Two earnings reports now carry a question the market has avoided for a year: can ASML and TSMC restore confidence in the AI trade?

The framing itself is the news. FXStreet poses the question in its preview of the two companies' results, and the phrasing implies the premise — that confidence in AI-linked equities has weakened and needs restoring. No verdict appears in the material. What follows is an assessment of what the question reveals, not what the numbers will show.

Why does this earnings season matter?

ASML (the Dutch maker of lithography machines, the equipment required to print advanced chips) and TSMC (the Taiwanese foundry that manufactures most of the world's leading-edge processors) sit at the top of the semiconductor supply chain. Their results function as upstream evidence. Demand visible in their order books either validates or contradicts the spending forecasts issued by AI chip designers further down the chain.

That proxy role cuts both ways. Strong equipment bookings would support the AI investment thesis. Weak guidance would hand skeptics their clearest data point yet, because neither company depends on narrative — both sell physical tools and wafers against firm orders.

What does the source actually claim?

Very little, and that restraint deserves note. The FXStreet item contains a headline question and no published figures, no forecasts and no quoted executives. It asserts two things only:

• ASML and TSMC are reporting earnings.
• Confidence in the AI trade — a term covering the basket of equities bought on AI demand expectations — is in doubt.

The second assertion is the load-bearing one. A year ago, previews of these same reports asked whether results could extend the rally. The vocabulary has shifted from extension to repair. Language moves before fundamentals in trade press, and this shift is measurable: the question presumes damage.

What is not said?

No definition of "confidence" appears. The word could refer to analyst price targets, options positioning, or institutional allocation. Each measure would fail differently under the same earnings print.

No date for either report appears in the material. No revenue, margin or booking estimates are cited. No comparison to prior quarters is offered.

Readers should also note the framing risk. Framing earnings as a test of an entire investment theme inflates two quarterly reports into a referendum, and quarterly reports measure one quarter — chip orders respond to capacity plans set years ahead, so a single print can mislead in either direction.

What is the honest benchmark?

Restoration of confidence requires more than a beat on consensus. It requires forward indicators — ASML's bookings figure and TSMC's capital-expenditure commentary — that confirm the AI demand curve operators have promised, alongside margins proving that demand is profitable rather than subsidized.

A comparison fixes the stakes: equipment and foundry results are to the AI trade what drilling data is to an oil field — the narrative prices the reserve, but the drill bit reports the yield. Both companies have reported through AI boom cycles before, and each time analysts debated whether peak orders had passed.

Until the numbers arrive, the only defensible statement is the source's own: the question of restored confidence is open. Anyone claiming a confident answer today is selling one.

Die Signal will assess the reported figures against this framework when the results are published.