Barclays re-evaluates Baidu: for the first time includes Kunlun Chip in the valuation system, significantly raises the target price by 47%

Against the backdrop of the continuous rise of AI narratives, Barclays' latest research report has conducted a critical valuation reconstruction of Baidu.

According to the Wind Trading Desk, although the judgment on Baidu's core advertising business remains cautious, driven by the upcoming expectation of the Kunlun chip's spin-off listing, Barclays has officially included Baidu's AI chip business into its valuation framework for the first time, significantly raising the target price from $100 to $147, an increase of 47%, while maintaining a "Neutral (Equal Weight)" rating.

Barclays stated that this means Baidu's valuation logic is shifting from a single main line of "search + advertising" to a dual-track structure of "advertising fundamentals + AI hard technology assets."

This is not just adding an asset, but a change in Baidu's valuation "coordinate system"

On the surface, Barclays has only done one thing: added the Kunlun chip asset to the existing valuation model and raised the target price by 47% based on this.

But in essence, this is a switch in the valuation coordinate system.

How has the market priced Baidu in the past? For a long time, Baidu has been placed in a very "narrow" framework:

• Core anchor: Search advertising

• Secondary variables: Cloud business, autonomous driving (highly discounted)

The overall pricing method essentially views Baidu as a mature internet company, layering on the dual discounts of policy uncertainty and slowing growth.

In this system, all "future narratives" are met with one question: "When can we fill the advertising gap?" As long as advertising struggles, it is difficult for other businesses to truly lift the valuation. What essential change has Barclays made this time?

For the first time, Barclays acknowledges a fact:

Baidu already possesses an "AI hard technology asset" that does not rely on advertising, traffic, or C-end user behavior, which means that Kunlun chip is not a "business supplement," but an independent branch of the valuation method. This is also why Barclays did not raise the assumptions for advertising and cloud, yet still significantly raised the target price.

In summary, the essence of this step is: Baidu is for the first time allowed to "tell a growth story without relying on advertising."

Why now? What is the "value trigger" for Kunlun chip?

Many people may ask: Hasn't Kunlun chip been in development for five years? Why is it being revalued now? Barclays' answer is that the key lies not in "technological breakthroughs," but in the simultaneous maturation of three external conditions.

1. Domestic AI chips have shifted from "policy tasks" to "real demand."

In the past few years, domestic chips were more about policy guidance, strategic reserves, and demonstration projects, but the situation has changed. With the explosive demand for large model inference, computing power costs becoming a commercial bottleneck, and HBM/high-end GPUs continuously restricted, "usable, deliverable, and scalable" domestic inference chips have become a necessity for the first time.

The keyword used by Barclays in the report is "surge in demand for domestically-made chips." This is a very critical change in expression.

2. Kunlun Core's positioning is very "realistic," rather than "benchmarking Nvidia"

Barclays did not pit Kunlun Core against training chips or H100-level stories; its implicit judgment is that Kunlun Core's advantage lies in the inference end, forming a closed-loop use in Baidu Cloud, search, autonomous driving, and industry applications. This means it does not need to become the "world's strongest," but rather a sustainable part of the Chinese AI ecosystem.

This is precisely the underlying logic of the 15x revenue multiple:

It is not a valuation of disruptors, but a valuation of high-certainty growth-type hard technology.

3. IPO is not financing, but "outsourcing pricing power"

This point is very important, but many people overlook it. The significance of Kunlun Core's IPO for Baidu does not lie in how much money it raises or how much cash flow it alleviates, but in one thing: outsourcing the question of "how much is Kunlun Core worth" from the company internally to the capital market.

Once listed: Kunlun Core will have independent financial reports, will be compared horizontally with Cambrian, Biren, and Suiyuan, and will obtain a market price that can be continuously refreshed and revised. This is why Barclays wants to get in early rather than wait until after the listing.

Why did this revaluation not get blocked by poor advertising?

This is one of the most confusing points for many investors.

Barclays made it very clear:

• Search is being eroded by AI chatbots

• There are structural downward risks in advertising monetization

• Even lowered revenue and EPS forecasts

So why can they still raise the target price? The answer is simple: the market is beginning to accept that Baidu's future value no longer comes solely from "user attention."

On a deeper level: This is a reordering of the "AI business model"

In the old internet era: Users → Traffic → Advertising → Cash Flow

In the AI era, a new value chain is forming: Computing Power → Models → Industry Solutions → Infrastructure-level Revenue

Kunlun Core is precisely positioned at the upstream, the most difficult part to replicate. This is also why the advertising business uses EBITDA multiples, while Kunlun Core uses revenue multiples; one is a pricing for declining cash flow, and the other is a pricing for growth-type assets.

What is the real insight for investors from this revaluation?

If we step outside "Baidu as a stock," this research report is actually sending a larger signal.

1. The "second valuation curve" of Chinese concept technology is emerging

• First curve: Platform, Traffic, Advertising

• Second curve: AI Infrastructure, Computing Power, Hard Technology

Whoever can independently extract the second curve has the opportunity to break the long-term valuation suppression.

2. The watershed in AI investment: shifting from models to "deliverable computing power"

The market is gradually realizing:

• Models will converge

• Computing power and chips are the long-term bottleneck

Barclays' attitude towards Kunlun Core this time is essentially anchoring "Chinese AI computing power assets" in advance.

3. Baidu's core risks have also become clearer as a result From now on, Baidu's biggest risk is no longer just the decline in advertising, but whether Kunlun Chip can truly form external customers, whether it can break free from the valuation discount of "only used within the Baidu system," and whether it can withstand comparisons with peers after going public