How Riot Monetized $198M in Bitcoin to Fund a 112MW AI Data Center: Lessons for C-Suites

How Riot Monetized Bitcoin to Finance an AI Data Center — Lessons for C‑Suites

Riot Platforms converted mined Bitcoin into near‑term capital to kickstart an AI data‑center build. The company sold about 2,201 BTC across November–December 2025, generating roughly $198.6 million, and finished the year holding 18,005 BTC. VanEck’s Matthew Sigel argued those proceeds are sufficient to fund Riot’s initial 112 MW Corsicana phase and jump‑start its 2026–27 AI build‑out.

Quick facts you can scan

  • BTC sold: 383 BTC in Nov 2025 (~$37M) and 1,818 BTC in Dec 2025 (~$161.6M); total ≈ $198.6M.
  • Year‑end BTC balance: 18,005 BTC.
  • Corsicana target: initial 112 MW phase aimed for Q1 2027 (company target / analyst coverage).
  • Production: rose from 428 BTC in Nov (14.3/day) to 460 BTC in Dec (14.8/day) — +8% MoM; down ~11% vs Dec 2024.
  • Hashrate: deployed hashrate rose 5% MoM to 38.5 EH/s (total computing power miners run to secure Bitcoin); operating hashrate ~34.9 EH/s (+27% YoY).
  • Fleet efficiency: 20.2 J/TH (joules per terahash — an energy‑efficiency metric for mining equipment).
  • Power credits: total jumped to $6.2M in Dec from $2.3M in Nov; energy credits and demand‑response payments are boosting cash flow.
  • All‑in power cost: fell to 3.9¢/kWh (down 1% MoM).

How the financing works — plain language

Riot mines Bitcoin as a recurring output. Instead of issuing equity or taking on large debt, Riot sells a portion of mined BTC to create cash for capital expenditure (capex — money spent to build the data center). CEO Jason Les says Riot will “sell monthly production to fund AI growth,” explicitly trading potential future BTC upside for immediate, non‑dilutive funding.

Matthew Sigel (VanEck, paraphrase): One season of selling mined BTC can fund Riot’s initial AI data‑center phase — showing miners are converting crypto inventory into capital for AI expansion.

This is equivalent to a manufacturer selling part of its inventory to finance a factory build rather than issuing new shares. It’s pragmatic: cash in hand speeds projects, preserves ownership, and avoids costly equity dilution — at the cost of exposure to BTC price timing and investor perception.

Worked example: the simple math

Proceeds from Nov–Dec 2025 sales ≈ $198.6M. VanEck’s analysis suggests that amount can cover the initial capex for the 112 MW Corsicana phase (site work, electrical infrastructure and major equipment to host AI racks). That’s the core thesis: monetize a produced commodity now to fund an asset that generates revenue tomorrow.

Sensitivity: timing risk in two scenarios

  • If BTC rallies 30% after the sales:

    The $198.6M in proceeds would represent an opportunity cost of roughly $59.6M (30% of proceeds) compared to holding — a real lost upside.

  • If BTC falls 30% after the sales:

    Riot avoided a notional loss of about $59.6M by selling into higher prices — the sales lock in value and remove downside risk for that portion of inventory.

These are illustrative numbers, but they show the tradeoff: selling secures capital and limits downside, while holding preserves upside at the cost of funding uncertainty.

Operational tailwinds that made this possible

  • Higher deployed hashrate: increased capacity gives Riot more predictable monthly BTC production to monetize.
  • Improved power economics: power credits (payments for reducing grid demand or using flexible load) surged to $6.2M in December, and all‑in power costs ticked down to 3.9¢/kWh — both enhance free cash flow.
  • Fleet efficiency: 20.2 J/TH reduces marginal cost per BTC mined, improving the margin on each sold coin.

Strategic implications for C‑suites

Using an operationally produced commodity to fund AI infrastructure is a real, repeatable financing option — but it’s not universal. It fits when an organization has:

  • Predictable recurring output that can be monetized.
  • Strong unit economics (low power costs, efficiency gains, grid programs).
  • A high‑confidence capex plan with measurable return on invested capital (ROIC).

Compared with alternatives:

  • Equity raises: preserve treasury but dilute existing shareholders.
  • Debt: leverages balance sheet and requires servicing regardless of project success.
  • Pre‑sales/colocation contracts or sale‑leaseback: transfer some construction risk and can provide upfront cash but may be complex to negotiate.

Risks and mitigation

  • Price timing and opportunity cost: mitigate with partial hedges (forwards, collars, options) and staged sale plans tied to project milestones.
  • Investor perception: selling treasury BTC can be read as a lack of confidence in long‑term price — offset with clear capital allocation messaging and disclosure.
  • Reliance on power credits: grid programs can be episodic or regional; plan for base case economics without extraordinary credits.
  • Regulatory and tax treatment: selling mined crypto has accounting and tax implications; coordinate early with tax and treasury teams.

What boards should ask before copying the playbook

  • Do we have stable, forecastable production?

    If production is volatile, monetizing output creates funding uncertainty and could force unplanned sales at worse prices.

  • Is our core unit economics robust without power credits?

    Test a downside case where credits are lower or absent — can the project still meet return hurdles?

  • Have we defined a hedging and governance policy?

    Decide how much to sell, when, and who signs off. Require scenario modeling (±30% BTC) and stress tests.

  • What will investors interpret the sales to mean?

    Prepare investor communications: explain the IRR on the AI asset versus expected BTC appreciation to justify the trade.

Key takeaways

  • Riot turned ~2,201 BTC into roughly $198.6M to fund an initial 112 MW AI data‑center phase. That’s a practical example of monetizing recurring digital output to finance AI infrastructure.
  • The model preserves equity and accelerates capex deployment — at the expense of BTC upside and some market‑timing risk.
  • Strong power economics and fleet efficiency made the math work; without those tailwinds, the approach is riskier.

If your board is evaluating similar options — monetizing an asset (digital or otherwise) to finance AI infrastructure — start with a 90‑minute scenario workshop: map capex needs, run BTC (or asset) sensitivities, and design a hedging and governance framework. This approach isn’t a silver bullet, but when the conditions align it converts recurring operational output into a disciplined pathway to AI capacity.

12