Riot Platforms (RIOT) Stock Forecast 2026 and 2030: The $647M Revenue Miner Selling BTC to Build an AI Empire

Riot Platforms just reported its best financial year in company history. Revenue up 72%. 5,686 bitcoin mined. The AMD data center lease operational. Starboard Value — a significant activist shareholder — publicly valued the AI pivot at up to $21 billion, compared to Riot’s current market cap of roughly $6.3 billion.

The stock trades at $18.24.

That number — $18.24 against what the bulls are saying is a $21 billion opportunity — is the central tension in every RIOT investment thesis right now. Whether you’re looking at Riot as a Bitcoin miner, a data center infrastructure play, or both, you have to explain why the market is assigning a fraction of the theoretical value Starboard describes.

This analysis tries to do exactly that, without pretending either the bull case or the bear case is simple.

From Riot Blockchain to Riot Platforms: The Identity Shift

People who have followed this stock for years will remember when it was called Riot Blockchain — a name it adopted in 2017 when it pivoted from being a veterinary biotech company into a Bitcoin miner. The blockchain rebrand was, candidly, a market narrative play as much as a business one: the company’s stock rose more than 400% in the days after the name change was announced.

The 2022 rename to Riot Platforms was a more substantive signal. Bitcoin mining was being reframed as a data center problem — not a financial speculation problem. Running a profitable mining operation requires scale, cheap power, efficient hardware, and thermal management at industrial levels. Those capabilities translate directly to hyperscale data center operations, which is where AI compute demand has been going. Riot wasn’t just a miner any more; it was a power-heavy infrastructure operator with specific assets that look increasingly valuable to hyperscalers.

In Castle Rock, Colorado, CEO Jason Les now runs what the company describes as “a Bitcoin-driven industry leader in the development of large-scale data centers.” Note which word comes first in the job description these days. It’s not “mining” — it’s “data centers.”

That reframing is either prophetic or premature. April 30, 2026 — the Q1 2026 earnings call — will provide the clearest data yet on whether it’s happening on the timeline that justifies the stock’s current valuation.

FY2025 in Full: Record Revenue, Record Loss

Riot’s full-year 2025 results, reported March 2, 2026, are a study in contrasts. The headline numbers are impressive. The cost structure underneath is the honest problem.

FY2025 highlights:

• Revenue: $647.4 million — record, up 72% from $376.7 million in FY2024
• Bitcoin mining revenue: $576.3 million (up from $321.0 million) — driven by higher BTC prices and increased operational hashrate
• Engineering segment revenue: $64.7 million
• BTC mined: 5,686 — with deployed hashrate reaching 38.5 EH/s by year-end
• BTC holdings at Dec 31, 2025: 18,005 BTC (~$1.58B at $87,498/BTC, with 14,028 unrestricted)
• Cash on hand: $309.8 million
• Total liquidity: Over $1.9 billion
• Net loss: ~$663 million — reversing from prior-year profitability

The $663 million net loss is the number that stops every RIOT bull mid-sentence. Revenue grew 72% and the company still lost $663 million. Why?

Three reasons compound each other. First: the average cost to mine one bitcoin in FY2025 was $49,645, compared to $32,216 in FY2024. A 54% increase in mining cost per coin. This happened because the global network hashrate grew 47% year-over-year — meaning more competition for the same fixed Bitcoin reward. Post-halving, the reward per block is 3.125 BTC. The entire industry is running harder for a smaller prize.

Second: depreciation and amortisation on Riot’s aggressive hardware investment. They ordered $779.5 million worth of MicroBT miners — totalling 49.2 EH/s in new capacity — that’s hitting the income statement as D&A even before some of that capacity is deployed.

Third: the AMD data center lease only became operational in January 2026. The entire investment thesis around Riot as an AI/HPC infrastructure play is built on an asset class that was generating zero revenue through the end of 2025. A ten-year lease worth an estimated $311 million in contract revenue — which Riot and Starboard both describe as a “starting point” for much larger ambitions — didn’t start generating a dollar until Q1 2026.

The market wants to see data center revenue before it gives credit for data center value. That’s not irrational. It’s a reasonable demand for evidence.

