Every time some ancient wallet from the early days wakes up and moves a few thousand coins, my mentions fill with the same question. Is this it, is the top in, should we be worried. And the honest answer is usually no, or at least not for the reason people think. The metric that actually tells you whether an old-wallet move matters is Coin Days Destroyed, and most people either ignore it or misread it in the exact same way every time.
What the metric is actually measuring
The idea behind Coin Days Destroyed is simple once you sit with it for a minute. A coin accumulates one coin day for every day it sits unspent in a wallet. Ten coins that have not moved in one hundred days carry one thousand coin days. The moment those coins are spent, that accumulated time gets reset to zero, and the amount that got wiped out is what we call destroyed. It is not that the coins vanish. It is that the holding-time clock they had built up is gone.
The reason this is more useful than plain transfer volume is that it weights every movement by conviction. A million coins that were bought yesterday and sold today barely register, because they had almost no coin days to destroy. A far smaller stack that has not moved in five years lights the chart up, because each of those coins is carrying years of accumulated time. So CDD is a volume metric that has been quietly reweighted toward the people who have held the longest, and those are usually the people worth paying attention to.
You will see it presented a couple of ways. Raw CDD is spiky and hard to read on its own, so most people look at a smoothed version, often a thirty or ninety day average, or a supply-adjusted variant that divides by circulating supply so you can compare across eras. There is also a metric called Dormancy, which is CDD divided by transfer volume, and it answers a slightly different question. It tells you the average age of the coins moving on a given day, which helps you separate a lot of young coins churning from a little bit of very old money quietly walking out the door.
Why a spike is worth a second look, not a panic
When CDD spikes, the plain-English translation is that long-term holders repositioned. Somebody who sat through multiple cycles decided today was the day to do something with coins they had been ignoring for years. That is genuinely information, because those holders tend to have the best average cost basis and the least emotional relationship with the price. When they move, it is rarely random.
But repositioning is not the same as selling, and this is where the panic-reads fall apart. Old coins move for a whole list of reasons that have nothing to do with someone dumping on the market. Before you read a spike as distribution, you have to rule out the boring explanations, because the boring explanations are the common ones.
- Custodian migration. Exchanges and custodians periodically re-shuffle their cold storage, consolidate wallets, or roll to new infrastructure. Those coins can be extremely old and completely internal. Nothing was sold. The custody just moved.
- Wallet or address upgrade. Holders move funds to a new address format, a fresh hardware wallet, or a multisig setup. From the chain it looks identical to a spend, and it will destroy a mountain of coin days, but the beneficial owner never changed.
- Estate, key rotation, or internal treasury moves. Funds, DAOs, and long-term holders shuffle for security or accounting reasons that have no market intent behind them.
The failure mode I see most often is somebody seeing a huge CDD print, reading a headline about a decade-old wallet awakening, and treating it as a confirmed sell signal. Half the time those coins land on a fresh self-custody address and never touch an exchange. The clock got reset, the metric screamed, and nothing actually happened to supply on the market.
A workflow for reading an old-coin move
Here is roughly how I walk through one of these before I let it change anything I do. It takes a few minutes and it filters out most of the noise.
- Confirm the spike is real on a smoothed basis. One noisy day of raw CDD is not a trend. If the ninety day average or the supply-adjusted line is not moving, you are looking at a single event, not a regime change.
- Trace the destination. This is the whole game. Follow the coins one hop. If they moved to a known exchange deposit address, that is a real distribution signal worth respecting. If they moved to another cold wallet or a fresh address that just sits there, it is almost certainly a migration or an upgrade and you can mostly ignore it.
- Check whether it clusters. One old wallet is an anecdote. A run of independent old wallets all activating over a couple of weeks, with a rising share landing on exchanges, is the pattern that actually precedes distribution. Age of the coins moving, via Dormancy, tells you if this is broad or a one-off.
- Frame it against context, not in isolation. Old coins moving into strength, while price grinds up and demand absorbs them, is very different from old coins moving into weakness while price is already sliding. Same metric, opposite meaning.
What history says usually follows
The base rate is the part people skip, and it is the part that keeps you sane. Large dormant-supply awakenings happen far more often than tops do. If every ancient-wallet move marked a top, we would have topped dozens of times. Most single old-coin moves are followed by nothing that a normal person would notice, because the coins were migrated rather than sold, or because the market absorbed them without flinching.
The awakenings that have historically mattered were not single events. They were sustained, where old supply kept getting spent for weeks, and directional, where a rising fraction of it verifiably hit exchanges. Even then, CDD is early and noisy rather than a clean sell button. It tells you experienced money is on the move. It does not tell you they are right, and long-term holders have handed coins to eager buyers near local tops plenty of times only to watch price keep climbing. Treat a spike as a reason to raise your attention, not to act on reflex.
The practical version of all this is a small habit. When an old wallet makes the rounds, do not react to the headline. Pull the smoothed CDD, trace where the coins went, and ask whether it is one wallet or a cluster. Inside Blockcircle I keep dormancy and old-supply movement next to whale flow and exchange balances so I can answer those questions in the same view instead of stitching three tabs together, but the discipline matters more than the tool. If you cannot confirm the coins reached an exchange and cannot find a cluster, you are almost certainly looking at plumbing, not a decision.