How is the Estimated Production Cost Calculated?

This chart uses a Difficulty Regression Model to estimate the average cost of producing one Bitcoin. Rather than attempting to track exact electricity prices or hardware specs for every miner worldwide, this model uses the network's own difficulty adjustment as a proxy for total mining expenditure.

Breaking Down Each Component

• Network Difficulty (D) — A protocol-level number that adjusts every ~2 weeks to keep block times at 10 minutes. When more miners join (deploying more capital and energy), difficulty rises. It is the single best on-chain proxy for how much real-world resource is being spent to secure the network. We use a simple moving average (adjustable via "SMA window" in the toolbar) to smooth out short-term noise.
• Damping Exponent (0.41) — If difficulty doubles, the production cost does not double — it rises by only about 33%. This is because mining hardware efficiency improves over time (newer ASICs produce more hashes per watt). The 0.41 exponent captures this historical rate of hardware improvement without needing to track specific ASIC models.
• Daily Issuance — The number of new BTC mined per day (~144 blocks × block reward). Currently 450 BTC/day (3.125 BTC × 144). Unlike the difficulty SMA, issuance uses the current day's value — so when a halving occurs, the production cost instantly doubles (same energy, half the coins). This creates the sharp step-up visible in the chart at each halving event.
• Scale Factor (adjustable) — A calibration constant (defaulting to 49, traditionally 50 in standard models) derived by fitting the model to historical bear market bottoms (2015, 2018–19, 2022), when price approached the true marginal cost of production and inefficient miners were forced offline. This value can be adjusted dynamically using the Scale input in the toolbar.

Why Does This Work?

Bitcoin mining is a competitive market. Miners only deploy hardware and consume electricity if they expect to profit. When the price drops below the cost of production, the least efficient miners shut down → hashrate falls → difficulty adjusts downward → costs decrease for remaining miners. This creates a natural dynamic floor beneath the price. The difficulty regression captures this feedback loop purely through on-chain data.

Data Sources & Updates

• Difficulty — Fetched from blockchain.info (full history, updated daily)
• BTC Price — Fetched from bitcoin-data.com (daily close)
• Issuance — Computed from the Bitcoin halving schedule (hard-coded block reward dates)

Both data feeds are cached server-side for 12 hours and refresh automatically on every page load. The chart requires zero manual maintenance — difficulty changes, halvings, and price movements are all captured automatically.