Buy BTC vs Mine BTC

Deploy the same capital two ways — buy spot bitcoin, or buy ASICs and mine — and compare outcomes across thousands of jointly-simulated, path-dependent futures. Network hashrate responds to the price path with a deployment lag (hashprice mean-reverts to a declining marginal-cost equilibrium), halvings shock miner revenue, and machine resale value rides hashprice. Mining is evaluated against buying on the same price paths, so every probability below is a paired comparison.

Model & methodology

BTC price follows a fat-tailed (Student-t) geometric random walk. The drift you set is the median annual growth rate.

Network hashrate is endogenous. Growth = base deployment rate + κ × log(hashprice / equilibrium hashprice), applied with a ~90-day machine-deployment lag. The equilibrium anchor is the cash cost of the marginal fleet (~25 J/TH at ~6.5¢/kWh), declining as fleet efficiency improves. A price rally lifts hashprice → hashrate accelerates months later → hashprice compresses back. A halving instantly halves hashprice → growth stalls (capitulation) until it recovers. This feedback is what makes mining returns path-dependent: the same terminal BTC price can produce very different mining outcomes.

Mining leg: your budget buys hashrate at the machine price; daily production = fleet share of (subsidy + fees) × uptime, net of pool fee; costs = energy + opex. Machines curtail on days with negative gross margin. HODL keeps mined BTC (selling only enough to cover costs); Sell converts production daily. Rig resale value decays at your depreciation rate and co-moves with hashprice (beta), floored at scrap value.

Buy leg: budget / spot price, held. Mix (Part 3): every allocation w×mine + (1−w)×buy is computed exactly per path; Sharpe/Sortino use the cross-simulation CAGR distribution vs your risk-free rate.

Calibration anchors (mid-2026): hashprice ~$30/PH/day (all-time low), fees ~0.65% of subsidy, industry cash cost ~$80k/BTC, machine prices ~500-day payback. Historical drift/vol fitted from this site's own price & hashrate data.

Part 1 Your assumptions

Loading market data…

Capital & horizon

Budget USD

Horizon years

Simulations

Strategy

Risk-free rate %/yr

Mining rig & site

ASIC model

Machine price $/TH

Energy (all-in) ¢/kWh

Energy escalator %/yr

Opex % of revenue

Pool fee %

Uptime %

Depreciation %/yr

Attrition %/yr of machines die

Count resale rig residual counts as wealth

Curtail at loss off on negative-margin days

Market model

BTC drift %/yr default

Path shape

Same total growth, different route — mining lives on the path, not the endpoint.

BTC volatility %/yr default

Distribution

HR response

How hard hashrate chases profitability — the path-dependency engine.

HR base growth %/yr

Deployment pace at normal margins; blank = neutral. The response adds κ×(hashprice vs breakeven, lagged 90d) on top — see Model & methodology.

Fee load % of subsidy

The wealth race — same capital, two strategies

Median path with 5–95% cone. Dashed markers = halvings. Mining wealth is marked-to-market (treasury + rig resale value).

Who is ahead, when

Share of simulations where mining leads buying at each point in time (paired, same price path).

Where you end up — terminal wealth

Distribution of final value across simulations. Dashed line = initial budget.

Mining minus buying, path by path

Outperformance in % of budget. Purple area = paths where mining won.

Path explorer — mining wealth relative to buying

40 sampled simulations, mine÷buy wealth ratio (log scale). Above 1× = mining ahead. Color = final winner.

BTC price forecast

The simulated futures: median + 5–95% cone.

Network hashrate forecast

EH/s — endogenous: it chases each price path with the deployment lag, and stalls after the halving.

Hashprice

$/PH/day — price ÷ hashrate: the mining leg's revenue engine.

BTC stacked by mining

Net treasury (HODL) vs the coins your budget buys on day one.

Rig resale value

% of purchase price: obsolescence decay × hashprice beta.

Which rig? — machine shootout

Every model runs your budget through the same simulated futures, so gaps between machines are real, not noise. Bars: median outperformance vs buying (whiskers = 25–75%).

Racing machines on identical futures…

Part 3 The optimal mix — risk-adjusted blending

Your allocation

Drag to blend capital between the two strategies. Metrics recompute from exact per-path outcomes.

MINING 50% 50% SPOT BTC

Risk / return frontier

Each dot = an allocation, 0% → 100% mining. ★ = max Sharpe. Risk and return from the cross-simulation CAGR distribution.

Risk-adjusted return vs allocation

Sharpe and Sortino across the blend. The peak is the optimizer's answer.

Part 2 No assumptions needed — the scenario map

Instead of trusting one set of assumptions, this section sweeps them. Every cell re-runs the paired simulation with common random draws, so differences between cells are pure scenario effects — not Monte Carlo noise. Advantage = median (mining − buying) terminal wealth, in % of budget.

Sweeping 300+ scenarios…

The decision surface — BTC drift × energy cost

Rotate me. Purple = mining wins, gold = buying wins. The ridge where the surface crosses zero is your breakeven frontier.

BTC drift × hashrate growth

The other axis of fate: how fast the network deploys machines. With the response engine on, this axis is the baseline deployment pace — actual hashrate still bends toward each price path.

P(mining wins) heatmap

Probability mining beats buying, per scenario.

Breakeven energy frontier

Max power price where mining still matches buying, per BTC scenario. Mining is the defensive trade: cheap power matters most when BTC disappoints.

Machine price × hashrate growth

Advantage (pp of budget) at the base BTC scenario.

What moves the answer — tornado

Change in median advantage when each input swings low → high (one at a time, same random draws).

Does the route matter? — path shapes compared

Same total drift, five different routes, identical random draws. Mining lives on the path.

What the map says