Tornado Cash uses smart contracts that accept ETH & other token deposits from one address and enables their withdrawal from a different address. To preserve privacy a relayer can be used to withdraw to an address with no ETH balance.

Tornado Cash is a fully decentralized non-custodial protocol allowing private transactions on Ethereum and other adjacent blockchains. Tornado Cash was co-founded by Roman Semenov in 2019.

TORN token holders own and govern the protocol. They currently take a cut of the total withdrawal fees paid (revenue).

github

etherscan

Infrastructure

Tornado Cash

# Tornado Cash Average fees per user (AFPU) Methodology

## 1. Overview

Average Fees Per User (AFPU) measures the average amount of fees paid per daily active user on Tornado Cash. This metric is calculated by dividing total daily fees by the number of daily active users, providing insight into the average fee contribution per user. The calculation uses withdrawal events from Tornado Cash privacy pools across multiple blockchain networks to identify both users and fee amounts from the same transaction data.

## 2. Data sources

The calculation relies on the following data sources:

- **Tornado Cash withdrawal events**: Parsed `Withdrawal` events from Tornado Cash smart contracts across Ethereum, Arbitrum, Avalanche, BSC, Gnosis, Optimism, and Polygon networks. These events contain both user addresses (withdrawal recipients) and fee amounts paid for the withdrawal service.

- **Token contract data**: Token contract addresses and metadata for various assets supported by Tornado Cash, including ETH, DAI, USDC, USDT, WBTC, and native tokens on each chain (AVAX, BNB, MATIC, xDAI). This includes decimal precision information for proper amount normalization.

- **End-of-day price feeds**: Token price data used to convert fees denominated in various tokens to USD equivalents for consistent aggregation across different assets and chains.

- **Chain availability data**: Information about when each blockchain network became available for data collection, used for filtering and ensuring accurate historical calculations.

## 3. Step-by-step calculation

### Component A: User identification from withdrawals

- **Purpose**: Identify daily active users based on Tornado Cash withdrawal activity
- **Logic**: Extract user addresses from Tornado Cash `Withdrawal` events across all supported chains and token pools. The `to` field from withdrawal events identifies the recipient address as the active user. Users are aggregated daily using timestamp truncation to day boundaries and counted using HyperLogLog sketches for efficient distinct counting across large datasets.
- **Key fields**: `block_timestamp` for daily aggregation, `to` as `user_address`, `transaction_hash` for deduplication, `chain` for multi-chain support

### Component B: Fee collection from withdrawals  

- **Purpose**: Calculate total fees paid for Tornado Cash withdrawal services
- **Logic**: Extract fee amounts from the same `Withdrawal` events used for user identification. The `fee` field contains the withdrawal fee amount in the native token denomination. Fees are aggregated daily and converted to USD using end-of-day price feeds, with proper decimal normalization based on each token's contract specifications.
- **Key fields**: `block_timestamp` for daily aggregation, `fee` as `amount`, `token_address` for price lookup and decimal handling, `chain` for multi-chain aggregation

### Component C: Daily user aggregation

- **Purpose**: Calculate distinct daily active users using scalable counting methods
- **Logic**: Aggregate user addresses by day using HyperLogLog sketches to efficiently handle large-scale distinct counting. Users are counted across multiple dimensions including project-level totals, chain-specific breakdowns, user type classifications, and product-specific segments. The HLL approach allows for approximate but highly accurate distinct counting without memory constraints.
- **Key fields**: `timestamp` (daily), `user_dau` (distinct count), `chain`, `user_type`, `product_id`

### Component D: Daily fee aggregation and USD conversion  

- **Purpose**: Calculate total daily fees in USD across all tokens and chains
- **Logic**: Aggregate withdrawal fees by day, applying decimal correction based on token contracts and converting to USD using end-of-day price feeds. Fees from different tokens and chains are combined into a single USD-denominated daily total. The conversion process handles varying decimal precisions across different token standards.
- **Key fields**: `timestamp` (daily), `fees` (USD amount), `chain_id`, `token_address`

### Final aggregation

- **Combining components**: Daily user counts and daily fee totals are joined by timestamp using the `calculate_arpu` macro, which performs the division operation to compute average fees per user
- **Currency conversion**: All fees are converted to USD using end-of-day price feeds before aggregation, ensuring consistent currency denomination across different tokens and chains
- **Time aggregation**: Both users and fees are aggregated to daily boundaries using timestamp truncation, with proper handling of timezone normalization
- **Final calculation**: AFPU is calculated as total daily fees (in USD) divided by total daily active users, with the result representing the average fee amount paid per user on each day

The final metric provides the average amount of fees (in USD) paid per daily active user of Tornado Cash, aggregated across all supported chains and token denominations.

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

Tornado CashTORN

Average fees per user (AFPU)

Average fees per user (AFPU)

Competitive landscape