The 99.8% Problem
Everyone says IPOs are how retail investors get early access to growth. They are wrong.
Bitari Inc. just filed an S-1 with the SEC for a $30 million Nasdaq listing under the ticker BIAI. The company runs bitcoin mining hosting services. It has nine months of financials. It plans to allocate 40% of net proceeds to acquisitions that don't exist yet.
Here's the number that matters: public investors will contribute 99.8% of the capital. They will receive 10% of the equity.
The chairman, Pei Zhao, through his holding company AI Power X Inc., paid $45,000 for his position. He owns 85.87% of the company. He will retain complete control after the listing.
Code doesn't care about fairness. Markets don't either. But this structure deserves scrutiny before anyone commits capital.
Context: The Mining Services Landscape
Bitcoin mining hosting is a mature, capital-intensive business. The model is simple: secure cheap electricity, deploy miners, charge hosting fees. Riot Platforms and Marathon Digital dominate the sector with billion-dollar market caps, massive hashrate, and institutional backing.
Bitari operates on a different scale entirely. Nine months of revenue: $8.37 million. That's less than 0.1% of the market share held by the sector's leaders.
The company's financial trajectory shows negative momentum. Revenue declined from $8.59 million to $8.37 million. Net income collapsed from $990,000 to $184,000. Operating cash flow sits at negative $690,000.
The S-1 filing reveals no proprietary technology, no patents, no competitive moat. This is a business that depends entirely on two external variables: bitcoin price and network hashrate difficulty. Neither favors small operators in the current cycle.
The company's tangible book value is $0.69 per share. The IPO price is $7. New investors face an immediate accounting dilution of $6.31 per share.
The Core: Dissecting the Capital Structure
The fund allocation plan reveals more about the strategy than any marketing material.
| Allocation | Percentage | Amount | |-----------|------------|--------| | Strategic acquisitions | 40% | $10.78 million | | Global expansion | 30% | $8.09 million | | New mining operations | 15% | $4.04 million | | Working capital | 15% | $4.04 million |
The 40% acquisition allocation is the critical piece. The company has identified no specific targets. There are no term sheets in progress. The S-1 provides no parameters for what types of companies might be acquired.
I audited a similar situation in 2023. The protocol had allocated $50 million for unspecified strategic initiatives. They eventually acquired a struggling DeFi project at a 300% premium to its market cap. The tokens I shorted after reading their transaction logs paid for my research.
Capital without targets is usually capital without discipline.
The 15% allocation for "new mining business and infrastructure" contains no technical roadmap. The company hasn't specified what type of miners they'd purchase, which jurisdictions they'd target, or how they'd secure competitive power rates. In a sector where operational excellence is everything, the lack of specificity signals a lack of planning.
The 30% global expansion allocation faces a similar problem. Brand recognition is minimal in the mining sector. Expansion without infrastructure partnerships or existing customer relationships is costly and uncertain.
The Controlled Company Loophole
Nasdaq rules allow certain companies to qualify as "controlled companies" when a single entity holds over 50% of voting power. The AI Power X entity crosses this threshold at 85.87%.
Controlled companies are exempt from several governance requirements: majority independent board, independent compensation committee, independent nominating committee. These exemptions remove the mechanisms that typically protect minority shareholders from conflicts of interest.
I've audited smart contracts with similar governance structures. The pattern is always the same: the operator has unrestricted access to user funds, and users have no way to challenge the operation. The Terra collapse in 2022 demonstrated what happens when a concentrated authority makes unreachable decisions with user capital. I survived that period by pre-allocating 60% of my portfolio to over-collateralized assets.
The S-1 doesn't explicitly state that the company will waive these requirements. But the structure allows it. That's enough uncertainty for most institutional investors to pass.
The 90% Non-Locked Position
Here's the detail that distinguishes this IPO from the rest.
Most IPOs impose lock-up periods preventing existing shareholders from selling for 90-180 days. This mechanism aligns the interests of existing shareholders with new investors. It prevents immediate dilution and protects price stability.
The S-1 includes no lock-up for existing shareholders. The 90% equity held by existing shareholders, including the chairman's 85.87% position, becomes immediately tradable after listing.
