Illustration: a woman and a man toast with a friendly AI robot analyst, a glowing bitcoin coin above it, and a city window showing a price ladder from $84K to $250K and $1 million.
TradingFloor AI Special
Intelligence edition · 26 September 2026

Bitcoin to $250K? To $1 Million? What It Would Mean for Four Treasury Stocks

Regulators are writing rules, AI agents are learning to trade, and four listed companies hold 846,000, 17,211, 2,878 and 182 bitcoin. We ran AI Deep Dives on Strategy, Coinbase, Smarter Web and Coinsilium and stress-tested each against the two big numbers.

Bitcoin trades near $84,200. For it to reach $250,000 it must rise about 197%; for $1 million, about 1,087%. Neither is a forecast we are making. Both are numbers serious institutions have put on the table, and both are worth understanding because a small group of listed companies turns bitcoin's price into an equity story. Three forces are converging. First, regulators: the US SEC has moved from enforcement to rulemaking, proposing Regulation Crypto Assets on 18 August 2026 (still a proposal, in its comment period). Second, adoption: US spot bitcoin ETFs took in about $433 million on a single day this month, and Coinbase reports that more than 97% of on-chain agentic transactions in Q2 ran on its x402 protocol. Third, AI: autonomous trading agents and bots are moving from novelty to daily tooling, which puts more machine-speed demand on exchanges, custody and settlement. We deep-dived the four names people ask about: Strategy ($MSTR), Coinbase ($COIN), The Smarter Web Company (#SWC) and Coinsilium (#COIN on Aquis). The results are more nuanced than the headlines. Strategy trades below the value of its bitcoin; Smarter Web and Coinsilium trade above theirs; Coinbase is not a treasury story at all. Our AI rated all four Hold, with elevated to high risk. Bitcoin at $250K or $1M would transform their balance sheets, but the same leverage cuts both ways. This is research, not advice, and every figure below is either sourced or labelled as our arithmetic.

Bitcoin does not need to go to $1 million to matter. These companies just need the market to believe it can.

A treasury company is a bet on one price, wrapped in a share price. The wrapper can trade at a premium, at a discount, or at a level that has nothing to do with the coins inside it.

The two numbers: what the people saying them actually said

$250,000 in 2026 is the bull case, and it is now a minority view. Tom Lee has held a $250K target for the 2026 cycle, arguing the four-year cycle is breaking down under ETF demand and institutional accumulation, though he has also warned 2026 would be 'jagged' and has at times pointed to $150K-$200K. JPMorgan has cited about $170K. Citi, Standard Chartered and Bernstein have all cut targets during 2026, so the centre of gravity among mainstream forecasters has moved lower. With bitcoin near $84K, the analyst range for this year runs from the mid-$20Ks to $250K. $1 million is a long-horizon claim. ARK's Cathie Wood has published a 2030 base case around $750K with a bull case near $1.25M, and has since lowered a separate headline target to about $1.2M, saying stablecoins are taking over some of the payments role bitcoin was expected to play. Michael Saylor has said $1M before the end of the decade. Bitwise's Matt Hougan has pointed to roughly a ten-year path. Reaching $1M by 2030 would need compounding of roughly 87% a year from today's price. What we take from it: the range of expert opinion is enormous, the higher targets depend on adoption events that have not happened yet, and several credible houses have moved down, not up. We treat $250K and $1M as scenarios to test balance sheets against, never as predictions.

Why regulation, ETFs and AI agents are the three levers

Regulation. On 18 August 2026 the SEC proposed Regulation Crypto Assets: a 'fit-for-purpose' framework with a startup exemption (up to $5M over four years), a fundraising exemption (up to $75M per 12 months) and a conditional safe harbor for delinking a crypto asset from the investment contract it was sold under. To be precise about what it is: a proposal, open for public comment for 60 days after Federal Register publication, and it concerns how crypto assets can be offered, not a blanket licence for trading. But the direction is clear. The SEC has moved from enforcement-by-guidance toward rules that companies can plan around, and clearer rules lower the cost of holding crypto for regulated firms. ETFs. US spot bitcoin ETFs recorded roughly $159.5M of net inflows on 17 September and $433M on 18 September 2026, led by Fidelity's FBTC and BlackRock's IBIT. A Coinbase institutional survey found 66% of institutional crypto investors favour ETFs for exposure. That demand competes with treasury companies for the same marginal dollar, which is one reason treasury stocks can trade at discounts. Glassnode also noted net selling from treasury companies in August, the first such outflow in its series. AI agents. Autonomous agents that read on-chain data, prepare orders and settle transactions are moving from experiment to routine. Coinbase said more than 97% of on-chain agentic transactions in Q2 2026 used its x402 protocol, and that Base is the top chain for bitcoin DEX spot volume. Third-party estimates put the AI trading-platform market at about $13.5B in 2025 growing about 20% a year, though such market-size forecasts vary widely. The investment logic is that machines trade around the clock, so demand for exchange, custody and settlement infrastructure grows with agent adoption, whether or not the price of bitcoin does.

