Liquidity providers deposit funds into Bancor and receive trading fees from traders (supply-side fees).

Bancor is a non-custodial exchange protocol built on Ethereum. Bancor was founded by Galia Benartzi, Guy Benartzi, Eyal Hertzog, & Yudi Levi and launched in 2016.

BNT tokenholders own and govern the protocol. They currently take a cut of the total trading fees paid by traders (revenue). Bancor's investors include DeFiance Capital, ParaFi Capital, Mechanism Capital, etc.

governance

github

twitter

discord

etherscan

website

Exchanges (DEX)

Bancor

# Bancor Average fees per user (AFPU) Methodology

## 1. Overview

Average fees per user (AFPU) measures the average daily fee revenue generated per active user on the Bancor protocol. This metric is calculated by dividing total daily fees by the number of daily active users, providing insight into revenue efficiency and user monetization. The calculation covers both Bancor v2.1 and v3 versions across Ethereum, handling the protocol's unique token conversion mechanics and fee structures.

## 2. Data sources

- **Bancor smart contract events**: Parsed `Converter_Conversion_event` from Bancor v2.1 and `BancorNetwork_TokensTraded_event` from Bancor v3 on Ethereum
- **Ethereum transaction data**: Transaction records including sender addresses (`from_address`) to identify users
- **Token metadata**: Token symbols, decimal places, and contract addresses for normalization
- **End-of-day price feeds**: Daily USD prices for fee tokens to enable currency conversion
- **Chain availability data**: Configuration data for filtering active chains and time periods

## 3. Step-by-step calculation

### Component A: Fee event extraction
- **Purpose**: Captures all fee-generating events from Bancor contracts
- **Logic**: Extracts fee events from two contract event types. For Bancor v3 `BancorNetwork_TokensTraded_event`, uses `targetFeeAmount` field and handles native token conversion by replacing the native token identifier (`0xeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee`) with WETH address. For Bancor v2.1 `Converter_Conversion_event`, uses `conversionFee` field with the same native token handling. Additionally processes BNT-specific fees from v3 events using `bntFeeAmount`, but sets amount to zero when either source or target token is BNT to avoid double-counting
- **Key fields**: `targetFeeAmount`, `conversionFee`, `bntFeeAmount`, `sourceToken`, `targetToken`, with fees always denominated in the target token

### Component B: Fee processing and USD conversion
- **Purpose**: Normalizes fee amounts and converts to USD values
- **Logic**: Applies decimal correction by dividing raw fee amounts by 10 raised to the power of the fee token's decimal places. Performs daily aggregation by truncating timestamps to day boundaries and summing fee amounts per transaction. Converts normalized amounts to USD by multiplying with corresponding daily token prices from end-of-day price feeds. Creates account names by concatenating source and target token symbols
- **Key fields**: Corrected `amount` values, USD-denominated `fees`, daily `timestamp` boundaries

### Component C: Trading event extraction for user identification
- **Purpose**: Identifies unique users from trading activities
- **Logic**: Extracts the same contract events used for fees but focuses on user identification. Joins trading events with Ethereum transaction data to obtain the transaction sender (`from_address`) as the user identifier. Applies the same native token to WETH conversion logic. Filters transactions starting from June 16, 2020 (Bancor v2.1 launch)
- **Key fields**: `user_address` from transaction `from_address`, trade amounts and tokens for context

### Component D: Daily active user aggregation
- **Purpose**: Counts distinct daily users across different dimensions
- **Logic**: Truncates timestamps to daily boundaries and identifies unique users per day using HyperLogLog (HLL) functions for scalable distinct counting. Aggregates users across multiple dimensions including project-level totals, chain breakdowns, user type classifications (all users are classified as 'Trader'), and product-level breakdowns. Uses HLL merge operations to combine user sets while maintaining distinct counts
- **Key fields**: `user_dau` counts, dimension identifiers (`chain_id`, `product_id`, `user_type`)

### Final aggregation
- **Combining components**: The `calculate_arpu` macro divides total daily fees by daily active users, matching records by date and dimensional attributes (chain, product). Both fee and user metrics are finalized with zero-filling for missing dates
- **Currency conversion**: Fees are converted to USD using end-of-day price feeds matched by timestamp, token address, and chain
- **Time aggregation**: All metrics use daily boundaries created by truncating timestamps to day level, with missing dates filled with zero values
- **Final calculation**: AFPU = Total daily fees in USD ÷ Number of distinct daily active users, with results filtered by chain availability rules to exclude inactive periods

*This methodology was automatically generated by Token Terminal's methodology analyzer.*

BancorBNT

Average fees per user (AFPU)

Average fees per user (AFPU)

Competitive landscape