Bitcoin's blockchain is a complete, public, and permanently audited record of every coin movement since January 2009. Unlike price data, which reflects sentiment and speculation, on-chain data measures actual behaviour: when people move coins, where they move them, and how long they hold them.
Three on-chain metrics have appeared consistently at every major Bitcoin bottom since 2015. None of them predict the exact low. All of them identify the accumulation phase that precedes recoveries.
Metric 1: Exchange outflows
Exchange net flow is the difference between coins entering and leaving centralised exchanges. When more coins leave exchanges than enter, it typically means holders are withdrawing to self-custody — a long-term holding signal. When coins flow into exchanges, holders are preparing to sell.
During bear market bottoms, exchange outflows tend to spike in a specific pattern: large, irregular spikes in outflow volume while price is still falling or moving sideways. This is characteristic of large buyers accumulating — withdrawing substantial amounts from exchange wallets to cold storage.
The pattern visible in the chart above appeared clearly in:
Late 2018 / early 2019: After BTC bottomed near $3,200 in December 2018, exchange outflows spiked repeatedly through Q1 2019 as price consolidated. The subsequent recovery to $13,000 by June 2019 took approximately six months.
March 2020: The COVID crash produced the sharpest exchange outflow spike of the prior cycle. BTC fell to $3,800 on March 12 and then saw massive outflow spikes over the following weeks. The recovery to new ATH took approximately 12 months.
Late 2022: After the FTX collapse drove BTC to $15,500 in November 2022, exchange outflows elevated significantly through Q4 2022 and Q1 2023 as entities moved coins off centralised platforms.
The key nuance: exchange outflows can also indicate fear (moving coins off exchanges for safety rather than accumulation). The distinguishing signal is where the coins go. Accumulation outflows tend to move to cold storage addresses that then sit dormant. Fear-driven outflows often move to different exchanges or DeFi protocols.
Metric 2: Long-term holder supply
Blockchain analytics can measure how long each coin (technically, each UTXO) has been sitting unspent. When coins that haven't moved in 12+ months continue not moving — even as price falls — it indicates a category of holders who are either completely indifferent to the price move or actively accumulating and holding.
Long-term holder (LTH) supply tends to reach cycle highs at or near price lows. The dynamic is intuitive: long-term holders don't sell during drawdowns, so their proportional share of total supply grows as short-term holders capitulate. When LTH supply starts declining — because those long-term holders begin distributing — it tends to coincide with the late stage of bull markets.
At the 2022 bottom, long-term holder supply reached approximately 76% of total circulating supply — a multi-year high. This level is consistent with extreme supply inelasticity: the holders who have survived the full drawdown and are still holding are overwhelmingly committed regardless of price.
Metric 3: UTXO realised price (the MVRV ratio)
The Market Value to Realised Value (MVRV) ratio compares Bitcoin's current market cap to its "realised cap" — an approximation of what all current holders paid for their coins. When MVRV is below 1.0, the average holder is at a loss. Historically, an MVRV below 1.0 has corresponded closely with cycle bottoms.
The realised value as a metric is imprecise — coins change hands at different prices, exchanges hold coins in consolidated wallets, and lost coins reduce the effective supply. But as a directional indicator, the MVRV ratio has a strong record at identifying when market price has fallen below the aggregate cost basis of holders.
MVRV at major cycle bottoms:
| Date | MVRV | Price |
|---|---|---|
| Jan 2015 | 0.62 | ~$175 |
| Dec 2018 | 0.70 | ~$3,200 |
| Mar 2020 | 0.88 | ~$3,800 |
| Nov 2022 | 0.76 | ~$15,500 |
An MVRV below 0.8 has historically been associated with high-conviction accumulation zones. This doesn't mean price can't go lower — the 2015 and 2022 lows were both extended over several months after MVRV first breached 1.0 — but it does indicate that the average existing holder is currently at a loss, which tends to limit further selling pressure from that group.
Why these signals lag
None of these on-chain signals fire before the price bottom. They identify the accumulation phase — which begins at or near the bottom and continues for weeks to months afterward. By definition, you cannot use them to call the exact low.
What they're useful for is distinguishing between two types of sideways price action:
Distribution disguised as accumulation: price is flat but coins are flowing onto exchanges, LTH supply is declining, and MVRV is elevated. This tends to precede further declines.
Genuine accumulation: price is flat but coins are leaving exchanges, LTH supply is growing, and MVRV is depressed. This is the on-chain signature of a recovery building.
The exchange outflow metric is the fastest-responding of the three and the one most directly integrated into the scoring model on this site. The MVRV and LTH supply metrics are slower — they're better suited for identifying the broad phase of the cycle than for timing specific entries.
What these signals can't tell you
On-chain data has significant false positive rates. Large exchange outflows can be:
- Exchange internal wallet reshuffling (not user movement)
- OTC desk activity (not necessarily bullish)
- Regulatory-related movements (exchanges preemptively relocating assets)
Long-term holder supply can increase for mechanical reasons — coins aging into the 12-month threshold as a cycle extends, not because of new accumulation.
The signals are most reliable when they converge: significant exchange outflows and growing LTH supply and MVRV below 0.8 at the same time. Any single signal in isolation should be weighted lightly.