STH / LTH Realized Price

The chart plots three lines on a shared logarithmic price axis: spot Bitcoin, the average cost basis of short-term holders (coins last moved within 155 days), and the average cost basis of long-term holders (coins older than 155 days). The readable feature is the shaded band between the two cost bases — its width is the gap between what recent buyers paid and what patient holders paid, and its orientation (which line sits on top) flips at the turn of every cycle.

Today the band runs from an STH cost basis of $72,900.06 to an LTH cost basis of $49,249.49, with spot at $65,837.03 — placing the network Between cost bases. The two cohort lines are nightly aggregates; only spot moves between refits.

Both lines are slices of realized capitalisation — the framework Nic Carter and Antoine Le Calvez introduced at the Baltic Honeybadger 2018 conference. Realized cap values each unspent transaction output (UTXO) at the price it last moved on-chain rather than at today’s market price; per coin, that yields realized price.

The 155-day cohort split traces to Rafael Schultze-Kraft and Kilian Heeg’s 2020 paper Quantifying Short-Term and Long-Term Holder Bitcoin Supply, which identified 155 days as the inflection point at which a coin’s probability of being spent flattens out. Their methodology applies a logistic transition (midpoint 155 days, 10-day width) rather than a hard cutoff so the cohort boundary breathes smoothly through time.

The two series compute as:

STH realized price = Σ (last_spent_price·value) over UTXOs aged < 155d ÷ STH supply
LTH realized price = Σ (last_spent_price·value) over UTXOs aged ≥ 155d ÷ LTH supply

Both series are taken from the upstream as already-aggregated cohort lines; per-UTXO last-spent prices are not reconstructed from the raw blockchain. The data sources page documents the upstream feed and the methodology page records the smoothing convention and known divergences.

Spot can sit in one of three places relative to the cohort band, and each carries distinct history. Above both cost bases is the bull-regime default — every cohort in aggregate profit, with the gap between spot and LTH a coarse measure of cycle maturity. Between the cost bases means recent buyers are underwater while long-term holders remain in profit, the transitional shape typical of late corrections and early bear markets. Below both means even long-term holders are at a paper loss — the regime that has bracketed the 2015, 2018, and 2022 cycle bottoms on the daily-close record (and the 2011–12 crash before them). The 2020 Covid flush is the only modern bottom that did not reach it, and it missed by about five percent.

Reading every canonical cycle anchor against today’s cohort series surfaces the regime pattern. The 2015 cycle low printed deepest at −42.9% below LTH; the 2018 trough at −27.5%; the 2022 post-FTX trough at −25.2%. The 2020 Covid sell-off is the sole exception on this dataset — the closest daily close to an LTH breach (17 March 2020) bottomed +5.3% above LTH cost basis and then recovered. Three of the four bottoms since 2015 confirm the regime; the 2020 flush is the warning that even a cleanly defined signal misses.

The single most informative read on this chart is not either line by itself — it is the spread between them. STH cost basis tracks recent flow; LTH cost basis tracks the long memory of the network. The spread’s sign and magnitude maps onto cycle phase with the kind of regularity the rest of the on-chain stack rarely manages.

Spread strongly positive (STH well above LTH). The bull regime. New capital is paying premiums over the older cohort’s cost basis. Every prior cycle peak has fired with the spread maximally extended.

Spread compressing toward zero. The late-bull or early-bear shape. New capital has slowed; long-term holders’ average cost is catching up to recent buyers’.

Spread negative (STH below LTH). The bear-market signature. Long-term holders’ average cost basis exceeds recent buyers’ because LTHs include coins picked up near the prior cycle top. Every Bitcoin bear since 2015 has ended only after the spread closed and crossed back, with the cross dates landing in early 2015, late 2018, and mid-2022.

Today’s spread is +48.0%. That places the network in the fully extended bull phase under this lens — cross-check against STH/LTH SOPR, which carries the same cohort logic on spent outputs rather than aggregate cost basis.

Across 5,777 daily closes in the cohort series, spot has closed below LTH cost basis on 674 of them — a residency of 11.7%. The figure is the bear-market anatomy of the asset in one number: long-term holders are, on average, in unrealised profit on the great majority of all trading days. The residency in below-LTH territory falls entirely inside the historic bear markets — the 2011–12 crash, mid-2014 to mid-2015, late 2018, and mid-to-late 2022 — with no sustained occupation outside a confirmed bear. That scarcity is exactly why a below-LTH print reads as high-conviction.

The 2020 Covid bottom never breached LTH. The closest daily close on btc oak’s series was 17 March 2020, with spot at $5,032.50 against an LTH realized price of $4,777.30 — a margin of +5.34%. A mechanical “below LTH = bottom” rule would have skipped that generational entry entirely. The signal is high-conviction when it fires; the absence of the signal does not preclude a cycle-trough print, and 2020 is the standing counterexample.

Lost and dormant coins haunt the LTH line. Chainalysis research has estimated 2.78 to 3.79 million BTC permanently lost; Sergio Demian Lerner’s Patoshi research identifies roughly 1.1 million Satoshi-era coins that have never moved since 2010 (recap on Yahoo Finance). Both categories sit in the LTH bucket with near-zero realized prices, biasing LTH cost basis down and LTH supply up; in particularly stressed bottoms this can make the LTH line look more “held” than the active float justifies.

In strong bull runs, the LTH line rises because of the calendar, not demand. The 155-day boundary is statistical, not economic. Coins purchased near a cycle peak age past the boundary into LTH continuously, lifting LTH realized price even when no holder has bought. Through 2021 and again into 2024 this mechanical drift accounted for a meaningful fraction of LTH cost-basis growth — the indicator was not capturing fresh long-term demand, it was capturing time passing.

Custodial reset events can briefly mis-state the band. When a custodian consolidates UTXOs, every affected coin is repriced to the consolidation-day value even though no economic ownership has changed. The aggregate effect is small relative to the float, but in pathological cases — the Coinbase cold-storage migration of December 2018 is the well-documented one, see Felipe at Paradigma Capital’s deep-dive — cohort cost bases can briefly drift from the network’s true position.

If you accumulate on a schedule, the LTH cost basis is the cleanest single reference for “is the patient cohort still net up?” A spot close below it is a strong — not certain — generational-bottom marker, and the residency figure above shows how rarely it appears. Treat below-LTH days as a tactical accumulation accelerator, not a forecast, and remember 2020: a real bottom can stop just short of the line.

If you are timing the cycle, the STH/LTH spread is a slow indicator — it moves on the order of weeks, not days. It will not catch sharp local tops or flushes, but it brackets cycle phase well, and the STH-crossing-back-above-LTH event has marked the end of every bear since 2015. Pair it with faster-moving cohort signals like STH/LTH SOPR and MVRV to call the turn the spread only confirms.