Q1 2026: The Numbers That Actually Matter Right Now

The most current operational picture comes from Riot’s Q1 2026 production update (April 2, 2026):

• BTC mined in Q1 2026: 1,473 BTC — down 4% year-over-year from 1,530 BTC in Q1 2025
• BTC sold in Q1 2026: 3,778 BTC at an average price of $76,626, generating $289.5 million
• BTC holdings at March 31, 2026: 15,680 BTC (5,802 restricted) — down 18% from Q1 2025’s 19,223 BTC
• Deployed hashrate at March 31: 42.5 EH/s — up 26% year-over-year
• Average operating hashrate: 36.4 EH/s — up 23% year-over-year
• All-in power cost: 3.0 cents/kWh — down 21% year-over-year

The headline that generated the most commentary: Riot sold 2.57 times more BTC than it mined in Q1 2026. Production was 1,473 BTC. Sales were 3,778 BTC. That’s a deliberate drawdown of treasury reserves, not routine income management.

Riot has been explicit about why. The proceeds — $289.5 million — are funding the AI/HPC data center buildout. January’s $289.5 million from BTC sales financed the acquisition of 200 acres of land at the Rockdale site and initial infrastructure for the AMD lease expansion. The company is converting BTC treasury into physical infrastructure.

Whether that’s the right trade depends entirely on what happens to both BTC and data center lease rates over the next three years. If BTC continues appreciating and lease rates for AI compute remain elevated, Riot sold treasury assets at a relative discount. If BTC declines and AI data center demand softens, Riot sold at roughly the right time and funded infrastructure that would otherwise require expensive equity dilution.

The 3.0c/kWh all-in power cost is genuinely impressive and is the reason analysts give Riot premium treatment relative to higher-cost miners. Power cost below $30/MWh in Texas is a structural competitive advantage that took years to build through grid relationships, grid stabilisation contracts (which generate significant power credits), and physical proximity to low-cost generation.

The Starboard Thesis: $9 Billion to $21 Billion in AI Value

In February 2026, Starboard Value — Riot’s fourth-largest shareholder with approximately 12.7 million shares — published a letter to CEO Jason Les and Executive Chairman Benjamin Yi that became the most-read document in Riot’s investor community.

The core argument: Riot’s Texas data center sites are among the most valuable undeveloped data center assets in the United States, primarily because they have the one thing that can’t be easily built right now — power connectivity at scale.

Corsicana has 400 MW currently built toward an ultimate 1 GW target. Rockdale has 700 MW of developed capacity, with the 200-acre acquisition in January 2026 enabling significant further expansion. Kentucky targets 232 MW. In total, Riot has approximately 1.4 gigawatts of additional capacity available to monetize beyond current mining and the initial AMD lease.

Starboard’s math: if Riot can monetize its power capacity at rates consistent with recent AI data center transactions, it could generate more than $1.6 billion in annual EBITDA from the data center business alone — before attributing any value to its Bitcoin mining or BTC treasury holdings. At a 10–13x EBITDA multiple, that implies data center equity value of $16–$21 billion, dwarfing the current $6.3 billion total market cap.

Starboard specifically named the AMD agreement as “early validation” and acknowledged Riot’s progress — but described the AMD deal as “only a starting point.” The letter urged Riot to act “with urgency” on additional tenant signings, warning that rivals (specifically Core Scientific) are moving faster on comparable pivots.

The stock jumped 9% on the day the Starboard letter was published.

The broader AI and data center buildout that is reshaping the value of energy infrastructure makes the strategic logic of Starboard’s argument compelling. Power-connected land in Texas is genuinely scarce. New grid interconnections take 3–5 years to approve and build. Riot’s existing infrastructure has a first-mover advantage that can’t be replicated quickly.

The AMD Deal: Concrete Numbers Behind the Thesis

The first confirmed data center lease — signed January 2026 with Advanced Micro Devices — gives Riot’s AI pivot a credibility anchor that purely aspirational positioning wouldn’t have.

