# Bitcoin Hashpoint Theory Lab

> 21 Bitcoin valuation models side by side, as of the 2026-10-24 daily close. Model outputs, not forecasts by Bitcoin Hashpoint. Not investment advice.

Page: https://dashboard.bitcoinhashpoint.com/lab · JSON: https://dashboard.bitcoinhashpoint.com/api/v1/lab/models · OpenAPI: https://dashboard.bitcoinhashpoint.com/api/v1/openapi.json
Generated 2026-10-25 08:35 UTC. Source: Bitcoin Hashpoint Theory Lab, https://dashboard.bitcoinhashpoint.com/lab

## Consensus

- Family-weighted median at the 2026-10-24 close: **$100,828**, 25th to 75th percentile $90,023 to $138,748 (5 families, one vote each).
- BTC close $80,772: -19.9% against the median.
- Family values: Trend fits $138,748 (3); Cycle $90,023 (1); Adoption $80,770 (1); Price structure $100,828 (2); Mining cost $176,652 (1).

| Horizon | Date | Median | 25th to 75th | Families |
|---|---|---|---|---|
| +30 days | 2026-11-23 | $89,586 | $82,921 to $107,275 | 4 |
| +180 days | 2027-04-22 | $111,706 | $94,898 to $133,531 | 4 |
| +365 days | 2027-10-24 | $128,206 | $125,379 to $158,175 | 3 |

Past the close only forward models vote; history-only models stop at the close.

## Models

| Model | Family | Value at close | Status | Note |
|---|---|---|---|---|
| Power law (`powerLaw`) | Trend fits | $157,534 | voting |  |
| Capital S-curve (`scurve`) | Adoption | $80,770 | voting |  |
| Production cost (`cost`) | Mining cost | $176,652 | voting |  |
| Log regression / Rainbow (`rainbow`) | Trend fits | $138,748 | voting |  |
| Logarithmic growth curves (`growth`) | Trend fits | $132,679 | voting |  |
| Halving-cycle overlay (`halving`) | Cycle | $90,023 | voting |  |
| Stock-to-flow (`s2f`) | Scarcity | $845,347 | excluded | 365-day walk-forward error 112.8% is over the 100% limit |
| Gold parity / wealth share (`wealth`) | Scenario | $80,772 | excluded | a slider scenario, out of the median at the default settings |
| Mayer multiple (`mayer`) | Price structure | $81,890 | voting |  |
| 200-week moving average (`wma200`) | Price structure | $119,766 | voting |  |
| Pi Cycle top (`pi`) | Cycle | n/a | no_value | markers only: it marks dates, has no price level and never votes in the median Signal: {"kind": "pi_cycle_top", "last_cross": {"date": "2021-04-12", "price": 59831.66}, "crosses": 4} |
| Realized price / MVRV bands (`realized`) | Holder cost | n/a | no_value | its input series ends 2026-10-23, before the 2026-10-24 close; the Lab reads it only on days its series covers Last value $89,452 on 2026-10-23. |
| Thermocap multiple (`thermocap`) | Holder cost | n/a | no_value | its input series ends 2026-10-23, before the 2026-10-24 close; the Lab reads it only on days its series covers Last value $53,012 on 2026-10-23. |
| NVT (`nvt`) | Network | n/a | no_value | its input series ends 2026-10-23, before the 2026-10-24 close; the Lab reads it only on days its series covers Last value $44,977 on 2026-10-23. |
| Metcalfe value (`metcalfe`) | Network | n/a | no_value | its input series ends 2026-10-23, before the 2026-10-24 close; the Lab reads it only on days its series covers Last value $7,600 on 2026-10-23. |
| Difficulty regression (`diffreg`) | Mining cost | n/a | no_value | its input series ends 2026-10-23, before the 2026-10-24 close; the Lab reads it only on days its series covers Last value $47,585 on 2026-10-23. |
| Energy value (`energy`) | Mining cost | n/a | no_value | its input series ends 2026-10-23, before the 2026-10-24 close; the Lab reads it only on days its series covers Last value $184,968 on 2026-10-23. |
| Hash ribbons (`ribbons`) | Mining cost | n/a | no_value | markers only: it marks dates, has no price level and never votes in the median Signal: {"kind": "hash_ribbons", "state": "recovered", "last_capitulation": "2026-06-08", "last_recovery": "2026-08-26", "series_end": "2026-10-23"} |
| Options-implied distribution (`options`) | Market-implied | n/a | no_value | market-implied: its curve starts at the snapshot, after the close; it votes at the horizons |
| Futures curve (`futures`) | Market-implied | n/a | no_value | market-implied: its curve starts at the snapshot, after the close; it votes at the horizons |
| Polymarket price markets (`polymarket`) | Market-implied | n/a | no_value | market-implied: its curve starts at the snapshot, after the close; it votes at the horizons |

