Bitdeer Has $2.9 Billion in Orders, Yet Its Shares Have Fallen Below $4
A company sitting on $2.9 billion in AI computing power orders should, in theory, be watching its share price surge.
Yet on November 18, 2026, Bitdeer Technologies Group (NASDAQ: BTDR) fell below $4 in intraday trading, touching a low of about $3.78.
Just one trading day earlier, the stock had closed at $4, down nearly 29% from the previous Friday.
With such a large order book in hand, why did the stock plunge in a matter of days, and what exactly is Wall Street worried about?
What Is Holding Up the $2.9 Billion Order Book?
Wall Street is not worried that Bitdeer cannot win business. The concern is whether it can deliver its massive contracted orders on schedule and collect the money without disruption.
First, it helps to understand the two numbers on Bitdeer's books.
At the end of the third quarter of 2026, Bitdeer's AI cloud business had only about $86 million in annualized recurring revenue, meaning the stable rental income generated by capacity already in operation.
But the total value of signed contracts that had yet to be delivered, its revenue backlog, had already reached about $2.9 billion.
That $2.9 billion is not cash Bitdeer can collect today. It must build the data center capacity and deliver the computing power over several years before the orders can gradually turn into revenue.
According to investment bank StoneX, the vast majority of this $2.9 billion backlog depends on two data centers in Malaysia.
The most important of these is the A201 data center in Johor Bahru, in Malaysia's state of Johor.
According to Bitdeer's third-quarter 2026 report, A201 has an IT load capacity of 21.7 megawatts.
It is a sizable computing power facility, with customers already reserving more than 70% of its capacity. This data center alone is expected to generate more than $1.7 billion in revenue over the next five years.
To fulfill the order, Bitdeer has ordered more than 100 top-end NVIDIA GB300 NVL72 racks, aiming to complete installation and deployment and begin generating revenue by the second quarter of 2027.
Given the scale of A201 alone, its delivery schedule could determine whether the broader $2.9 billion order book ultimately materializes.
The capacity was contracted and the equipment had been ordered, seemingly making this a straightforward deal. But two developments in mid-November suddenly cast uncertainty over the project.
Tencent Named as Two Policy Hurdles Emerge
Two reports released in quick succession around mid-November prompted Wall Street to revisit its assumptions.
First, on November 15, 2026, The New York Times cited a person familiar with the matter as saying that Tencent was likely the prospective computing power customer behind Bitdeer's A201 data center in Johor, Malaysia.
Bitdeer did not respond to The New York Times' request for comment at the time.
Bitdeer has not disclosed the project's customer list in its public filings, leaving outsiders unsure how much capacity Tencent might use or how much money could be involved.
Three days later, on November 18, The Block reported that StoneX analyst Mark Palmer had published a research note.
While maintaining a Buy rating and expressing confidence in the company's fundamentals, Palmer cut his Bitdeer price target from $22 to $15.
He also reduced his 2027 revenue forecast for Bitdeer by about $230 million to $1.77 billion, while cutting his forecast for adjusted EBITDA, a measure of core operating profit, by about $150 million to $616.5 million.
Why did the analyst lower his earnings expectations for next year immediately after The New York Times identified Tencent?
Together, the two developments exposed a distinct risk in Bitdeer's customer mix.
A comparison with its peers makes the issue clearer.
For U.S. neocloud companies that mainly serve domestic customers, the computing power business comes down to four challenges: raising capital, securing electricity, buying GPUs and bringing capacity online on schedule.
Once those four steps are complete, they can begin collecting revenue.
Bitdeer is different. It is building data centers in Malaysia to meet Chinese technology companies' demand for overseas computing power.
Chinese technology giants urgently need computing power outside China, and that demand has helped Bitdeer secure large orders early, giving it a significant advantage in expanding its market.
But it also leaves Bitdeer facing two additional policy hurdles that its U.S. peers do not.
The first is whether the chips can reach Malaysia on schedule.
In his report, Palmer warned that U.S. export license requirements for advanced chips shipped to Malaysia could directly delay Bitdeer's GPU deliveries.
Under the schedule he had previously seen, deliveries of NVIDIA GB300 systems were due to begin in November. If license approvals take longer, the equipment's arrival will be pushed back.
The second hurdle is whether the original customers can use the capacity even if the chips arrive and the data center is completed.
The New York Times reported on November 15 that the U.S. government was drafting new rules governing overseas data centers that provide Chinese customers with computing power powered by advanced GPUs.
One day later, on November 16, the Bureau of Industry and Security issued a related interim rule requiring licenses for advanced chips shipped to five countries, including Malaysia, as well as for remotely providing computing power to Chinese customers.
In Computing Power Heart's view, this is the variable Bitdeer needs to watch most closely.
Bitdeer has not disclosed how many Chinese customers it has or what share of total revenue or contracted orders they represent.
But if Chinese customers account for a large share of the Malaysian orders, a policy change could disrupt several otherwise separate contracts at the same time.