Terms of the AMD deal:

• 10-year initial lease at Riot’s Rockdale facility
• Initial capacity: 25 MW of critical IT load
• Option to expand: up to 175 MW more (total 200 MW)
• Estimated contract revenue: ~$311 million across the initial term
• Riot acquired 200 acres of land underlying the Rockdale site (fee simple acquisition) to enable the data center buildout supporting AMD’s requirements
• Data center revenue from the AMD agreement began generating in January 2026

The AMD deal is significant for three reasons. First, AMD is a credible, financially strong tenant with genuine long-duration AI compute demand. This is not a startup counterparty. Second, the 200 MW option represents 8x the initial capacity, giving Riot significant revenue upside if AMD exercises even a portion of the expansion rights. Third, the deal demonstrates Riot’s ability to execute a complex infrastructure transaction — land acquisition, regulatory navigation, infrastructure build-out, and long-term lease negotiation — which is exactly the capability skeptics questioned when the AI pivot was first announced.

The gap between the current AMD deal (25 MW, $311M over 10 years) and Starboard’s $1.6B EBITDA vision is substantial — approximately 56 additional AMD-scale deals to fill 1.4 GW of available capacity. That’s why analysts who believe the thesis still give RIOT a Strong Buy while not taking $42 price targets at face value.

RIOT Key Data (April 2026)

Sources: GlobeNewswire (Riot Platforms FY2025 results); Riot Platforms IR; Yahoo Finance — RIOT; TipRanks; StockAnalysis.com

Why RIOT Has Underperformed Bitcoin

This point is worth making directly before the prediction tables, because many investors come to RIOT expecting it to track BTC closely and are surprised when it doesn’t.

Over the past 12 months, Bitcoin itself is up substantially from its cycle lows. RIOT’s market cap has increased 163% over the same period — so it’s not languishing in absolute terms. But Bitcoin miners structurally have a problem that becomes more acute over time: they’re leveraged to Bitcoin’s price on the upside, but they have fixed costs (power, hardware, salaries, D&A) that compress margins when mining difficulty rises. The 2024 halving cut the block reward from 6.25 to 3.125 BTC. The same amount of global computing power is now competing for half the bitcoin. That’s why Riot’s average mining cost per BTC jumped 54% in FY2025 even though BTC prices were significantly higher than in FY2024.

The only sustainable escape from this trap is either: a) Bitcoin’s price rises faster than mining difficulty (possible in bull cycles, uncertain over time), or b) the company diversifies revenue away from pure mining into data center infrastructure that isn’t dependent on BTC price at all. That’s precisely what the AMD deal and Starboard thesis are about.

Bitcoin’s price trajectory remains the single largest variable in RIOT’s near-term stock price. If BTC moves from current levels toward $120,000–$150,000, miner economics improve dramatically even without data center contributions. If BTC declines toward $60,000–$70,000, mining cost-per-coin approaches or exceeds revenue-per-coin for all but the most efficient operators — and Riot, with its 3.0c/kWh power advantage, would be among the last standing, but still loss-generating.

RIOT Stock Forecast 2026

April 30, 2026 — Q1 2026 earnings — is the most immediate catalyst for RIOT stock.

Consensus estimates expect Q1 revenue of approximately $127.91 million (range: $105.45M–$141.60M) and an EPS of approximately -$0.55 (range: -$1.98 to -$0.04). The wide ranges reflect genuine uncertainty: the Q4 2025 EPS miss of -$1.88 against a -$0.21 estimate (-780% surprise) shows how imprecise current mining company forecasting is.

What the Q1 call needs to deliver for bulls:

1. Data center revenue disclosure: How much has the AMD lease generated since January? This is the first quarter where any data center revenue will appear. Even modest numbers — a few million dollars — validate the transition narrative.
2. Additional tenant progress: Any announced letter of intent or signed agreement with a second data center tenant beyond AMD would be an immediate re-rating catalyst. Analysts are watching for signals of deal pipeline.
3. BTC treasury update: The Q1 BTC sale volume (3,778 BTC) raised legitimate questions about the pace of treasury liquidation. The call needs to frame this as strategic, not distressed.
4. Power cost hold: If 3.0c/kWh is sustainable into Q2, mining economics remain among the best in the industry.

For the full-year 2026 stock view, the two scenarios are relatively distinct:

Bull case for 2026: AMD lease revenue visible in Q1, a second major data center tenant signed by mid-year, BTC price holding above $85,000–$90,000, and the stock re-rating from “expensive miner” toward “infrastructure company at a discount.” In this scenario, RIOT reaches $28–$38 by year-end as the Starboard thesis gains credibility with actual revenue.

