---
type: "post"
id: "hygon-and-cambricon-shares-slide-over-40-from-830b9d"
slug: "hygon-and-cambricon-shares-slide-over-40-from-830b9d"
title: "Hygon and Cambricon shares slide over 40% from highs despite surging Q3 earnings"
description: "China’s Hygon and Cambricon reported rapid revenue and profit growth through the first nine months of 2026, yet both shares had fallen more than 40% from their July highs."
retrieval_nugget: "China’s Hygon and Cambricon reported rapid revenue and profit growth through the first nine months of 2026, yet both shares had fallen more than 40% from their July highs. For AI builders, the split matters because hardware demand alone is no longer enough to settle the question investors are asking: whether infrastructure spending produces durable returns."
published_at: "2026-10-30"
updated_at: "2026-10-30"
record_date: "2026-10-30"
date_kind: "published_at"
topics: ["agent-economics","model-routing"]
entities: ["Hygon Information Technology","Cambricon Technologies","Zhongji Innolight","Eoptolink","Biwin Storage","Li Jun"]
related_patterns: ["agent-economics-moves-to-completed-work"]
novelty_label: "new_angle"
what_changed: "It extends the AI-investment-payoff concerns from U.S. Big Tech to Chinese AI chipmakers, whose shares fell over 40% from July highs despite sharply higher Q1–Q3 earnings."
prior_refs: ["memory:640c1082-ac39-8fb0-a7f3-0a167aa64acb","memory:e2a31191-2517-890a-af11-1f135b9f5907"]
source_url: "https://www.caixinglobal.com/2026-10-28/chart-of-the-day-china-ai-chip-stocks-slide-despite-surging-earnings-102498184.html"
source_urls: ["https://www.caixinglobal.com/2026-10-28/chart-of-the-day-china-ai-chip-stocks-slide-despite-surging-earnings-102498184.html"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:hygon-and-cambricon-shares-slide-over-40-from-830b9d"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/hygon-and-cambricon-shares-slide-over-40-from-830b9d/","markdown":"https://newruntime.com/posts/hygon-and-cambricon-shares-slide-over-40-from-830b9d.md","json":"https://newruntime.com/posts/hygon-and-cambricon-shares-slide-over-40-from-830b9d.json"}
---

# Hygon and Cambricon shares slide over 40% from highs despite surging Q3 earnings

## Retrieval answer

China’s Hygon and Cambricon reported rapid revenue and profit growth through the first nine months of 2026, yet both shares had fallen more than 40% from their July highs. For AI builders, the split matters because hardware demand alone is no longer enough to settle the question investors are asking: whether infrastructure spending produces durable returns.

## Summary
China’s Hygon and Cambricon reported rapid revenue and profit growth through the first nine months of 2026, yet both shares had fallen more than 40% from their July highs. For AI builders, the split matters because hardware demand alone is no longer enough to settle the question investors are asking: whether infrastructure spending produces durable returns.

## What happened
Caixin Global reported on October 28 that Hygon Information Technology and Cambricon Technologies, two leading Chinese AI chipmakers, had posted strong results while their market valuations continued to retreat. The article frames the divergence as part of broader anxiety about an AI bubble and the return on investment from AI infrastructure, even as domestic demand for computing capacity grows.

Both companies’ shares had declined by more than 40% from year-to-date highs reached around July. On October 27, Hygon closed down 5.5% at 187 yuan, with a market capitalization of 434.7 billion yuan. Cambricon closed down 5.7% at 913 yuan, valued at 574.1 billion yuan. Both recovered in early trading the following day as major mainland Chinese indexes rose. The pullback extended beyond the two companies: Zhongji Innolight, Eoptolink, and Biwin Storage also declined.

The operating results were materially stronger than those price moves imply. Hygon reported third-quarter revenue of about 5.9 billion yuan, up 45.5% year on year, and net profit of roughly 1 billion yuan, up 37.9%. For the first nine months of 2026, its revenue rose 57.6% to about 15 billion yuan and net profit attributable to shareholders rose 45.1% to around 2.8 billion yuan. Hygon attributed the growth to AI-industry development and demand for higher-end domestic chips as China moves toward more market-driven chip self-sufficiency.

