MVRV Z-Score Bottom Signals: Historical Backtest
CryptoRebateHub Editorial Team
The MVRV Z-Score entered its green zone five times (2012, 2015, 2019, March 2020, 2022); each marked a cycle low area, but it is a zone, not a timing signal.
Direct answer: The MVRV Z-Score compares Bitcoin's market cap with its realized cap and scales the gap by the standard deviation of market cap. Per Glassnode's metric guide and CryptoSlate's review of Glassnode data (checked 2026-11-01), the score has dropped into its green zone in five episodes: 2012, 2015, early 2019, March 2020 and from June 2022 into early 2023, and each coincided with a cycle low area rather than a single day. Use it as a valuation zone that tells you how much downside has already been priced, not as a buy button; the zone has lasted anywhere from a few weeks to more than 300 days.
What the MVRV Z-Score measures
Three inputs build the metric, and all three come from on-chain data rather than exchange order books.
- Market cap is the spot price multiplied by the circulating supply.
- Realized cap values every coin at the price it last moved on-chain, so it approximates the aggregate cost basis of all holders.
- Standard deviation of market cap is computed over Bitcoin's full history and normalises the gap between the two caps.
Glassnode's metric guide (checked 2026-11-01) states the formula as MVRV Z-Score = (Market Cap − Realized Cap) ÷ Standard Deviation of Market Cap and credits the metric to the analyst Awe_andWonder in October 2018, building on the MVRV ratio work of Murad Mahmudov and David Puell. When the score is high the market trades far above the holders' aggregate cost basis, which has historically marked tops; when it is low or negative the market trades near or below that cost basis, which has historically marked bottoms. The MVRV Z-Score explainer on this site covers the mechanics in more depth, and the MVRV zones guide covers interpretation.
One point matters before any backtest: neither Glassnode's guide nor LookIntoBitcoin's chart page publishes a numeric cut-off for the green or red box (checked 2026-11-01). The boxes are drawn visually. If you adopt a rule such as "green means below zero", it is your convention, and you should write it down before you test it.
Worked example: reading the score
The score is a standard-deviation test, so a change of one point means the gap between the two caps moved by one standard deviation. The table uses hypothetical numbers to show how the three inputs combine.
| Scenario (hypothetical) | Market cap | Realized cap | Std. dev. of market cap | Z-Score | Reading |
|---|---|---|---|---|---|
| Market below cost basis | 500 bn | 600 bn | 400 bn | −0.25 | Green-zone territory: average holder underwater |
| Neutral | 800 bn | 600 bn | 400 bn | 0.50 | Mid-cycle, no extreme |
| Stretched | 2,600 bn | 600 bn | 400 bn | 5.00 | Approaching historical top territory |
Two features follow from the maths. First, a negative score requires market cap to fall below realized cap, which only happens after a deep drawdown, because realized cap itself drifts upward as coins change hands at higher prices. Second, the denominator grows as Bitcoin's history lengthens, so the same percentage gap produces a smaller Z-Score today than it did in 2015; comparing absolute peaks across cycles therefore understates how stretched recent cycles were.
The five green-zone episodes
CryptoSlate's research note on the Glassnode data set (checked 2026-11-01) lists how long each episode lasted. The durations, not the entry dates, are the part most readers miss.
| Episode | Approximate period | Time in green zone | What happened next |
|---|---|---|---|
| 1 | 2012 | about 212 days | The 2012 halving followed in November; the 2013 bull cycle came after |
| 2 | 2015 | more than 300 days | Longest green-zone stay on record; the recovery started late 2015 |
| 3 | Early 2019 | about 133 days | Market lifted out of the zone in spring 2019 |
| 4 | March 2020 | a few weeks (the note gives 20 to 37 days) | The shortest episode, driven by the COVID liquidity shock |
| 5 | June 2022 to early 2023 | roughly half a year | The score recovered above zero in early 2023 |
The pattern is consistent: a negative score has never appeared outside a major drawdown, and every episode sat inside the eventual bottoming range. What the table does not show is a precise bottom. In 2015 the price kept drifting for months after the first negative print, and in 2022 the score went negative in June while the eventual low came in November. The old version of this article quoted 12-month returns after each entry; we have removed those figures because we could not reproduce them from a primary source in this review, and the Bitcoin returns tool lets you compute forward returns from any date yourself.
