On August 10, 2025, Bitdeer (Nasdaq: BTDR) filed a prospectus supplement that unlocked a $1 billion at-the-market (ATM) equity offering. The market read it as a bullish signal: capital for the Tydal AI data center, a pivot from Bitcoin mining to high-performance computing. The code never lies, only the auditors do. And here, the code is not Solidity—it is SEC filings, balance sheets, and the fine print of a $1.3 billion letter of credit. Tracing the silent bleed from 2017’s broken logic, I see a pattern: hype masking structural dilution. This is not a project announcement. It is a funding mechanism with a 30% dilution floor and a timeline that stretches into 2027. The question is not whether Bitdeer can build a data center. The question is whether the math of this capital raise leaves existing shareholders with anything but a promise.
Context: The Protocol Behind the Pivot
Bitdeer is a Bitcoin mining company that has extended its business into ASIC chip design and now AI data centers. The Tydal project, located in Norway (likely, based on the partnership with Volta Tydal AS), is a planned 75 MW AI data center with two phases: Phase 1 completion target December 31, 2026; Phase 2 by March 31, 2027. The project is backed by a $1.3 billion letter of credit from JPMorgan Chase and other financial institutions. But the $1 billion ATM offering is separate—it is an equity sales agreement that allows Bitdeer to sell up to $1 billion of its Class A ordinary shares at market prices, with no fixed schedule. As of August 10, the company had 227.4 million shares outstanding, with an additional 159.6 million shares reserved for the ATM. Using the example price of $10.88 per share, the ATM could issue approximately 91.9 million new shares, representing a 40.4% dilution relative to current shares. This is not a debt facility. It is a dilution tap.
Core: The Systematic Teardown of the Capital Structure
Let me break this down with the precision of a forensic audit. The ATM is not a one-time raise. It is a standing authorization to sell shares at the company’s discretion. The prospectus supplement states the funds may be used for "data center, AI cloud, ASIC R&D, working capital, and other general corporate purposes." This is a kitchen sink of uses. The $1 billion is not allocated to Tydal. It is a pool of potential dilution. The market naively conflates the two: the $1.3 billion letter of credit for Tydal and the $1 billion ATM for anything. The letter of credit is conditional—a "support" not a cash deposit. If Bitdeer fails to meet project milestones, the agreement can be terminated. This is a real risk. The Tydal project has not broken ground. The target start date is 2026, meaning at least 18 months of capital burn before any AI revenue. Where does the cash come from? The ATM. But the ATM is equity, not debt. Every share sold reduces the claim of existing holders.
The dilution math is brutal. At a $10.88 share price, the 91.9 million new shares would increase the total share count to 319.3 million. That is a 28.8% dilution on a post-money basis. But the ATM is not a fixed exercise. Bitdeer can sell shares at any price. If the stock price drops to $5, the company would need to issue 200 million shares to raise $1 billion, pushing dilution to 46.8%. If the stock price rises to $20, they would issue only 50 million shares, a 22% dilution. The market is pricing in the AI narrative, but the ATM creates a ceiling: every upward move incentivizes the company to sell more shares, suppressing the stock price. This is the silent bleed. The company has already used the ATM to raise $160.7 million since January 2025, selling 9.05 million shares. They are not novices. They will use it again.
Luna’s death was a math error, not a market crash. The same logic applies here. The error is not in the code of a smart contract but in the capital structure of a listed company. The market assumes Bitdeer will use the ATM to fund a high-return project. But the project has no revenue forecast in the filing. The Tydal data center is a capital expenditure with a two-year construction timeline. The return on equity depends on future AI compute demand, which is volatile. If the project fails to generate returns above the cost of equity (which is effectively the dilution cost), then every dollar raised via ATM destroys shareholder value. The absence of a projected IRR in the filing is a red flag. The company provides no scenario analysis. The market is expected to trust the narrative. Forensics reveal the truth markets try to bury.
The regulatory layer adds another dimension. Bitdeer is a US-listed company, so its ATM is SEC-compliant. But the legal risk lies in the gap between stated use and actual use. The prospectus supplement gives management broad discretion. If they use the $1 billion to cover operating losses or to buy Bitcoin instead of building AI infrastructure, they are within the legal framework. Investors have no recourse. The code never lies, only the auditors do. And here, the auditor is the SEC, which has limited capacity to police forward-looking statements. The letter of credit from JPMorgan is a positive signal, but it is not a guarantee. It is a credit facility that can be revoked if Bitdeer fails to meet conditions. The project is still in the planning phase, with no environmental permits for the Norway site disclosed. Complexity is just laziness wearing a tech suit.
Contrarian: What the Bulls Got Right
I am not a permabear. The bulls have a point. Bitdeer’s vertical integration is a real advantage. They own ASIC chip design, mining hardware, and now potential AI compute. The Tydal project, if executed, could provide a clean energy backbone for AI training, leveraging Norway’s hydropower. The $1.3 billion letter of credit is not just a paper promise. JPMorgan’s involvement suggests institutional vetting. The company has a track record of building mining farms. The pivot to AI is not a new idea—Core Scientific and IREN have done it. Bitdeer may be late to the party, but it has a unique asset: proprietary ASIC chips that could be repurposed for AI inference. The bull case is that the ATM is a tool, not a trap. If the stock price rises, dilution is minimal. If the Tydal project generates high returns, the dilution is justified. The market is pricing in a 2027 revenue stream, not a 2025 dilution event. The bull says: ignore the noise, focus on the asset.
The contrarian angle is that the bulls are correct about the asset but wrong about the timing. The project is at least 18 months from revenue. The ATM is live now. The company will likely sell shares aggressively in the near term to fund the construction phase. This is a classic pattern: early dilution to fund later growth. The problem is that the market is not discounting the dilution properly. The 30% potential dilution is not priced into the stock because the narrative is too strong. The bulls are buying the story, not the math. But the math always wins. The code never lies. The Tydal project may be a real asset, but the ATM creates a capital structure that rewards the company at the expense of existing shareholders. The bull case is valid only if the project generates returns above the cost of equity. That is a high bar for a project with no revenue forecast.
Takeaway: The Accountability Call
This is not a scam. It is a structured dilution. The market will eventually price it in—either through a stock price decline or through the realization that the AI narrative is a capex-heavy, long-duration bet. The on-chain trace here is not a wallet address but a SEC filing. The signal is clear: Bitdeer is using the ATM to fund a speculative project with no guaranteed return. The question is not whether the data center will be built. The question is whether the math of the capital raise will leave existing shareholders with a net positive outcome. The answer is not in the press release. It is in the quarterly reports of 2027. Until then, the silent bleed continues. Follow the gas, not the hype. The gas here is the ATM sales. And they are already burning.