Below fair value, coins leaving exchanges

What this means

The average holder is sitting at or below their own cost basis, and BTC is moving off exchanges on net rather than onto them. Historically this combination has shown up during and just after forced-selling events, not during ordinary pullbacks.

The v1 rule

Valuation reads "below-basis" when MVRV is under 1.0, or MVRV's percentile against its own full history is at or below the 25th percentile - whichever condition is met first. Flow direction reads "accumulating" when the net exchange flow (30-day) and exchange reserve change (90-day) signals combine, under v1's two-vote system, to a net negative reading: more BTC leaving exchanges than arriving. This is a pinned v1 classifier rule, not a permanent definition - a future version could set these thresholds differently.

What would change the state

Valuation would need MVRV's percentile to climb back above the 25th percentile (and MVRV itself above 1.0) to leave below-basis for mid-range. Separately, the flow and reserve signals would need to flip to a net positive reading - more BTC arriving at exchanges than leaving - for the direction label to flip to distributing. The two dimensions move independently; either can change while the other holds.

Historical duration and transitions

Entered 9 times in the comparison window.

Closed periods measured (N): 9
Mean duration: 30.4 days
Median duration: 29.0 days
Shortest / longest: 5 / 68 days

Historically transitioned into

Below fair value, coins moving to exchanges: 6 times
Mid-range valuation, coins leaving exchanges: 3 times

Comparisons use the institutional-era market from August 2020 onward; a metric's own percentile-in-history uses full available history.

What followed, historically

Of 9 prior periods in this state, what followed

30 days latermedian +4.8% · worst −27.8%

90 days latermedian −5.6% · worst −37.7%

180 days latermedian +21.8% · worst −43.9%

Bar: middle half of past outcomes. Tick: median. Line: worst case.

Table view

Horizon	Median	Q1 / Q3	Worst case
30d	+4.8%	−7.0% / +7.5%	−27.8%
90d	−5.6%	−13.1% / +22.8%	−37.7%
180d	+21.8%	−3.1% / +35.1%	−43.9%

Every prior period (9)

Comparisons use the institutional-era market from August 2020 onward; a metric's own percentile-in-history uses full available history.

Conditioned on: core state (valuation, flows, leverage, halving position); sentiment and short-history modifiers excluded.

Description of the past, not a forecast.

Every period in this state (9)

Period	30d	90d	180d
2026-06-21 – 2026-08-19	+4.8%	+28.2%	n/a - not yet resolved
2026-03-26 – 2026-04-07	+13.0%	-11.4%	+25.4%
2026-03-01 – 2026-03-11	+3.8%	+12.2%	+18.2%
2023-01-20 – 2023-02-10	+7.5%	+24.7%	+32.0%
2022-11-08 – 2023-01-15	-7.0%	+22.8%	+55.3%
2022-10-07 – 2022-11-05	+7.2%	-13.9%	+44.3%
2022-08-05 – 2022-09-04	-14.1%	-13.1%	+2.1%
2022-06-20 – 2022-07-30	+13.2%	-5.6%	-18.5%
2022-06-08 – 2022-06-13	-27.8%	-37.7%	-43.9%

Every period is listed here, including the ones where the base rate didn't hold.