry three different risk and time profiles. Size them accordingly. Infrastructure tends to be more stable. Growth stocks and application-layer plays carry more variance.
  • Keep a portion boring on purpose. The reallocation I am describing is not a wholesale pivot. It is a considered addition to a base that stays slow and steady. Do not let macro excitement push you into over-rotation.
  • Watch for liquidity events. A potential OpenAI IPO, a potential SpaceX listing, and continued rare earth ETF development in European markets are moments worth planning for in advance.
  • Do not mistake volatility for opportunity without a thesis. Some growth stock charts right now look extraordinary. That is not a reason to buy. The question is whether you understand why they are moving and whether that reason holds over the period you intend to hold.
  • Talk to someone qualified before acting on any of this. Genuinely. I am thinking out loud about my own allocation, not giving you a roadmap for yours.

Are there themes I missed? Probably. Biotech, longevity tech, and defence and aerospace are all conversations worth having in a future piece.

For now: long on structural disruption, honest about short-term risk, and no longer pretending that the allocation I built in a different macro environment is still the right one.

Share

If you read this far, the next one is for you.

The thinking behind the decisions, on Finance & Investing and everything adjacent to it. No schedule, no filler. You get it before anyone else does.

More on Finance & Investing

What Financial Metrics Actually Matter Right Now (And Why I Keep Second-Guessing Myself)

Revenue growth was the gospel of 2018. Then survival was the gospel of 2022. Now everyone's betting on AI productivity that hasn't fully shown up in the numbers yet, and the S&P PE ratio suggests the market believes the story completely.

Why I Can't Buy a Piece of OpenAI, And Why That's a Problem

VCX went from $40 to $550 in days on NYSE. I watched it from Vienna. I couldn't buy a single share, not because I don't understand the risks, but because MiFID II says no.

VCX Fell 80%. Now I Can Actually Buy It, Here's What I'm Thinking

The fund that went from $40 to $550 and made European investors feel left out is now sitting at $113, and I found a way in through a Swiss bank.

Learn the underlying concept

Learn

Two AI agents need one live memory file

If two AI coding agents share a repo but not a single mutable memory layer, the user becomes the message bus. Here is the failure mode, why it happens, and the operating model that fixes it.

Learn

The distribution map pattern: one config that drives all publishing outputs

Multi-platform publishing workflows accumulate per-platform hacks over time. The distribution map pattern replaces them with a single declarative config: topic → platform list, which becomes the source of truth for what gets generated and where it goes.

If this resonated, I'd be happy to talk about it.

Find me →
← Back to articles