On November 5, Bitdeer Technologies Group (NASDAQ: BTDR) closed at $5.31, down nearly 19% for the day after touching an intraday low of $5.03.

Compared with its 12-month high of around $23.30, the stock has lost roughly 77%. It has fallen nearly 30% over the past five trading sessions. Investors who bought at $8 to $10 are now sitting on substantial losses.

Yet Bitdeer's industrial expansion has not stopped.

In August, the company announced that its Tydal project in Norway had secured a roughly $4.7 billion, 16-year AI/HPC data center contract. AI infrastructure projects in Malaysia, the US and elsewhere are also moving forward.

More notably, Computing Power Heart has learned exclusively from people familiar with the project that Bitdeer's planned 1 GW AIDC campus in Kuantan, Malaysia, has received approval for 1 GW of power capacity. Bidding for the EPC contract covering the first 200 MW data center phase has concluded, and a contractor has been selected.

Bitdeer's stock is plunging even as its large AI data center projects continue to advance. Why?

In my view, this is more than a matter of market sentiment. Wall Street is reassessing the financing capacity of the entire AI infrastructure sector, and Bitdeer sits at the intersection of two valuation frameworks: Bitcoin mining and AI data centers. It has advantages that neoclouds lack, but faces the same unavoidable pressure from capital expenditure.

More importantly, Jihan Wu needs to answer a question: the company can keep expanding its asset base, but when will existing shareholders begin to benefit from the value those assets create?

Both Valuation Frameworks Are Breaking Down

Bitdeer now finds itself in an awkward position. It is no longer a pure-play Bitcoin miner, but the capital markets do not yet value it as a mature AI data center company.

In the past, that dual identity was an advantage. When Bitcoin was rallying, Bitdeer could pitch its mining capacity, proprietary mining rigs, BTC reserves and operating cash flow. When AI was in favor, it could highlight its power resources, data centers, GPU deployments and long-term leases. Both narratives worked, and both offered the prospect of a valuation premium.

Now, the opposite is true.

Start with crypto mining. Bitcoin mining hashprice recently fell as low as roughly $33 per PH per day. For miners heavily dependent on spot mining revenue, margins have been squeezed sharply. Electricity costs, mining rig depreciation and operating expenses do not fall in step with hashprice. Lower mining profits also mean less internal cash flow may be available to fund the AI transition.

Then there is AI. The market's attitude toward miners pivoting to AI has changed. In the past, a miner could command a valuation premium simply by announcing that it controlled several hundred megawatts of power or planned to convert a mining farm into an AI data center. Investors are now asking more practical questions: Has the power actually been connected? Has the facility been built? Have customers signed contracts? When can it be delivered? How much more capital is required?

According to Bitdeer's financial results for the second quarter of 2026, quarterly revenue was approximately $229 million, up about 47% year over year, while its net loss still reached $92.3 million. The business is clearly expanding, but revenue growth has yet to translate into consistent net profits.

At the same time, the company still needs to raise capital to support its mining rig business, AI infrastructure and other expansion projects.

Financing is not inherently a problem for a fast-growing company. But for public-market investors, the form of financing matters greatly. The lower the share price, the more shares the company must issue to raise the same amount of equity capital, increasing dilution for existing shareholders.

That creates a practical problem: the business can keep growing while the value attributable to each share does not necessarily rise with it. This is one of the central valuation pressures Bitdeer now faces.

Kuantan's 1 GW Campus Awards EPC Contract for First 200 MW

Even as Bitdeer's share price continues to fall, Jihan Wu's AI data center expansion in Southeast Asia is making tangible progress.

Computing Power Heart has learned exclusively from people familiar with the project that Bitdeer is advancing a planned 1 GW AIDC campus in Kuantan, Malaysia.

According to information provided by people familiar with the project, the campus has received approval for 1 GW of power capacity and is planned to include five data center buildings, each with approximately 200 MW of capacity. Bidding for the EPC contract covering the first 200 MW phase has concluded, and a contractor has been selected. Bitdeer is expected to serve as the ultimate project owner and appoint the EPC contractor to carry out construction.

This means the Kuantan project is no longer merely at the land-banking or long-term planning stage. It has begun selecting contractors for concrete engineering work.

For the AIDC industry, this is a significant development. A 1 GW data center campus involves far more than a few server buildings. Every component, from grid connections, substations and power distribution systems to cooling, network infrastructure and final server deployment, requires substantial capital and long-term coordination. This is particularly challenging as AI data center construction worldwide faces constraints involving transformers, power distribution equipment, construction resources and grid-connection timelines.