## Scorecard (walk-forward)

Walk-forward: at each month-end each model is rebuilt from data knowable on that date and scored on the close 30, 180 and 365 days later. History-only models carry their latest value forward. This is the honest test; in-sample errors flatter every model. Month-ends 2016-01-31 to 2026-08-31 (128), data through 2026-10-10. A model whose 365-day walk-forward error is over 100% stays drawn and scored but leaves the median.

| Model | 30d error | 180d error | 365d error | 365d within 25% |
|---|---|---|---|---|
| Power law | 45.6% | 47.1% | 48.2% | 28.2% |
| Capital S-curve | not walk-forward scored: the S-curve inputs are not rebuilt per date | | | |
| Production cost | 65.4% | 47.2% | 42.1% | 35.0% |
| Log regression / Rainbow | 45.6% | 47.1% | 48.2% | 28.2% |
| Logarithmic growth curves | 60.9% | 56.6% | 53.2% | 28.2% |
| Halving-cycle overlay | 38.8% | 41.1% | 40.3% | 32.6% |
| Stock-to-flow | 67.7% | 89.0% | 112.8% | 11.1% |
| Gold parity / wealth share | not scored: a slider scenario | | | |
| Mayer multiple | 22.0% | 39.9% | 59.5% | 14.5% |
| 200-week moving average | 31.8% | 41.7% | 44.8% | 29.1% |
| Pi Cycle top | markers only: no price level to score | | | |
| Realized price / MVRV bands | 23.6% | 39.0% | 50.6% | 26.5% |
| Thermocap multiple | 46.0% | 55.1% | 63.3% | 5.1% |
| NVT | 45.2% | 54.5% | 67.7% | 12.8% |
| Metcalfe value | 75.1% | 78.6% | 82.3% | 4.3% |
| Difficulty regression | 45.1% | 52.2% | 56.8% | 11.1% |
| Energy value | 96.7% | 70.2% | 57.9% | 30.8% |
| Hash ribbons | markers only: no price level to score | | | |
| Options-implied distribution | scoring begins once 30 days of snapshots exist (1 so far) | | | |
| Futures curve | not scored: the options median carries the family's score | | | |
| Polymarket price markets | not scored: the options median carries the family's score | | | |
| **Family median** | 27.6% | 31.6% | 41.2% | 25.6% |
| **Every-model median** | 31.2% | 31.7% | 41.1% | 30.8% |

## Method

Each model's central value at the latest daily close, from the Lab page's own code at its default settings. The consensus is the family-weighted median: each family's median, then the median and 25th to 75th percentile across families. Past the close only forward models vote. A model whose 365-day walk-forward error is over 100% stays drawn but leaves the median.

Inputs: Bitcoin Hashpoint Theory Lab: daily BTC closes (Bitstamp), Coin Metrics and blockchain.info on-chain series, the box's own Bitcoin node, Deribit and Polymarket snapshots, run through the Lab page's own model code.

Each model's formula and origin is in the JSON at https://dashboard.bitcoinhashpoint.com/api/v1/lab/models/{id}.

## Cite

Source: Bitcoin Hashpoint Theory Lab, https://dashboard.bitcoinhashpoint.com/lab

Model outputs, not forecasts by Bitcoin Hashpoint. Not investment advice.