The scenario table: what each company's bitcoin would be worth

This is our arithmetic, not a target. It multiplies each company's reported bitcoin by three prices, using bitcoin at $84,218 (25 September 2026). It ignores debt, preferred stock, dilution, taxes and every operating business, so it is a ceiling on 'coin value', not on the share price. Strategy ($MSTR): 846,000 BTC (8-K, as at 20 Sep 2026; average cost $75,416) Now $71.2B | at $250K $211.5B | at $1M $846B Coinbase ($COIN): 17,211 BTC (Q2 2026 results) Now $1.45B | at $250K $4.3B | at $1M $17.2B The Smarter Web Company (#SWC): 2,878 BTC (company statement, 29 May 2026) Now $242M | at $250K $720M | at $1M $2.88B Coinsilium (#COIN, Aquis): 182 BTC (company announcement, 22 Jun 2026) Now $15M | at $250K $45M | at $1M $182M Against today's market capitalisations, our arithmetic gives these multiples of bitcoin value to market cap: Strategy about 1.3x-2.1x now (the two market-cap figures in our data disagree, so we show a range), rising to roughly 3.9x-6.3x at $250K and 16x-25x at $1M. Smarter Web about 0.76x now, 2.3x at $250K, 9x at $1M. Coinsilium about 0.85x now, 2.5x at $250K and 10x at $1M. Coinbase's bitcoin is small next to its business, so we do not compute a treasury multiple for it: $1.45B of bitcoin compares with $1.2B of revenue in a single recent quarter. Read it carefully: those multiples describe balance-sheet leverage, not share-price forecasts. Markets reprice treasury companies for financing costs, dilution and sentiment, and the stocks have historically swung more than bitcoin itself in both directions.

AI Deep Dive: Strategy ($MSTR) - the giant trading below its coins

Our AI's verdict: Hold, buy-rating 55/100, risk 58 (elevated), 3-6 month horizon. Headline: 'a bitcoin treasury trading at a discount, with an AI pivot and high volatility.' The bull case: the 846,000 BTC are worth about $71B at today's price, 10.7% above the $76K average cost in our data, and the stock trades at roughly 0.48x-0.75x that value on a basic market-cap view (0.84x on an enterprise-value view including net debt of about $6.0B). Continued purchases and an AI-software narrative add optionality. The bear case: statutory net loss of about $4.0B in 2025, deeply negative free cash flow, rising debt, and a share price that can decouple from the coins whenever financing gets harder. If bitcoin falls below $70K, our AI's bear scenario has the discount widening further. Its trade plan is technical and conditional: hold, with confirmation above $183 and invalidation at $156. Scenario tags in the deep dive: bull (bitcoin above $90K, discount narrows) 35%, base 40%, bear (below $70K) 25%. Data caveat: holdings come from a public treasury list cross-checked against Strategy's 8-K; the two market-cap figures in our data disagree, so treat the discount as a range.

AI Deep Dive: Coinbase ($COIN) - the picks-and-shovels play, not a treasury

Our AI's verdict: Hold, buy-rating 61/100, risk 66 (elevated). Headline: 'a crypto exchange with a bitcoin treasury, facing regulatory uncertainty and mixed market reactions.' Coinbase is the cleanest way to own the AI-agent-plus-regulation thesis without owning a treasury strategy. It holds 17,211 BTC after adding 819 in Q2 2026, but its business is trading, stablecoins, subscriptions and Base. Q2 2026 was weak: revenue fell to about $1.2B from $1.5B a year earlier, with a net loss of $359.5M ($1.36 a share), the third consecutive miss versus analyst estimates. Management points to record trading-volume market share, diversified revenue and agentic finance. A data warning we want on the record: a third-party treasury list paired with a stale market-cap field made our automated NAV read look like a 90% 'discount'. Our AI flagged it as implausible and marked the read 'unclear'. It is a data artefact, not a signal, and we have not used it. Other risks in the deep dive: a CME lawsuit, possible SEC rule changes, and 74% annualised volatility. Scenario tags: bull (bitcoin above $90K, regulatory headwinds ease) 30%, base 50%, bear (below $75K) 20%.