Securing land and power is the first hurdle. Finding a capable EPC contractor and completing construction is the second. Delivering on time and ensuring that customers continue paying rent is the third hurdle, and the one that ultimately determines investment returns.

Still, approval for 1 GW of power capacity does not mean that all of that power has been physically connected, nor does it mean the campus already has 1 GW of deliverable IT load. Likewise, completion of EPC bidding for the first 200 MW phase does not mean that all construction has begun. As of publication, the Kuantan project's detailed construction schedule, capital expenditure, grid-connection progress and end customers have yet to be confirmed.

Based on the information currently available, however, Bitdeer's Southeast Asian AIDC expansion is moving from project reserves toward concrete execution. That makes one question even more pressing: if major projects are advancing on the ground, why does the capital market remain unwilling to assign the company a higher valuation?

Wall Street's Real Concern Is Who Will Pay for It

To understand Bitdeer's falling share price, it must be viewed in the context of the broader US financing environment for AI infrastructure.

Wall Street is no longer concerned only with whether AI demand can keep growing. It is asking a more practical question: Who will pay to build all these AI data centers?

Over the past two years, Bitcoin miners, neoclouds and infrastructure developers have flooded into the AI market. Some have borrowed to buy GPUs, some have raised capital to build facilities, and others have locked in power and land ahead of time, betting on AI computing power demand over the next decade or more. They share one defining feature: enormous upfront capital expenditure and long payback periods.

The problem is that capital markets are recalculating the economics. Some high-risk AI infrastructure financing now faces all-in funding costs in the double digits, while concerns have also emerged over the terms of high-yield debt and convertible bond financing.

It is important to distinguish among a convertible bond's coupon rate, issuance yield and all-in financing cost after accounting for the value of the equity conversion option. They are not the same. But the market's message warrants attention: AI infrastructure remains a sought-after asset class, yet capital is not unlimited, and not every company can obtain low-cost financing.

For a company building AIDC capacity at scale, an increase in financing costs from 5% to 10% is far from trivial. If a project requires $1 billion in financing, a five-percentage-point difference in interest rates translates into an additional $50 million in annual interest expense. If construction takes two or three years and customer rent has not yet begun flowing steadily, those costs will continue eroding future returns.

More troubling, higher financing costs also raise a project's minimum required return. A long-term lease may look enormous on paper, but after construction costs, financing expenses and ongoing operating costs, the profit left for shareholders may be far smaller than expected.

This is why the market is scrutinizing the balance sheets of neoclouds and Bitcoin miners pivoting to AI. CoreWeave, IREN and Nebius have different business models, but all must balance capital expenditure, financing capacity and future customer revenue. Investors once readily paid a premium for revenue expected to grow rapidly over the next several years. Now they are increasingly focused on how much capital is required to generate that revenue.

A $4.7 Billion Contract Is Not $4.7 Billion in Cash

Bitdeer faces the same issue.

In August, the company announced that its Tydal project in Norway had signed a long-term AI/HPC data center leasing and services agreement worth approximately $4.7 billion. The contract covers 121 MW of IT load over 16 years. It is unquestionably a substantial deal.

But a large contract value does not mean the capital markets will necessarily value the company at the contract's full amount.

First, the direct tenant is a Volta subsidiary, not an investment-grade technology giant such as Microsoft or Google. Second, disclosed terms give the tenant the right to exit in year 10 without paying a termination fee. Although the contract includes a letter-of-credit support arrangement of approximately $1.3 billion, that credit protection remains subject to agreed conditions. More importantly, substantial capital investment is still required between signing the contract, delivering the project and generating stable rental income.

The $4.7 billion contract therefore cannot be treated simply as $4.7 billion in locked-in cash revenue. Its ultimate value depends on customer performance, construction costs, the financing structure and long-term cash flow.

Seen in that context, the logic of the 1 GW Kuantan project becomes clearer: the larger the project, the greater its potential asset value, but also the more funding it may require upfront. Whether Jihan Wu can secure projects is no longer the capital market's only concern. The next critical test is whether he can build them with sufficiently low-cost capital while avoiding excessive dilution for existing shareholders.

Mining Still Generates Cash, Setting Bitdeer Apart From Neoclouds

That said, it is not entirely fair to group Bitdeer with every neocloud.

