3 Trade Ideas For The Back Half Of 2026

I grew up in Michigan watching the auto industry get left for dead more than once.

Every time the experts announced the funeral early, they were wrong on timing.

They were also wrong on the size of what came next….

That same setup is sitting in front of me right now, spread across three sectors that almost nobody treats as related.

Software, biotech, and consumer discretionary all run on one shared engine, and that engine is artificial intelligence.

Watching any one of them in isolation hides the actual trade.

By the end of this piece, you’ll know where the flywheel spins fastest.

We’ll also look at how the catalysts line up for the back half of 2026, and where I’m playing defense.

Positions matter more than opinions, so I’ll show you both.

Three Sectors, One Flywheel

Software spent the better part of two years getting punished on one fear.

Investors decided AI agents would gut the seat-based subscription model.

Valuations compressed hard. Some names came all the way down to broad market multiples.

Cheap software with rising revenue per user looks like a mispricing to me, not a broken business model.

The real value in this cycle sits in the workflow layer above the model.

Proprietary data, business logic, permissions, audit trails, and security all live up there.

Any competitor can rent the same model. Nobody can rent fifteen years of a customer’s messy internal data.

I argue with AI more than I probably should. I like poking at the spots where it sounds confident and turns out to be flat wrong.

Even a skeptical read says the platform leaders in cloud, cybersecurity, and vertical software are set up to monetize this cycle.

Enterprise budgets are rotating from raw infrastructure toward applications that produce something measurable.

Biotech already ran with a version of this story. The sector delivered one of the stronger recoveries of 2026 on gene editing progress, heavy M&A, and a visible pickup in financings and IPOs.

Pharma is buying because a patent cliff is a math problem with a due date. Revenue rolls off on a known day, and buying a pipeline beats building one.

AI is the accelerant underneath all of it. It has moved out of pilot programs and into real R&D work like target identification, molecular design, and trial optimization.

Compressed timelines change the economics of every program. Each month shaved off development is a month added to the commercial life of the drug.

Consumer discretionary is the laggard here, and the caution makes sense to me. The K-shaped economy is real, essentials inflation hasn’t fully cooperated, and near-term fundamentals are soft.

Here’s the bridge that keeps getting missed. Longer healthspans free up lifetime earning power, and that money eventually finds its way into discretionary categories.

Software productivity gains support wages and profits, which feed straight back into consumer demand. Software is also rebuilding retail itself through personalization and e-commerce, which lets scaled operators take share even in a weak tape.

Where The Setups Are

Narratives are easy to write. Positioning costs real money, so here’s what I’m tracking into the back half of 2026:

• Software earnings that show actual AI monetization through consumption pricing, higher net revenue retention, or expanding revenue per customer, instead of AI mentions on the call

• Biotech trial readouts paired with continued M&A as pharma keeps refilling pipelines

• Genuine stabilization in consumer confidence and real incomes, not one hot month of data

• Margin evidence that AI productivity is reaching income statements across industries

• Rotation out of pure AI infrastructure and into AI adopters, which broadens the trade well past the usual mega-cap names

That last bullet matters most to me. Infrastructure gets paid first in every technology cycle, and adopters get paid second.

The second wave tends to last longer than the first one.

Trinity Terminal has been flagging setups building across all three sectors while valuations reset and sentiment lags the fundamentals. Divergence between price and data is usually where the money lives.

Playing Defense Without Losing The Thesis

I’m not blind to the risks, and you shouldn’t be either. Software could see slower AI returns in certain sub-segments, and biotech carries clinical and regulatory risk that never fully goes away.

Consumer discretionary gets worse in a hurry if inflation re-accelerates or the labor market cracks.

None of that changes my intermediate or long-term stance. Drawdowns are normal, and every cycle I’ve traded through served up at least one that felt terminal while it was happening.

Perma-bears have been calling the top for months. I’d rather be short with the intention of getting long than sit on the sidelines waiting for permission that never shows up.

Short-term hedges are how I fund patience. They let me hold the core thesis without white-knuckling every red candle.

My highest-conviction setups favor software platforms with real data moats, biotech names with near-term catalysts, and consumer names levered to affluent or digitally native demand.

Selectivity inside each sector still decides the outcome. Size the position so one bad headline can’t force you out of a good idea.

Infrastructure got paid first in this cycle. The adopters get paid second, and that second wave is the one I am positioned for.

Picking the sector is the straightforward part. Finding the name inside it where institutional money is already accumulating takes a score.

AMD read 4.3 before the trade war narrative reversed. It ran 389%. HUT 8 (HUT) read 5.2 before the AI infrastructure story reached the crowd. It ran 132%.