Reserve Risk

Reserve Risk plots one ratio: spot price over the accumulated HODL bank, on a logarithmic right axis, with spot price muted on a log left axis for orientation. The sage band at or below 0.002 is the strong-conviction zone — where price is cheap relative to the conviction long-term holders have already banked. The rust band above 0.008 is the conviction-exhaustion zone, where those holders have historically begun selling into strength. Horizontal reference lines mark both thresholds.

Today’s reading is 0.00113, placing the network in the Strong conviction regime, with an accumulated HODL bank of 58,047,216. The point of the chart is the ratio’s asymmetry across cycles, which the data sources page details — the floor has held while the ceiling has collapsed.

The headline formula is short:

Reserve Risk = Price / HODL Bank

The HODL bank is the interesting part. It accumulates “conviction” from Coin Days Destroyed — the chain-level primitive that captures how many coin-days a spend extinguishes (a 1 BTC coin held for 100 days then spent destroys 100 BTC-days). Hauge weights each spend by the spot price at the time of the spend versus a reference price drawn from the running historical median, building up an integral of conviction shown by long-term holders sitting through depressed prices.

The intuition. When old coins move at low prices, they add little to the HODL bank. When they sit through long stretches of pain, the bank grows. When they finally move at elevated prices, the bank only barely shrinks — the conviction-bank is a near-monotonic accumulator. Spot price divided by that accumulator is therefore a direct expression of “how much are we paying for this stack of conviction right now”. That monotonicity is exactly why the ratio’s ceiling compresses cycle over cycle, the effect documented below.

Coin Days Destroyed itself is an early-Bitcoin primitive — the metric was introduced as “Bitcoin Days Destroyed” on the Bitcoin wiki by user ByteCoin in 2011. Reserve Risk is one of several on-chain valuation indicators Hauge built on top of that primitive.

Reserve Risk is most informative at the extremes — but the two extremes are no longer equal. Sustained readings at or below 0.002 have fired at every cycle bottom on record and continue to do so; that floor is the trustworthy half of the indicator. Sustained readings above 0.008 fired only at the 2013, 2017 and April-2021 peaks and have missed every top since. The middle range — between 0.002 and 0.008 — covers most days and carries weak signal on its own; pair the read with cycle-shape and cohort-spread indicators (RHODL, MVRV, STH/LTH realized price) for context.

Reading every canonical cycle anchor against the live series surfaces the same regime-decay pattern that drives the rest of the realized-cap family — but in Reserve Risk the compression is more dramatic than in any other indicator on the site. Cycle peaks at 0.037, 0.039, 0.032, 0.0094, 0.0055, 0.0025. Cycle troughs at 0.00104, 0.00159, 0.00120, 0.00079.

The cleanest way to see the regime shift on this chart is the per-cycle Reserve Risk peak, in order. Pulling the maximum reading inside each cycle’s topping window from the live series:

Four of six cycle peaks on record cleared 0.008 — and three of those four were pre-2018. The November 2013 peak printed the highest Reserve Risk in the modern series at 0.0387; April 2013 hit 0.0373; December 2017 hit 0.0323. The April 2021 peak just cleared the threshold at 0.0094, but the November 2021 peak fell back to 0.0055 and the March 2024 pre-halving high topped at 0.0025 — well inside Neutral territory. The 2024 peak is roughly an order of magnitude lower than the 2013/2017 peaks.

The denominator is doing the work. The HODL bank is a near-monotonic accumulator — every additional bear-market base coin adds conviction, and that conviction never fully evaporates even when older coins eventually move. By 2024 the bank had been compounding for ~13 years of Bitcoin history, against a relatively young denominator-stock at the 2013 peak. Even very large dollar-denominated coin movements at the modern cycle tops barely dent the HODL bank, so the ratio (price / bank) compresses naturally.

Lost coins amplify the effect. Every Patoshi-era coin that has not moved since 2010 contributes maximum value-days-destroyed to the HODL bank but never spends, so it adds to the denominator forever. With the permanently-lost float running to several million BTC, those never-moved coins act as a standing floor under the HODL bank, structurally lowering every Reserve Risk print.

A practical corollary: a “wait for Reserve Risk > 0.008” rule would have missed the November 2021 and March 2024 peaks consecutively. The bottom-side threshold (≤ 0.002) has held up far better — every cycle bottom on the record has fired sub-0.002, with troughs clustered tightly between 0.00079 and 0.00159.

The 0.008 ceiling has compressed by an order of magnitude. Three of six cycle peaks cleared it cleanly (2013×2 at ~0.038, 2017 at 0.032). One barely cleared (2021-Apr at 0.0094). Two did not (2021-Nov at 0.0055, 2024-Mar at 0.0025). The threshold was anchored on the 2010s blow-off-top shape and is now well outside the modern operating range. A “wait for 0.008” rule would have missed the entire November 2021 cycle and would currently be missing the 2024 cycle as well — two consecutive top-side misses.

The HODL bank is invisible to off-chain custody. Coins held inside ETF wrappers, lending platforms, or exchange custody appear as on-chain HODL only when the underlying coin sits unmoved at a custodian address. ETF creation and redemption can churn the same underlying position multiple times per week without any change in the “real” conviction of the end holder. The 2024 reading is biased downward by post-Jan-2024 spot-ETF flows that age-reset coins via custodial reshuffling — a structural break the metric has no way to see.

Early-cycle data is unreliable. Reserve Risk’s denominator was small enough through 2012 that the ratio took on values not directly comparable to modern readings — the very first series point, 17 August 2010, prints above 1.0 because the HODL bank was essentially empty. The chart’s display window starts in 2013 once the bank had accumulated enough mass to make the denominator stable. We keep the underlying series back to 2010 for completeness, but treat any apparent extremum in the first two years as a denominator artefact, not a signal.

2020 Covid was a regime exception in both directions. The Mar 2020 trough at 0.00120 was clean (Strong-conviction band). But spot price recovered fast enough that there was no clean “peak” window in the 2020 cycle — the run from $5,000 to $69,000 happened in about 20 months without a clear pause. Reserve Risk printed 0.0094 on 13 March 2021, then 0.0055 on 20 October 2021, with a not-very-elevated mid-cycle in between. The signal was directionally correct but the cycle’s shape was unusual.

If you accumulate on a schedule, Reserve Risk at or below 0.002 is one of the highest-conviction tactical accumulation signals in the realized-cap family. Every cycle bottom on the record has fired here — and unlike the raw MVRV sub-1 trigger, the deep-bottom Reserve Risk band has not compressed cycle over cycle. The 2022 post-FTX trough at 0.00079 was the lowest reading on the entire series, which suggests bottom-side conviction signals remain intact even as top-side ones have dulled. This is the half of the indicator you can still lean on.

If you are timing the cycle top, the 0.008 ceiling is no longer a working signal — two of three modern peaks have fallen short, and the trend is toward further compression. Do not wait for it. Pair Reserve Risk’s bottom-side read with RHODL for the realized-HODL distribution lens, MVRV for the price-to-cost-basis ratio, and SOPR for the realised-profit cohort spread. Reserve Risk alone is a slow signal — it moves on the order of weeks, not days.