What Is Cost Basis Distribution in Crypto?
A price chart tells you where the market has been. A cost basis distribution tells you who is winning, who is trapped, and where supply is likely to appear.
Most crypto traders stare at the same candlestick chart. It shows price over time, but it hides something more useful: the cost at which the people holding the asset actually bought. Cost basis distribution, often called chip distribution (筹码分布), reconstructs that missing layer.
The idea is simple. Every trade has a price. If you accumulate volume by price instead of by time, you get a map of where the market has built positions. Add a turnover model so old positions fade, and the map starts to resemble the current holder base.
Why price alone is not enough
Two coins can both trade at $10 after a long base. In the first, 80% of holders bought below $8 and are sitting on profit. In the second, 80% bought above $12 and are underwater. The candles look similar. The behavior ahead does not: the first coin has holders who can take profit, the second has holders waiting to break even.
This is the question chip distribution tries to answer: at the current price, how much supply is in profit, how much is trapped, and where are the concentrations?
How a chip distribution is built
- Collect candles. ChipEye uses up to 400 Binance candles for the selected timeframe.
- Split each candle across price. Volume is spread over the candle's high-low range, with more weight near the typical price. This is a triangular distribution, a standard CYQ approximation.
- Decay old chips. Each new candle's turnover removes a slice of existing chips. A 20-day average turnover means roughly 1/20 of the old base is replaced per day. Faster turnover makes old costs disappear sooner.
- Normalize. The final array sums to 100%, so every price level can be read as a share of current chips.
The five numbers worth knowing
| Metric | What it means | How to use it |
|---|---|---|
| Profit ratio | Share of chips whose cost is below spot. | Extremes flag crowded winners or a washed-out base. It is not a timing signal on its own. |
| Average cost | Quantity-weighted average entry of all chips. | A magnet for price. Above it, the average holder is winning; below it, bounces meet supply. |
| 90% cost range | Middle 90% of cost basis after trimming both tails. | Narrow range = converged costs. A break from a tight range often trends. |
| 90% concentration | Range width relative to its midpoint. | Below 15% is tight for altcoins, above 40% is dispersed. |
| Chip peaks | Local concentrations of supply. | The nearest peak above spot is resistance; the nearest below is support. |
The four patterns that matter
Low single peak
Chips concentrate near the low of the loaded range. Long sideways turnover has pulled costs together. This is a potential accumulation base, but it needs confirmation: a high-volume break above the peak that holds on the retest.
High single peak
Supply has rotated to the top of the range. If price loses the base of the peak, a large block becomes trapped at once. That is why high single peaks are treated as distribution risk rather than a launchpad.
Dual peak
One peak above and one below. The upper peak is exit supply; the lower peak is cost support. Price often ranges between them. A real break of either edge is more informative than a move in the middle.
Dispersed chips
After a large move, costs are spread across a wide band. Support and resistance at any single price are weak. The better trade is often to wait for a new concentration to form.
A practical three-step read
- Locate spot relative to the main peak. Above it, the peak is support; below it, the peak is overhead supply; inside it, expect chop.
- Measure the nearest peak. Distance and mass matter together. A peak 2% away holding 25% of chips is a much bigger obstacle than a thin peak 15% away.
- Confirm with volume and futures. A breakout with rising open interest and normal funding is healthier than one driven by short covering or crowded longs.
What chip distribution cannot do
It cannot tell you the future, and it cannot see every holder. It also degrades on short histories: a coin with 40 days of data has not traded enough for the turnover model to be representative. For those coins, treat the output as a rough estimate and check token unlocks separately.
How ChipEye calculates it
ChipEye rebuilds the distribution server-side for the top 200 Binance USDT perpetuals every 10 minutes. You can switch between spot and futures volume, daily and 4-hour candles, and scrub the chart to see the distribution as it stood at any candle close. Data is public Binance market data; the model is a quantitative estimate, not investment advice.
FAQ
Is chip distribution the same as on-chain cost basis?
No. On-chain cost basis uses wallet flows and UTXOs or account balances. Chip distribution uses exchange trade volume at price with a turnover decay model. It is an exchange-market estimate, not a wallet-level record.
Which timeframe should I use?
Use daily chips for swing and position views, and 4-hour chips for tactical timing. Shorter timeframes react faster but carry more noise and depend more on the turnover setting.