Three layers, one displayed price

For any tracked asset, CounterTools may have up to three independent price signals at any moment:

• Direct DEX price — recent on-chain trades against XCP (or PEPECASH, or BTC).
• Multi-hop derived price — recent trades against another asset that itself has a fresh price.
• Dispenser anchor — open dispenser listings selling for BTC, MAD-filtered to reject outliers.

Each signal carries a confidence tier (HIGH/MEDIUM/LOW) based on its source, recency, and the volume of the underlying trade. The chart endpoint exposes all three series independently; the headline floor displayed elsewhere uses the most conservative qualifying value.

Direct DEX price

Counterparty's on-chain DEX matches buy and sell orders directly between addresses. When an order matches against XCP, the resulting trade gives us a per-unit XCP price for the asset. We multiply by the current XCP/USD rate to express it in dollars.

We require trades to meet a minimum volume threshold (per-asset, e.g. 5000 BITCORN, 10000 PEPECASH) to filter dust. Volume thresholds are documented per-currency and can be overridden when a market has shifted enough to need recalibration.

Multi-hop pricing

Many Counterparty assets don't trade against XCP — they trade against PEPECASH, BITCRYSTALS, other currencies, or sub-assets. For those, we derive an XCP price by routing through whichever counterparty has a fresh anchor.

Concretely, if BITCORN trades against PEPECASH at a known rate, and PEPECASH itself has a recent direct XCP price, then:

The BTC↔XCP conversion (when going through a BTC-denominated dispenser anchor) always uses the live global rate, so the error in one hop doesn't snowball with the error in another. We don't chain more than one hop — two-hop derivations multiply too much noise.

Time tolerance for anchor lookup

Quote-currency anchors aren't available at every block. When we need a PEPECASH/XCP rate at a given moment, we look for the nearest anchor within bands:

• ±24 hours → HIGH confidence on the derived rate
• ±7 days → MEDIUM (median of anchors in window)
• ±30 days → LOW (median of anchors in window)
• Beyond 30 days → no derivation, mark stale

Dispenser anchor — MAD outlier rejection

Dispensers are posted-price listings — anyone can deploy a dispenser at any rate. Some are clearly accidental (a wallet leftover at a 1000× off rate), some are scams, and most are genuine market prices. We use a statistical filter to reject outliers without manual curation.

We compute MAD on the 90-day rolling window of dispenser sales. The 90-day scope matters: it lets us detect outliers at the current price level rather than rejecting today's legitimate price as "wrong" because it happens to be 10× different from a 2022 value. Sample-size tiers:

DEX floor — the $10 minimum order filter

Separately from the historical-price methodology, when we compute the current floorfor an asset (the lowest available ask any buyer could hit right now), we ignore DEX orders whose total "get-side" value is under $10 USD.

This single threshold replaces a much hairier per-asset volume-threshold system. It's self-adjusting as prices drift — what counts as "dust" in PEPECASH terms changes as PEPECASH/USD changes — and it works identically for every asset on the platform without hand-curated overrides.

Confidence tiers

Every stored rate carries a confidence label so consumers (the API, the chart, your portfolio) can decide how much weight to give it.