How to use the zone in a plan
- Decide your convention in advance: which provider's chart you follow and what counts as "green" (for example, a score at or below zero). Write it in your plan so that you cannot move the goalposts later.
- Check the current reading on the cycle dashboard, which pairs the Z-Score with other long-cycle indicators.
- Treat a green reading as permission to deploy a larger slice of a pre-set budget, not as a signal to deploy everything at once. The thermometer DCA playbook describes a zone-by-zone schedule.
- Confirm with at least one independent signal: realized losses, exchange reserves, the 200-week moving average or miner stress, as summarised in the on-chain bottom signals guide.
- Set the exit rule for the plan before entering: a score back above zero does not mean the drawdown is over; it means the discount to cost basis has closed.
- Account for execution cost. A plan that buys in many tranches during a green episode pays more fees; the fee comparison tool shows standard VIP 0 rates such as Binance spot 0.10% maker and 0.10% taker (checked 2026-11-01), and the DCA backtest lets you test tranche sizes.
Mistakes and risks
- Mixing MVRV ratio and MVRV Z-Score. The ratio is market cap divided by realized cap and sits near 1 at the cost basis; the Z-Score is a standard-deviation measure and sits near 0 there. A threshold such as "below 0.1" belongs to neither metric and appeared in older versions of this article; ignore it.
- Treating the first negative print as the low. Historical episodes lasted from a few weeks to more than 300 days.
- Ignoring the shrinking denominator effect. Later cycles print lower peak scores partly because the standard deviation term keeps growing, so top signals calibrated on 2013 or 2017 may not repeat.
- Assuming the sample is large. Five episodes over 15 years is a small data set; a structural change such as ETF ownership or a different liquidity regime can produce a bottom that never prints a negative score.
- Provider differences. Realized cap calculations differ slightly between data providers, so the day a score crosses zero can differ by provider.
Related tools and guides
- Cycle dashboard
- MVRV Z-Score: what it is
- Bitcoin MVRV Z-Score zones and common mistakes
- The Bitcoin bottom checklist: seven signals
- 200-week moving average tool
- Has Bitcoin bottomed? Judge with nine indicators
- Bitcoin DCA backtest
FAQ
Is a negative MVRV Z-Score a buy signal?
It is a valuation condition, not a timing signal. In every historical episode the price sat in its eventual bottoming range, but the zone lasted from a few weeks to more than 300 days, so a single negative print says little about the next month.
What is the difference between MVRV and MVRV Z-Score?
MVRV is a plain ratio of market cap to realized cap. The Z-Score subtracts realized cap from market cap and divides by the standard deviation of market cap, which turns the gap into a standard-deviation measure that can be compared across cycles.
Where do the green and red zones come from?
They are visual bands drawn by chart providers; Glassnode's and LookIntoBitcoin's documentation describe them without numeric cut-offs (checked 2026-11-01). Any numeric threshold you use should be stated as your own rule.
Does the score work for other coins?
Glassnode publishes it for several assets, but the Bitcoin series has the longest history and the only sample of full cycles. For assets with a short history, the standard-deviation term is unstable.
How often should I check it?
Weekly is enough. The metric moves slowly because realized cap changes only when coins move on-chain; daily checking adds noise rather than information.
Bottom line
The MVRV Z-Score is one of the few on-chain metrics with a clean record at cycle lows: five green-zone episodes, five bottoming ranges. Its value lies in telling you how far the market has fallen relative to the holders' aggregate cost basis, which is useful for sizing a pre-planned buy schedule. It cannot tell you the week of the low, it depends on a small sample, and it should be paired with independent confirmation and a fee-aware execution plan.
Educational and research content only, not investment, tax or legal advice. On-chain metrics depend on provider methodology and can be revised; verify current readings on the provider's own chart before acting.