Bitdeer has an easily overlooked advantage: it is not wholly dependent on external financing while waiting for AI projects to generate revenue. It also operates a business that has been running for years: Bitcoin mining.

According to financial disclosures for the first half of 2026, Bitdeer's self-mining revenue reached approximately $315 million, up about 226.5% year over year. The company mined 2,694 BTC in the second quarter.

That means Bitdeer still has a business capable of generating ongoing operating revenue even while its AI data center projects remain under construction.

More importantly, Bitdeer can develop its own mining rigs. It does not rely entirely on external suppliers for equipment and can allocate resources across chip development, mining rig manufacturing and mining farm operations. On its second-quarter earnings call, management said spending on wafers for its mining rigs was funded by cash generated from crypto mining and asset-backed financing, rather than equity issuance.

This is a key distinction between Bitdeer and some pure-play AI infrastructure startups. Those companies may need to continually raise capital to buy GPUs and lease data center space, then wait for customer revenue to cover their costs. Bitdeer, by contrast, can use its existing mining business to help fund its AI infrastructure transition.

Of course, mining revenue is not the same as net profit, nor can all of it be used to build AI data centers. When hashprice falls, the mining business itself comes under pressure. But as long as mining continues to generate positive free cash flow, it may reduce Bitdeer's dependence on external financing.

Bitdeer also has options. Where certain power resources are better suited to crypto mining, it can continue operating mining rigs. Where long-term returns from AI hosting are more attractive, it can consider converting the relevant infrastructure into AIDC capacity.

That flexibility does not mean a mining farm can be turned directly into an AI facility. AI data centers require higher standards for power reliability, networking, cooling and customer service, and conversions still demand substantial capital and engineering work. But this operating base does distinguish Bitdeer from pure-play neoclouds.

This is also why Bitdeer's entire business cannot be assessed solely through a neocloud valuation framework. It combines the cyclical operating cash flow of Bitcoin mining with the long-term asset value of AI data centers. If the two can work effectively together, Bitdeer may prove more resilient than some AI infrastructure companies that depend entirely on external financing.

The condition is that cash from mining cannot be continually consumed by inefficient expansion. Using mining to fund AI should be one of Bitdeer's advantages, not a justification for repeatedly asking investors to finance new projects.

Investors Who Trusted Jihan Wu Have Suffered Heavy Losses

From Tydal in Norway to Kuantan in Malaysia, Jihan Wu's long-term strategy is clear: he does not want Bitdeer to remain merely a computing power reseller. He wants the company to control more critical assets across the AI infrastructure value chain. Land, power, data centers, mining rigs, GPU deployment and computing power services can generate a degree of synergy.

In Norway, Bitdeer has already adopted a model in which it owns data center assets and provides customers with long-term hosting services. If the Kuantan project proceeds as planned, it could scale that model further. Controlling its own land, power and data center infrastructure would give the company greater operational flexibility. Whether it provides long-term hosting to major customers or deploys its own GPUs for an AI Cloud business, it would have more options.

But the cost is equally clear. Land costs money. Grid connections cost money. Substations, power distribution and cooling systems cost money. Building server facilities costs money. If Bitdeer also buys GPUs, capital expenditure will rise further.

This path can create valuable long-term assets, but it is also extremely capital-intensive. Many observers focus only on the headline value of AI data center leases without seriously calculating how much capital the company must invest upfront to secure those contracts.

I have long believed that the biggest advantage Bitcoin mining farms bring to an AIDC transition is not simply that miners understand electricity, cooling and server facilities. Their real advantage is experience building and operating high-density computing infrastructure. But AI data centers are far more complex than traditional mining farms. A mining farm can adjust equipment operations based on Bitcoin and electricity prices, while an AI data center must meet stringent requirements for power reliability, networking, cooling and customer service. Securing land and power is only the first step. Delivering the facility to customers and consistently generating the expected return on investment is the real test.

I have met many Chinese Bitcoin miners, and some have long placed considerable trust in Jihan Wu. That trust did not come from nowhere. From Bitmain to Bitdeer, Wu has accumulated experience in mining rigs, mining farms, power resources and the Bitcoin value chain that most entrepreneurs cannot easily replicate.

That is precisely why some miners built large positions in Bitdeer at prices between $8 and $10. The stock is now trading near $5. Someone who bought at $8 is down more than 30% on paper, while someone who bought at $10 has lost nearly half. These investors have suffered severe losses, and many bought largely because they trusted Jihan Wu.

