Fundamentals set the direction. Technicals shape the timing.

Start by asking whether the thesis has changed

SemiTracker does not begin by asking whether a rally still has room to run. It asks: did this development change the industry's fundamentals, or only market sentiment? It places each news item, price move, or research update in the supply chain, then checks whether the impact reaches earnings expectations or valuation.

The method separates direction from timing. Supply-chain position, adoption, and supply-demand shifts identify where value may accrue. Expectations, positioning, and price structure help determine timing. The goal is to distinguish broad sector exposure from company-specific advantage at each stage of the cycle.

What it checks first

This agent uses a clear source hierarchy. The most recent Semiconductor Tracking material comes first. When newer evidence conflicts with an older view, the newer evidence takes precedence and is dated accordingly.

• Use the latest evidence on fast-moving topics For issues such as H20 policy changes, DeepSeek, the memory cycle, CPO adoption, and geopolitics, the agent follows the full timeline and gives priority to the most recent well-sourced evidence.
• Long-form beats short notes; research must be source-tiered Relevant source text is weighed in order: long-form pieces, short notes, then field research notes. Research content must be labeled as first-hand source, primary research, institutionally cross-checked, or media secondhand — rumor never gets passed off as fact.
• Web search is a data layer, not an opinion layer News, filings, market data, and sell-side research are inputs. The agent extracts the relevant facts, then evaluates them through its core frameworks instead of treating a recent report as a conclusion.
• Gaps in the source material are stated clearly If the available material does not cover a ticker or topic, the agent labels the result as framework-based inference rather than implying that a published view exists.

Five core frameworks

It does not analyze a stock in isolation. Every question is placed across three layers: macro, industry, and company. Macro conditions shape liquidity and risk appetite; industry dynamics define the cycle and adoption curve; company execution determines who converts demand into earnings.

1. 01

   Macro, industry, and company

   First determine whether an external variable has genuinely moved through the supply chain, then identify whether it affects valuation, order momentum, or a single company's earnings.

2. 02

   The AI capability formula

   Breaks AI progress into compute, networking, memory, data, algorithms, and application demand. A headline matters only if it changes a constraint, shifts economics along the chain, or creates a measurable benefit for a company.

3. 03

   Three-stage penetration

   The same technology roadmap deserves different PE bands in its introduction, acceleration, and maturity stages. Early on you price the potential, mid-stage the earnings leverage, late-stage the delivery and the competitive landscape.

4. 04

   EPS × PE split

   First separate earnings impact from valuation impact. If EPS expectations are changing, focus on delivery and operating leverage; if the multiple is changing, focus on expectations, positioning, and crowding.

5. 05

   Fundamentals select the exposure; market structure informs timing

   Fundamentals identify the relevant companies and the direction of the thesis. Positioning and price structure help frame timing. The output uses ranges, conditions, and position-sizing considerations rather than precise entry or exit points.

Mapping the transmission chain: did the news reach your positions?

Semiconductor rallies often begin with an event, but portfolio impact depends on transmission. A ban, a round of price increases, a technology roadmap, or an order book must pass through policy or demand, a supply-chain link, company revenue, margins, and valuation before it changes the investment case.

So it asks: does this hit upstream equipment, materials, chips, packaging, optical modules, servers, power, memory — or the application side? If it only stirs distant sentiment without reaching orders, prices, or profits, it's most likely noise. If the sourcing is strong, the chain is clear, and either EPS or PE has started to move, it's worth digging further.

Locating the stage: beta or alpha

CPO, advanced packaging, HBM, NAND, DRAM, switch silicon, and the server chain cannot all be assessed with the same yardstick. Early in adoption, valuations reflect potential; during acceleration, earnings leverage matters more; as adoption matures, competition, cost curves, and cash flow become decisive.

The framework identifies the stage first, then separates broad sector exposure from company-specific advantage. A price increase is not automatically a buy signal; its significance depends on whether it reflects tighter supply, recovering demand, or a short-lived inventory adjustment.

What the output includes

A complete answer avoids an industry-data dump. It asks whether the thesis has changed, where the change has reached the supply chain, how price action affects timing, and what evidence would invalidate the view.

1. 01

   Thesis assessment

   Starts by stating whether the core thesis has changed. If the development affects sentiment but not fundamentals, it says so; if the impact is material, it traces the next link in the chain.

2. 02

   Transmission chain positioning

   Maps the event across macro conditions, industry dynamics, and company execution to identify its effect on valuation, cycle momentum, orders, prices, margins, or inventory.

3. 03

   Technical guidance

   Written in three parts: expectations, response, sizing. Ranges instead of price points, conditions instead of gut calls.

4. 04

   Failure conditions

   States which changes would weaken or invalidate the thesis, along with the evidence required to revisit the conclusion.

Try asking

Start with an event, a supply-chain link, or a technology roadmap. The agent will assess the thesis, trace the impact, and identify the variables that matter next.

Who it's for

It suits people who already have a supply-chain question, position exposure, or a technology-roadmap hypothesis. You don't need an answer first — just spell out the event and the link you care about.

• Investors following AI hardware, semiconductors, servers, optical modules, memory, and advanced packaging
• Anyone assessing whether a headline has changed the outlook for their holdings
• Researchers who want to tell sector beta, stock alpha, and sentiment noise apart
• Anyone re-rating supply chain opportunities with penetration stages and the EPS × PE framework

Boundaries: it won't force a call

It won't quote current live prices, capacity, shipments, or the size of a price hike from memory. On valuation or capacity questions, it needs you to supply current data, or it reasons through the framework from currently retrievable data; numbers from historical articles always carry their dates, with a reminder that they belong to the moment of publication.

It won't guarantee returns, give precise targets or exact entry and exit timing, fabricate survey figures, contract details, or prices, or treat a lone rumor as fact. When the available evidence is insufficient, it leaves the question open instead of guessing.

This material is compiled from company filings, news, published institutional research, and social media. Nothing here about any sector or stock is investment advice. Markets carry risk; invest with care.

SemiTracker

Bring an AI hardware, semiconductor, or supply-chain question. The agent will test the thesis, map the transmission path, locate the cycle, and separate signal from noise.

Built from public research and post-mortems to demonstrate a semiconductor analysis framework. It does not represent any source author or provide investment advice.

After CPO's rally, what would justify further upside? What would extend or end the memory price cycle? Which supply-chain link could benefit next?

Ask SemiTracker

Send