I am "The Little Accountant," a long-time tracker of Robo.ai.
Recently, $Robo.ai Inc(AIIO)$ has been issuing announcements at a rapid pace. They just secured chip distribution rights, and today they announced a joint venture with Tachyon9 to build an AI data center. Many friends in my messages are excited, viewing this as "major bullish news."
However, as someone accustomed to reading financial reports, my first reaction is usually caution: Is the battlefront stretching too long? Can the cash flow sustain this? If the business lacks focus, will it become a "jack of all trades, master of none"?
Today, I want to set aside emotions and re-examine the Robo.ai acquisition of Jidu (Jiyue) $百度(BIDU)$ $吉利汽车(00175)$ purely from the perspectives of Financial Logic and Commercial Closed Loops, using the previously acquired "high-end chip distribution rights" as a variable combined with today's "self-built data center" news. Let’s skip the sentiment and calculate the Cost Ledger and the Risk Ledger.
I. Tesla's "Dojo Anxiety" vs. Robo.ai's "Cost-Reduction Maneuver"
To understand why Robo.ai needs to establish a joint venture with Tachyon9 to build a data center, we must first understand why Tesla built the Dojo Supercomputing Center.
1. $特斯拉(TSLA)$ Tesla's Ledger: Cloud Services are Too Expensive Elon Musk's decision to develop the Dojo chip in-house and build a supercomputing center was driven by one core logic: External computing power is too expensive and uncontrollable. As FSD (Full Self-Driving) data volumes grow exponentially, if Tesla continued to rent computing power from Amazon AWS or Microsoft Azure, the massive OpEx (Operating Expenditure) would devour the profits from car sales. Tesla's choice: Endure massive short-term CapEx (Capital Expenditure to build Dojo) in exchange for extremely low long-term unit training costs.
2. Robo.ai's Modified "Dojo Strategy" Robo.ai's current operation is essentially copying Tesla's homework, but with a "Middleman" buff added—Chip Distribution Rights.
The Tesla Model: Self-design chips -> Outsource to TSMC -> Assemble servers. Costs are low, but the R&D cycle is extremely long, and the risk is huge.
The Robo.ai Model: Secure high-end chip distribution rights (First-tier wholesale price) -> Sell to its own JV data center (Internal Transfer Pricing).
Financial Deduction: Compared to other computing centers that need to snatch graphics cards from the secondary market at a premium, Robo.ai's hardware procurement costs for building a data center are theoretically 15%-20% lower. This means that while Robo.ai lacks the capability for in-house chip design, through channel advantages, it compresses the D&A (Depreciation and Amortization) of the data center to levels comparable to self-research.
Conclusion: This is a clever "Dimensional Strike." Robo.ai is using trading advantages to replace the technical advantages Tesla used to achieve "low-cost computing power."
II. Jiyue Auto: Not Just a "Model 3," But a "Data Collector"
With the data center in place, the significance of Jiyue Auto (Jiyue) on the financial statements changes. It is no longer just inventory waiting to be sold, but the "Super Terminal" in the entire closed loop.
1. Data Closed Loop: Tesla's Moat Tesla's strength lies in having millions of cars on the road continuously transmitting data back to feed the backend supercomputer (Compute), training a stronger FSD model (Model), which is then pushed back to the vehicles via OTA. This is the "Closed Loop Logic" we are talking about: Jiyue generates data -> Transmits to Tachyon9 JV Data Center for training -> Model Iteration -> Feed back to Jiyue.
2. Financial "Left Hand to Right Hand" For Robo.ai, without this self-built data center, the smart driving training costs accompanying every Jiyue car sold would be a cash outflow paid to external cloud vendors. But now, this money becomes a "left hand to right hand" transaction: the computing fees paid by Jiyue become the revenue of the JV data center. From the perspective of the consolidated group statement, external procurement costs (OpEx) vanish, settling instead as internal assets.
III. Risk Warning: Tesla Has "Cash Capability," Does Robo.ai Only Have "Borrowing Power"?
While the logic perfectly benchmarks Tesla, as an accountant, I must throw a bucket of cold water: The financial foundations of the two are completely different.
1. The Disparity in Cash Flow Before betting on Dojo, Tesla already had global hits in the Model 3 and Model Y, generating terrifying Free Cash Flow (FCF) sufficient to support heavy asset investment. In contrast, Robo.ai just acquired Jidu, and Jiyue's sales are still in the ramp-up phase. Investing in a heavy-asset data center at this moment increases the pressure on the capital chain exponentially.
2. The Difference in Fault Tolerance If Musk fails with Dojo, he still has money from selling cars to fall back on. If Robo.ai fails to ramp up Jiyue sales, the data center it built becomes water without a source. Jiyue's delivery volume is the premise of all this logic. If the cars don't sell, there is no data return, and the data center will have to compete for customers in the open market against Amazon and Alibaba Cloud. In that scenario, Robo.ai completely loses its unique commercial closed-loop advantage.
IV. The Little Accountant's Summary: A High-Odds "Imitation Show"
Robo.ai's series of operations reveals massive ambition. It is attempting to use a trinity of "Trade (Chips) + Infrastructure (Data Center) + Manufacturing (Jiyue)" to replicate the cost advantages Tesla achieved only through "Full-Stack Self-Research."
My Judgment:
Strategic Score: S-Class. The logical closed loop is beautiful and theoretically constructs an extremely deep cost moat.
Execution Difficulty: SSS-Class. Opening three battlefronts simultaneously with limited funds is dancing on a tightrope.
Advice to Investors: Do not just stare at the "Good News." Stare at the Cash Flow Statement in the next quarter. See if the quick money earned from the Chip Distribution Business can actually cover the appetites of the two "gold-swallowing beasts": The Data Center and Car Manufacturing. If it can, this is the prototype of the next giant; if not, it is a gamble.
(Disclaimer: This article is a logical deduction based on public information and represents the personal views of "The Little Accountant" only. The market involves risks; investment should be cautious.)
Comments