I still believe Jihan Wu is one of the few entrepreneurs in the Bitcoin mining industry with genuine experience across technology, industrial operations and capital markets. Bitdeer also has advantages that are difficult for other companies to replicate: mining rigs, mining farms, mining revenue, power resources and AI data center projects already under development.

That is exactly why I am even more disappointed by Bitdeer's performance in the capital markets. Its communication with investors and management of market expectations have fallen far short.

I am not saying the company must guarantee a rising share price. Share prices are determined by the market, and no listed company can guarantee returns for investors. But a company that is undertaking large-scale capital expenditure, raising capital continuously and simultaneously operating Bitcoin mining, mining rig development, AI Cloud and data center infrastructure businesses has a responsibility to explain its capital allocation strategy more clearly.

For example: How much more capital does the Norway project require? How much of the financing has been secured? Is each delivery phase proceeding on schedule? How are the existing Malaysian projects progressing? When will they begin generating actual revenue? How much new financing will be needed over the next year? Will the company continue raising capital through share issuance? Most importantly, how will management balance business expansion against per-share value for existing shareholders?

These questions cannot be answered solely through a succession of project announcements.

To be clear, Bitdeer has not failed to disclose information altogether. It regularly publishes operating data and continues to announce project updates. But disclosing large volumes of project data does not mean investors can clearly assess future funding gaps and shareholder returns. Investors in a capital-intensive company need a coherent framework connecting project scale, construction costs, financing sources, delivery schedules and expected cash flow. If that framework remains unclear, the capital markets will naturally demand a larger risk discount.

Especially after such a steep share-price decline, management should communicate the company's underlying value more proactively. Could Bitdeer reasonably repurchase shares? Can it rely more heavily on project financing to reduce its dependence on issuing stock at depressed prices? Can it disclose the capital expenditure, financing arrangements and expected cash flows of major AIDC projects more clearly? And is management willing to take concrete action to demonstrate confidence in the company's long-term value?

Market capitalization management is not about manipulating the share price. It is about ensuring that the capital markets understand and recognize a company's underlying operating value. Jihan Wu has considerable personal credibility among Chinese Bitcoin miners, and that trust is a valuable corporate asset. But personal credibility cannot replace investor relations at a listed company, much less transparent financial disclosure.

Miners may trust Jihan Wu. But the capital markets will ultimately pay only for verifiable cash flow and shareholder returns.

Four Questions Bitdeer Must Answer on November 10

Bitdeer plans to report its third-quarter 2026 results on November 10.

In my view, the most important issues in this earnings report will not be how many additional megawatts of projects the company announces or how much the headline value of its AI contracts has increased. They will be several more practical questions.

First, how much cash flow can the mining business still generate, and can that cash continue supporting AIDC construction?

Second, can revenue growth from the AI Cloud business gradually improve overall profitability?

Third, are the financing, construction and delivery of major projects such as Tydal in Norway proceeding according to plan?

Fourth, as Bitdeer continues to expand, how will it control capital expenditure and the dilution risk facing existing shareholders?

If Bitdeer answers these questions clearly, the market may reassess its value. If the answers remain focused mainly on the project pipeline, future revenue and ambitious expansion plans, then even another multibillion-dollar contract may not ease investors' concerns.

I still believe Bitdeer is one of the companies worth following over the long term as the global Bitcoin mining industry pivots toward AI infrastructure. It has mining farm infrastructure, energy resources and industry experience that new entrants will struggle to replicate quickly. The award of the EPC contract for the first 200 MW phase of the 1 GW Kuantan project also indicates that its AIDC expansion continues to move forward.

But I also believe Jihan Wu and Bitdeer's management must take public-market investors more seriously. In the past, Bitcoin miners trusted Wu to build mining farms. Now Wall Street needs him to prove something else: that he can not only build AI data centers, but also make those assets generate real returns for shareholders.

No matter how large the business becomes, the capital markets will not keep paying for the story if shareholders cannot participate in the growth in value.

Information about the Kuantan project has not been publicly confirmed by the listed company, and its actual progress remains subject to subsequent project disclosures. Conflict-of-interest disclosure: The author owns shares in Bitdeer Technologies Group (NASDAQ: BTDR), is currently sitting on an unrealized loss, and has lost heavily. This article reflects both an industry assessment and the genuine experience of an ordinary retail investor.