This creates a scenario: 90% of the company's stock becomes available for sale while only 10% is held by new investors. If the chairman decides to sell any portion of his position, the market will absorb the supply.
The incentive structure is clear: the new investors are the exit liquidity.
The AI Narrative
The ticker "BIAI" suggests AI association. The company's name references AI Power X Inc. as a major shareholder.
The S-1 contains no substantive AI technology. No algorithms. No data infrastructure. No machine learning deployment. The company is a bitcoin mining hosting provider.
The "AI+mining" narrative has been a popular theme in public markets over the past year. Mining companies have seen valuations surge when they announce AI-related partnerships or infrastructure upgrades. The market wants to believe that excess energy capacity can be redirected to compute operations.
The reality is more complex. Mining infrastructure is optimized for specific workloads. Transitioning to AI compute requires significant hardware investment, cooling system upgrades, and potentially different facility designs. The energy requirements differ, and the latency tolerances are different.
I audited an "AI-driven trading bot" in 2025 that claimed 30% monthly returns. The bot was executing high-frequency trades on decentralized exchanges, paying excessive gas fees, and generating minimal alpha. The narrative was the product. The code was just a wrapper.
Bitari may have similar plans to leverage the AI narrative to justify a higher valuation. The stock price of $7 per share gives the company a market cap of roughly $30 million. That's a multiple of approximately 3.5x annualized revenue, which is not absurd for the sector. But when you factor in declining revenue, negative cash flow, and a weak balance sheet, the AI narrative is the only justification for a premium.
The Path Forward
If the IPO succeeds, the structure looks like this: a 99.8% of new capital from public investors, an instant 90% dilution in book value, a chairman with absolute control, and a company with deteriorating financials.
The float will be approximately 4.3 million shares out of 43 million total. Low float is a recipe for price manipulation. With the majority of the shares held by a single entity, the price can be moved with small volume.
The historical pattern for such structures is predictable: the stock lists at $7, remains stable for a few weeks, and then the sell pressure begins. The chairman's shares are unrestricted. The company's financial deterioration continues. The "AI" narrative fades when no AI products emerge.
Arbitrage is just patience wearing a speed suit. The arbitrage here isn't a price difference between two exchanges. It's the gap between the $7 IPO price and the underlying reality.
The Contrarian View
There is a possibility this works out.
The company could surprise investors. The management might announce an actual acquisition target with revenue and infrastructure. Bitcoin price could rally significantly, boosting mining profitability across the sector. The AI narrative might be validated if the company successfully pivots toward high-performance computing.
Any of these events could support the stock price above the IPO level.
But none of these scenarios address the structural problem: the majority shareholder has no economic alignment with the new investors. If the company succeeds, the chairman captures 85.87% of the upside. If the company fails, the chairman can sell his position into the market, and the new investors absorb the losses.
Algorithms don't get scared. But the people who write them do.
The risk asymmetry is fundamental to the structure. It cannot be fixed by a good quarter or a promising announcement.
Speed is the only shield in a flash loan. In traditional markets, the equivalent speed is the lock-up period. Without it, the shield is gone.
The Takeaway
Bitari's IPO is a study in capital structure design. It's not a question of whether the business can be profitable—mining hosting can be a solid business. It's a question of who captures the value.
The structure captures value for the chairman and the existing shareholders. The new investors assume all the risk with no protection.
I audit the logic, not the hope.
The logic here is clear: 99.8% of the capital, 10% of the equity, no lock-up, no governance protections, no acquisition targets, no AI substance, and a deteriorating financial picture.
The company lists on Nasdaq. The SEC will review the S-1 thoroughly. The offering is legal, and the disclosures are sufficient under the rules. The rules don't protect against bad deals.
Trust the stack, verify the exit.
For the new investors, the exit is the problem. The float is small, the controller holds 85.87%, and there's no restriction on their selling. The exit is the chairman's game.
The $30 million raised is real. The business is real. The risk is the structure.
If the market rewards this structure, expect more companies to follow the same template: a 0.2% founder contribution, a 90% control position, an unrestricted float, and a compelling narrative to fill the gap. The problem is the pattern itself.
The question is whether the market will learn the lesson before the next one.
