The weekly five. Written before the week it covers, so no prices and no forecasts. Each entry names the metric, what it measures, and the level that would actually change the read. Free-tier data is available from Glassnode Studio, CryptoQuant and mempool.space.
1. Realized cap
Realized cap values every coin at the price it last moved, which makes it a measure of capital actually committed rather than of market sentiment. The reading to take is the slope. Rising realized cap means new money is entering at current levels. Flat realized cap alongside a rising market cap means the same capital is being remarked upward. The second situation is far more fragile than the first. Mechanics in Realized Cap vs Market Cap.
2. Coin days destroyed
This weights every coin that moves by how long it sat still, so a five-year-old coin moving registers far more than a five-day-old one. Elevated readings mean patient supply is being spent, which is the behaviour that has historically accompanied cycle distribution. Low readings mean the churn is all short-term. What matters is a sustained elevation, not a single spike, which is usually one entity reorganising. See Coin Days Destroyed Explained.
3. Percent supply in profit
The share of all coins currently worth more than they last moved at. This is a direct measure of how much latent selling pressure exists, because coins in profit can be sold without realising a loss. Extremes are the useful part: readings above roughly 95 percent have clustered near cycle highs, and readings below 50 percent near capitulation lows. The middle of the range says very little. Covered in Percent Supply in Profit Explained.
4. Funding rates
Perpetual futures funding tells you which side is paying to hold its position. Persistently positive funding means longs are paying shorts, which indicates leveraged optimism and creates the conditions for a squeeze in the other direction. The number to watch is not today’s rate but how long it has stayed in one direction. See Funding Rates and Perpetual Futures Explained.
5. Mempool congestion
Pending transaction backlog and the fee rate required for next-block confirmation. This is the only genuinely real-time item on the list. Sustained congestion means people are paying for urgency, which is information that confirmed-transaction metrics only surface later. It also drives miner revenue, which feeds into miner selling. Explained in The Bitcoin Mempool Explained.
How to use this
These five were chosen to be non-overlapping: capital flow, holder behaviour, positioning, leverage and network demand. Adding a sixth metric that correlates with one already on the list adds confidence without adding information, which is the most common way onchain analysis goes wrong. The metrics guide has the full catalogue if you want to substitute.
What would make this list wrong
Every metric here has a failure mode worth naming in advance, because a metric you cannot imagine being wrong is one you will over-trust.
Realized cap can rise on internal transfers between custodians if the labelling is incomplete, which looks like new capital and is not. Coin days destroyed spikes on estate transfers, exchange migrations and long-dormant wallets consolidating for security reasons, none of which carry any market opinion. Percent supply in profit says nothing about willingness to sell, only about ability, and holders who intend to keep their coins for a decade do not care that they are in profit.
Funding rates can stay stretched far longer than any position can survive, and reading an extreme as an imminent reversal has been expensive many times. Mempool congestion rises on inscription waves and consolidation activity that have nothing to do with demand for Bitcoin as an asset.
None of this makes the metrics useless. It means each one answers a narrower question than the headline suggests, and knowing the narrow question is the whole skill.
Common questions
Why do these five change week to week? They rotate deliberately. Different market conditions make different metrics informative, and a fixed list trains you to ignore whichever one is currently flat.
Can I automate this? Partly. Most providers offer alerts on free tiers, and the free dashboard guide covers assembling them in one place.
What if two metrics disagree? That is the normal case and it is useful. Disagreement between positioning and holder behaviour usually means a move is leveraged rather than funded.
Research and education only. Onchain Decoded does not make price predictions.
Leave a Reply