Bitcoin On-Chain Metrics: MVRV, SOPR, and NUPL for Market Timing

Three on-chain metrics separate Bitcoin cycle tops from bear-market noise better than any chart pattern. Each measures a different facet of holder behavior, and together they form a rules-based framework for sizing exposure.

MVRV (Market Value to Realized Value)

MVRV divides market cap by realized cap (the cost basis of all coins, valued at the price when last moved). It compares what the market pays today versus what holders actually paid.

• Below 1.0: holders are underwater on aggregate — historically a capitulation zone, BTC bottomed here in 2018, 2019, and 2022.
• 1.0–2.0: neutral accumulation band.
• Above 3.5: distribution zone; every prior cycle top (2013, 2017, 2021) printed MVRV above 3.5–4.5.

Use the 30-day SMA to filter daily noise. A reading under 1.2 with price above the 200-day MA is a high-probability accumulation trigger.

SOPR (Spent Output Profit Ratio)

SOPR tracks whether coins moved on a given day were sold at a profit or loss. SOPR = 1.0 is the breakeven line.

• SOPR < 1.0 for multiple consecutive days signals forced selling and capitulation — a buy zone when it reclaims 1.0.
• SOPR spiking above 1.10–1.15 means long-term holders are taking large profits; pair this with elevated MVRV to time exits.
• During the 2022 bear market, SOPR below 1.0 for 30+ days marked the November bottom region.

Track long-term-holder SOPR (LTH-SOPR, coins held >155 days) separately; a drop below 1.0 there is rarer and more reliable than the all-coins version.

NUPL (Net Unrealized Profit/Loss)

NUPL = (market cap − realized cap) / market cap. It expresses aggregate unrealized P&L as a percentage.

• < 0: capitulation (fear).
• 0–0.25: hope and fear range — accumulation.
• 0.25–0.5: optimism.
• 0.5–0.75: belief — late-cycle.

• 0.75: euphoria — distribution zone.

A NUPL print above 0.75 has preceded every major BTC top. Combine with MVRV > 3.5 and LTH-SOPR > 3 for a triple confirmation before trimming.

Rules of use

• Treat single-day prints as noise; require 7-day averages.
• On-chain signals lag price by days, so use them for bias, not entries.
• Re-balance exposure in steps: scale in when two of three metrics hit accumulation bands, scale out when two hit distribution bands.