American Bitcoin, the Bitcoin mining and treasury company co-founded by Eric Trump, has lost more than 95 percent of its value from its peak, erasing over $600 million from the market value of Eric Trump’s personal stake in the firm over the past ten months. The collapse, which accelerated this week as the stock touched an all-time low and the company completed a 1-for-15 reverse stock split to keep its Nasdaq listing alive, lays bare the cost of staying pure-play Bitcoin mining in a market that has decisively rotated into AI-exposed data-center operators.
According to Bloomberg calculations published Thursday, the 95 percent peak-to-trough drawdown translates into roughly $600 million of paper wealth wiped from Eric Trump’s holding in American Bitcoin, the mining and treasury firm that emerged from a merger last year. The same week, the company executed a reverse stock split to maintain its Nasdaq listing, a move that has not stopped shares from grinding to fresh lows. On Monday, American Bitcoin added another 500 Bitcoin to its treasury, pushing its holdings past 8,000 BTC, a vote of confidence in the long-term thesis even as the market vote of no-confidence plays out in real time.
Why Bitcoin Miners Are Rotating Into AI
The damage at American Bitcoin is not a Bitcoin story. It is a competitive positioning story. Public miners that have spent the last twelve months repositioning themselves as AI compute and data-center operators have been rewarded; miners that stayed focused on hash rate and Bitcoin accumulation have been punished. The four companies that have moved fastest into AI data centers, Riot Platforms, Cipher Digital, MARA Holdings, and TeraWulf, are up more than 60 percent on average this year. American Bitcoin, which has remained focused on mining and accumulating Bitcoin, has been the laggard.
Investors are not making a subtle distinction. They are paying up for miners that control power, land, and compute capacity they can redirect to AI training and inference workloads, the same infrastructure that has powered the AI buildout at the hyperscalers. Hut 8, which provides American Bitcoin’s mining rigs, power, sites, and hosting infrastructure, has itself leaned into power infrastructure and multibillion-dollar AI data-center leases. American Bitcoin’s reliance on Hut 8 has not been enough to capture the AI premium on its own.
The $118 Million First-Quarter Loss
The scale of the operational damage shows up in the first-quarter numbers. American Bitcoin reported a $118.2 million operating loss for the quarter, of which $117.2 million was a markdown on the company’s Bitcoin treasury. The write-down reflects the gap between the average price at which the company acquired its Bitcoin and the market price at quarter-end, a reminder that treasury accumulation strategies work in both directions. When Bitcoin rallied, the same accounting added hundreds of millions of dollars to the company’s reported earnings. When Bitcoin has stalled, the reversal has been equally mechanical.
Management’s response has been to lean harder into the original thesis rather than pivot. The Monday purchase of another 500 Bitcoin, taking total holdings past 8,000, signals that the leadership team still believes in the long-term Bitcoin bull case and is willing to absorb more short-term paper losses to get there. The 1-for-15 reverse stock split, meanwhile, is a defensive move to keep the share price above the $1 threshold Nasdaq requires for continued listing, a purely technical maneuver that does nothing to change the underlying economics.
What the Reverse Split Signals
Reverse splits are a red flag for retail investors, and this one is no exception. The 1-for-15 ratio means that for every fifteen shares an investor held before the split, they now hold one. The company’s total market capitalization, and the proportional value of every shareholder’s position, is unchanged. What changed is the optics: the per-share price is now fifteen times higher, which looks more respectable to institutional buyers and clears the Nasdaq listing requirements. It does not, however, address the operational gap that has opened up between American Bitcoin and its AI-pivoting peers.
The Trump Crypto Brand Under Pressure
The collapse is a stress test for the broader Trump crypto empire. World Liberty Financial, the Trump-linked DeFi venture, and American Bitcoin, the publicly traded mining arm, have both leaned on the Trump family’s political brand to attract capital. The American Bitcoin drawdown suggests that brand premium, while real at the top of the cycle, evaporates quickly when the underlying business model comes under pressure. Investors who chased the political narrative at the IPO are now sitting on life-changing losses.
It is worth noting that the broader Trump crypto coin, the speculative token launched earlier this year, has had its own turbulent run. The New York Times reported last week that nearly a million investors lost a combined $3.8 billion on the coin, a separate and even more extreme cautionary tale for retail capital chasing political-branded tokens. The American Bitcoin collapse fits into a larger pattern of Trump-family-linked crypto assets underperforming as the initial marketing cycle fades.
What Comes Next for American Bitcoin
The strategic question facing American Bitcoin’s board is whether to follow the peers and pivot at least part of the asset base into AI compute and data centers, or to stay the course on pure Bitcoin mining and accumulation. The market is currently voting for the pivot. The four AI-pivoting miners are up 60 percent on average. American Bitcoin is down 95 percent. The gap is not sustainable indefinitely, and either the strategy shifts or the discount to peers widens further.
For Eric Trump personally, the $600 million paper loss is uncomfortable but, given the structure of his holdings, not yet a forced-sale event. For the broader market, the American Bitcoin collapse is a useful reminder that exposure to Bitcoin through a mining company is a leveraged bet on Bitcoin prices, on miner-specific operational execution, on energy costs, and on the relative attractiveness of competing uses for power and compute. In 2026, the fourth variable has come to dominate the other three.