Short-Term Holder NUPL

The paper profit or loss of recent buyers โ€” coins that changed hands within the last 180 days. Below zero this fast-moving group is underwater. Every point shown is real up to the date on the badge above the chart.

Take **only** the coins that moved within the last 180 days. Compare what they are worth now with what they were bought for, and express the gap as a share of today's value.

This group is the market's weakest hands, so their pain matters. For example, โˆ’0.10 means recent buyers are 10% underwater as a group โ€” historically that is where panic selling has clustered and short-term bottoms have formed.

Above about 0.3, recent buyers are sitting on large gains, which has usually come during the excited part of a rally. This flips between profit and loss far faster than the patient-owner version.

In words: for coins that moved within the last 180 days, take today's value minus the price they were bought at, then divide by today's value.

Every point on screen was counted by **our own Bitcoin node** โ€” a full copy of the blockchain that we run ourselves. The line starts in late August 2010: in Bitcoin's first priced days most coins in this group still dated from before any price existed, so the group's average cost was close to zero. Like our NUPL and MVRV charts, it starts once price against that cost had stayed at 10 or below for a full month. A day our node could not count is left as a gap; it is never filled in from another source.

Coins join and leave this group purely by ageing, so the reading changes even when nobody trades. Note on the age split: our node's history before 2026-08-17 was replayed with a 155-day split between short- and long-term holders; from 2026-08-17 it uses 180 days, so the line can shift slightly on that date.

Computed by Safu from its own Bitcoin node's short-term-holder realized price (coins that moved within the last 180 days), set against the daily BTC closing price we store. Our node's rows for this group run back to 2009; the line starts in late August 2010, once the group's average cost had settled into a range a market can produce. No outside data is used.

Computed in-house from Safu's Bitcoin node ยท 1 point/day since Aug 2010 ยท recomputed daily