Bear case for 2026: Q1 data center revenue disappoints (too small to move the narrative), BTC declines below $75,000 under macro pressure, no additional data center tenant announcements, and the stock gives back recent gains to retest the low-$12 to low-$14 range. The $663 million net loss and the Q4 EPS miss both weigh on sentiment in this scenario.

Piper Sandler’s April 21, 2026 target raise from $21 to $23 is consistent with the “base to moderate bull” range — cautious optimism that the business is directionally correct without pricing in full execution on the AI thesis yet.

RIOT Stock Forecast 2027–2030

The 2030 investment thesis for RIOT is cleaner than almost any other Bitcoin miner — precisely because Riot has physical assets that don’t require BTC to be valuable.

1.7 gigawatts of power-connected land in Texas is a real asset whose value is determined by the data center market, not the crypto market. If AI compute demand continues growing — and there’s little credible reason to expect it to reverse — the scarcity of grid-connected industrial land in Texas ensures Riot’s infrastructure appreciates in real terms. This is not a financial bet on crypto; it’s a real estate and energy bet with crypto optionality.

The stablecoin and DeFi infrastructure build-out of 2026 intersects here too: as on-chain financial infrastructure demands more compute for settlement, compliance, and AI-driven risk assessment, the physical infrastructure that Riot is building hosts some of that demand. Coinbase’s own institutional infrastructure expansion illustrates how crypto-native companies are increasingly competing for the same data center capacity as traditional hyperscalers.

For 2030, the key questions:

Has Riot successfully leased a significant fraction of its 1.7 GW available capacity? Even 30% of that — 500 MW — at competitive AI data center rates would generate EBITDA that would dwarf current mining revenue. At $10M/MW annually (a reasonable mid-range data center lease rate), 500 MW would produce $5 billion in annual revenue.

Has BTC gone through another halving cycle? The April 2028 halving will cut rewards from 3.125 to 1.5625 BTC per block. Miners who haven’t diversified revenue by 2028 will face another profitability squeeze. Riot’s data center pivot is essentially a hedge against the 2028 halving’s impact on mining economics.

What is BTC’s price by 2030? This remains the wildcard. Bitcoin’s supply schedule — with only 21 million total coins ever, and diminishing daily issuance — argues for higher prices over time as institutional adoption grows. Bitcoin’s long-term price trajectory intersects directly with Riot’s mining margin and BTC treasury value — currently 15,680 BTC worth approximately $1.1 billion.

The extreme 2030 scenario — RIOT trading above $100 — implies a market cap north of $50 billion, which would require both full data center monetisation and continued Bitcoin appreciation. It’s within the range of theoretical outcomes if Starboard’s thesis executes. It’s not a base case.

Is RIOT Worth Buying in 2026?

Riot is genuinely one of the most interesting stocks at the intersection of Bitcoin mining, energy infrastructure, and AI data centers. No other position in the public equity market gives you all three exposures simultaneously with as much operational scale.

The bull case is real: 3.0c/kWh power cost, 42.5 EH/s deployed hashrate, 15,680 BTC treasury, a working AMD data center lease, and 1.7 GW of available infrastructure that Starboard values at up to $21 billion — against a current market cap of $6.3 billion.

The bear case is also real: $663 million net loss in FY2025, a Q4 2025 EPS miss of -$1.88 against a -$0.21 estimate (a -780% surprise), treasury being drawn down to fund infrastructure, and an AI data center transition that is still in its first operational quarter with a single anchor tenant.

The April 30 earnings call is the most important near-term event. If data center revenue is visible and any progress on additional tenants is announced, the re-rating thesis gets legs. If it’s another large net loss with no new tenant announcements, the bear case gains force.

For investors with a 2–4 year horizon and comfort with Bitcoin volatility, RIOT at $18 is a bet on one of the most capital-intensive transitions in the mining sector. The infrastructure exists. The power contracts exist. The AMD relationship exists. What’s still unproven is Riot’s speed of execution on signing additional data center tenants — which is what April 30 needs to start addressing.