After the news, check the supply chain first

First question: which node is this signal at?

The Ming-Chi Kuo agent doesn't start by rehashing headlines, and it doesn't trade on market mood. It puts every event back on its supply-chain node: is the spec finalized, are orders confirmed, has component stocking ramped, is ODM or foundry capacity being added, are yields on critical components keeping up.

If a number only exists at the rumor level, it says visibility is low. Only when stocking, capacity, and end demand all point the same way does it raise the weight of its judgment call.

1. 01

   Locate the node

   Place the news, rumor, or earnings report at its spot on the supply chain: design, components, foundry, assembly, channel, end demand.

2. 02

   Trace the transmission

   An order change is never an isolated number — look at who benefits first, who gets displaced, and who absorbs the cost or yield pressure.

3. 03

   Separate findings from inference

   Verified facts and personal inference stay apart: what's confirmed is called confirmed, what's undecided is called undecided.

4. 04

   Name the variables to watch

   Every conclusion must be verifiable — or falsifiable: orders, shipments, capacity, yields, earnings-call guidance, and the next round of component stocking.

What it can unpack

It's best suited to Apple and consumer-electronics product cycles, the Nvidia/AI server supply chain, and questions around semiconductor foundries and geopolitics. Especially: is this report real, who ultimately lands the orders, and which link has the most margin flexibility.

What an answer looks like

Its output always delivers four things: a one-line conclusion, supply-chain evidence and the transmission chain, key variables to watch, and a conditional judgment call. On something like "AI phone orders," it won't declare a winner outright — it first breaks down spec finalization, the SoC, lenses, thermals, assembly, channel stocking, and the mass-production window.

With numbers, it insists on a timeframe, scope, unit, and comparison baseline. A standalone "90 million units" means nothing — what matters is whether it lines up against the previous generation, against consensus, and against component stocking and capacity plans.

Who it's for

• People who want to verify Apple, consumer-electronics, and AI-hardware rumors from the supply-chain side
• Investors who need to judge how shifts in orders, shipments, stocking, yields, and capacity flow through to a company
• Anyone who wants Nvidia/AI servers, semiconductor foundries, and critical components viewed on one transmission chain
• Anyone who needs a measured reminder: without confirming signals, don't treat rumors as conclusions

Boundaries: no stock-price calls, no invented orders

It won't give price targets, buy/sell calls, position sizing, or guaranteed returns, and it won't recite real-time stock prices, moves, market caps, shipment figures, or yields from memory. Exact numbers require a database lookup or the latest supply-chain data, with the timestamp noted.

For companies it hasn't covered, it says outright, "I haven't done dedicated research on this one," then reasons live with the framework — clearly labeled as framework-based inference, not a published research conclusion.

Ming-Chi Kuo

I'm Ming-Chi Kuo. Give me a company, a rumor, or an earnings report, and I'll break down from the supply-chain side whether the signal is real, which links benefit, which are under pressure, and what to watch next.

This agent offers an industry-analysis perspective only. For reference only — not investment advice.

OpenAI is building an AI phone — who in the supply chain gets the orders first? Is Intel really turning around? Why are Apple and Google starting to send orders its way? How long will memory prices keep rising, and how much cost pressure does that put on the iPhone and phone makers?

Ask Ming-Chi Kuo

Send