AI Deep Dive: The Smarter Web Company (#SWC) - the London bitcoin proxy

Our AI's verdict: Hold, buy-rating 59/100, risk 72 (high). Headline: 'bullish momentum meets bearish fundamentals.' The bull case: strong price momentum, outperformance of its peer group over three months, and 2,878 BTC, worth about $242M against a market capitalisation near $320M on our figures (about £236M). It listed on the London Stock Exchange main market in February 2026 and has proposed 'MORE' preferred shares to raise £15M-£25M for further bitcoin buying, a plan that would dilute or add senior claims and should be weighed as such. The bear case: deeply negative earnings on the operating business, negative free cash flow, about £22.5M of net debt, 90% annualised volatility and a history of drawdowns near 40% over 12 months. Bitcoin held is currently below its average cost in our data. Valuation note: our automated read used a public list showing 2,747 BTC, which is stale versus the company's own 2,878 BTC statement. With the newer figure the shares trade at roughly 1.3x bitcoin NAV (a premium), not the 1.38x in the raw deep dive. The exact multiple depends on the holdings date. Scenario tags: bull (bitcoin above $90K) 40%, base 40%, bear (below $75K or dilutive financing) 20%.

AI Deep Dive: Coinsilium (#COIN on Aquis) - the micro-cap with 182 coins

Our AI's verdict: Hold, buy-rating 54/100, risk 75 (high). Headline: 'a venture firm focused on blockchain and AI, trading at a premium to its bitcoin treasury.' Coinsilium holds 182 BTC through its Gibraltar subsidiary Forza, worth about $15M at today's price against a market cap near £13.3M. Bitcoin is up about 31% since its last holdings announcement (22 June 2026), so the true NAV has likely risen since. Net debt is low, but the operating business is small: net losses widened to about £5.6M in 2025 on almost no revenue, and cash is thin at about £1.4M. Why it matters for the $250K/$1M question: at 182 BTC the leverage to bitcoin is large relative to the company's size. Our arithmetic gives 2.5x today's market cap at $250K and 10x at $1M. The same size means the shares are illiquid, volatile and exposed to funding risk and dilution. Our AI's data caveats: the holdings figure is 96 days old, the GBP/USD rate was 29 days old at the time, and NAV counts only the bitcoin named in announcements. Scenario tags: bull (bitcoin +20% plus portfolio funding news) 40%, base 40%, bear (bitcoin -15%) 20%.

What would make the thesis wrong

Bitcoin can fall as easily as it can rise: the stocks above are leveraged to that price in both directions. Three concrete ways this story breaks. Regulation stalls. Regulation Crypto Assets is a proposal. Comment periods, litigation, a change in Commission priorities or Congress can slow or reshape it. Treasury companies lose their funding advantage. Strategy has traded below the value of its bitcoin on some measures; if the market stops paying premiums, raising money to buy more bitcoin becomes expensive, and Glassnode has already seen treasury companies net selling. Smarter Web's and Coinsilium's premiums can disappear quickly. Stablecoins and ETFs absorb the demand. ARK itself trimmed a long-term target because stablecoins are taking part of bitcoin's expected payments role. ETFs give institutions bitcoin without owning a company's balance sheet. For each company, the conditions that would change our AI's mind are set out on its deep-dive page.

Read the full AI Deep Dives and the sources

Deep dives are AI-generated research from evidence in our database; each page shows its data, scenarios and caveats. External sources are listed for the figures quoted above. Source cut-off 26 September 2026. Scenarios are illustrative, not forecasts; nothing here is investment advice, and capital is at risk.
Alpha · edition-aware

Question the markets.

Challenge a conclusion, compare assets, or explore what would invalidate the setup. Alpha reads this edition and remembers your follow-ups.

Evidence-led market intelligence, not personal investment advice. Company references are a market map, not recommendations; facts and conditions may change.