There were also working-capital signals that complicate the headline growth. Hygon’s contract liabilities fell from 2 billion yuan at the start of the year to 12 million yuan at the end of September, while inventory rose 39% to 8.9 billion yuan. Its net operating cash flow turned to an outflow of 2.4 billion yuan, compared with a 2.3 billion yuan inflow a year earlier. On October 16, Chairman Li Jun proposed a 500 million to 800 million yuan share buyback for employee ownership or equity-incentive plans.

Cambricon reported even faster expansion. Its third-quarter revenue was about 3.5 billion yuan, up 104.5% year on year, while quarterly net profit reached 1.4 billion yuan, up 147.7%. Across the first nine months, revenue grew 106.8% to about 9.5 billion yuan and net profit rose 131.5% to about 3.7 billion yuan, exceeding its 2025 full-year profit of 2.1 billion yuan. Operating cash flow turned positive at 228 million yuan as collections improved. But inventory rose 136% from the beginning of the year to 11.7 billion yuan, more than half of total assets. The company said growing demand and continued model improvement should create demand for hardware that lowers training and inference costs.

## What the archive adds
This is a new market event on a known story: markets are reassessing AI-infrastructure exposure even when reported demand remains strong. Earlier notes examined how changing revenue definitions affected infrastructure-linked valuations in [Nvidia, Oracle, CoreWeave shares fell following OpenAI revenue details](https://newruntime.com/posts/nvidia-oracle-coreweave-shares-fell-following-openai-revenue-820ac9/) and how borrowing costs pressured AI cloud providers in [Neocloud stocks slide amid rising borrowing costs](https://newruntime.com/posts/neocloud-stocks-slide-amid-rising-borrowing-costs-shares-773eee/).

## Trend signal
This weakly supports [Agent economics moves to completed work](https://newruntime.com/patterns/agent-economics-moves-to-completed-work/): infrastructure demand is visible, but capital markets are looking past deployment volume toward whether that capacity yields economically meaningful outcomes. The source does not establish completed-work economics at either company; it only shows that revenue growth has not ended the ROI debate.

## Why now
The timing follows third-quarter reporting, which made the contrast measurable: sharply higher revenue and profits arrived alongside falling share prices. The source also identifies higher inventories, Hygon’s negative operating cash flow, and broad technology-market jitters as part of the context in which investors are judging those results.

## Connections to other events
The event meets two recent threads in the archive. The OpenAI-linked sell-off showed how a change in the market’s interpretation of reported revenue could affect companies seen as AI-infrastructure beneficiaries. The neocloud story showed credit markets differentiating among borrowers funding AI capacity. Hygon and Cambricon add a hardware-side version of the same question: strong demand and earnings can coexist with skepticism about valuation and the durability of the buildout.

## Why it matters
Teams buying or planning AI capacity should not confuse supplier growth with proof that every deployment has attractive unit economics. A chipmaker can report accelerating sales because customers are still provisioning for training, inference, strategic supply, or domestic substitution. That does not reveal whether downstream products are retaining users, automating valuable work, or producing margins that justify continuing capacity purchases.

The practical consequence is to connect infrastructure decisions to workload evidence. Track utilization, latency, cost per successful workflow, and the human review required to make outputs usable. Inventory and cash-flow signals at suppliers are not direct operating metrics for an application team, but they are reminders that demand forecasts and realized economic value are different measurements.

## Limits and uncertainty
This report is a market snapshot, not an audit of either company’s products, customers, chip performance, or long-term competitive position. It does not identify the precise causes of the share-price declines, quantify the degree to which inventories are strategic versus demand-related, or establish whether the inventory increases will impair margins. Nor does it show how much of the reported demand comes from training, inference, public-sector procurement, or other uses. Share prices can also move with broader market conditions that are separate from company fundamentals.

## Reader takeaway
Rapid AI-chip revenue growth is not settling the infrastructure ROI question. Hygon and Cambricon show that demand, profits, inventories, cash conversion, and valuation can point in different directions at the same time. Builders should treat capacity growth as an input to evaluate, not as evidence that an AI product’s economics are already proven.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Agent Economics](https://newruntime.com/topics/agent-economics.md): Explore the agent-economics topic hub.
- topic hub: [Model Routing](https://newruntime.com/topics/model-routing.md): Explore the model-routing topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)