The report naming Tencent gave investors their first concrete look at how this policy risk might play out. How many orders could be affected will remain unclear until the company provides further disclosure.
If the Malaysian projects are held up, then, how steep a price could Bitdeer pay?
StoneX Models Three Scenarios
Rather than speculate, StoneX's November 18 report modeled three scenarios of increasing severity for potential delays or restrictions affecting the Malaysian projects.
The first scenario, assigned a 40% probability:
The U.S. approves the chip shipment licenses, or, even if the original customers cannot use the capacity, Bitdeer quickly leases the computing power to other customers. The entire project is delayed by only one quarter.
In this case, payments would arrive three months later than planned, with little impact on the broader outlook.
The second scenario, assigned a 35% probability:
The project comes online two to three quarters late, and Bitdeer must cut computing power prices by 15%.
Why would a delay of more than six months require a 15% discount?
The reason is straightforward. Bitdeer has already committed real money to building the data center and ordering more than 100 racks.
If licensing delays push delivery back by more than six months, market prices for high-end chips may have changed. Alternatively, if the original customers cannot use the capacity, Bitdeer would have to quickly find new buyers for a substantial amount of computing power.
Either way, buyers would gain bargaining power. To avoid leaving a completed data center idle and burning cash, Bitdeer might have to offer a 15% discount to secure contracts.
That would materially reduce the project's profitability.
The third scenario, assigned a 25% probability:
This is the most difficult assumption: the A201 data center cannot move forward at all, while restrictions also affect the use of Bitdeer's existing GPU fleet and cause it to lose some customers.
In that case, Bitdeer would have to search globally for new customers to absorb the capacity. If replacement customers demand later lease start dates or lower prices, both the total contracted value and the expected payment schedule would be disrupted.
Taken together, the three scenarios show that:
A delayed launch affects when revenue can be recognized;
A 15% price cut reduces the deal's ultimate profitability;
Losing customers means Bitdeer must find new buyers to cover the data center and equipment investments it has already made.
With these risks now in plain sight, what can Bitdeer do to ease investor concerns and restore its share price?
Three Steps to Regain Investor Confidence
With its shares below $4, management's immediate priority should not be telling a new story. It should break the $2.9 billion AI cloud backlog into delivery milestones, recognized revenue and cash collections that investors can verify quarter by quarter.
In Computing Power Heart's view, the company can start with three practical steps.
First, publish a verifiable delivery timeline for the Malaysian projects.
What markets fear most is uncertainty.
For each project, Bitdeer could disclose when the first GB300 systems are expected to arrive, when the racks will be powered on, when customers will accept delivery and when billing will begin, while clearly identifying which steps are still awaiting regulatory approval.
If A201 genuinely faces a delay, the company should publish a revised timeline early and explain candidly the potential range of impact on 2027 revenue.
It should update the market each time it meets a milestone.
Second, avoid depending on a single customer category and line up backup buyers for the computing power in advance.
To the extent permitted by confidentiality agreements, Bitdeer could provide more detail on customers' geographic distribution, customer concentration and the value of contracts potentially exposed to policy changes.
At the same time, it should accelerate customer development in North America, Europe and Southeast Asia, building a reserve order pipeline for any data center capacity that might otherwise sit idle.
If customers do need to be replaced, investors will want answers to three questions:
How much capacity can the new customers take, when will they begin paying, and how much of a discount will they demand? Preparing these alternatives would help keep the market calm.
Third, align spending with customer payments and prioritize projects that can begin billing immediately.
Bitdeer's original crypto mining business continues to generate profits, providing financial support for its AI transition. But its AI projects must also begin producing cash flow as quickly as possible.
Management should clearly disclose how much customer prepayment it has actually collected, how much project financing it has secured and how much more it expects to invest over the coming quarters.
It should first complete and monetize projects with firm customer commitments, secured funding and clear delivery conditions.
Further expansion should proceed in stages and strictly track contract signings, prepayments and financing progress.
This would both reduce the financial strain of borrowing to fund construction and lower the risk of diluting existing shareholders.
If Bitdeer can do only one thing right now, it should get the first Malaysian customers onto its GPUs and show the corresponding billing and payment progress in its next earnings report.
Once Wall Street can see the $2.9 billion order book being delivered in batches and revenue turning into cash quarter after quarter, the market will have a concrete basis for reassessing the true value of Bitdeer's AI business.
This article is based on public disclosures by listed companies, industry materials and information obtained by Computing Power Heart from people connected to the project. Information about the Kuantan project has not yet been publicly confirmed by the listed company, and its actual progress remains subject to subsequent disclosures. This article does not constitute investment advice. Conflict-of-interest disclosure: The author owns shares of Bitdeer Technologies Group (NASDAQ: BTDR), is currently sitting on an unrealized loss, and the loss is substantial. The views above reflect both industry analysis and the genuine experience of an ordinary retail investor. This article does not constitute investment advice.

Comments
00No comments yet. Be the first to weigh in.