Lenders deposit funds into  Fraxlend and receive interest payments from borrowers (supply-side fees).

Fraxlend is a non-custodial lending protocol built on Ethereum. Fraxlend is part of Frax Finance and was launched in 2022.

FRAX tokenholders own and govern the FRAX Finance protocol. Fraxlend is a part of Frax Finance. They currently do not take a cut of the total interest paid by borrowers (revenue).

twitter

discord

website

Lending

Fraxlend

# Fraxlend Active users (weekly) Methodology

## 1. Overview

Weekly Active Users (WAU) for Fraxlend measures the number of unique users who interact with the Fraxlend lending protocol within any 7-day period. This metric uses a sliding window approach where each day's WAU represents the count of distinct users who performed any Fraxlend action (borrowing, repaying, depositing, or withdrawing) during that day and the preceding 6 days. The calculation employs HyperLogLog (HLL) probabilistic data structures for efficient cardinality estimation across multiple dimensions including project-level, chain-level, user type, and product breakdowns.

## 2. Data sources

- **Fraxlend v1 contract events**: Parsed Ethereum events from Fraxlend lending pair contracts including `Pair_BorrowAsset_event`, `Pair_RepayAsset_event`, `Pair_AddCollateral_event`, `Pair_RemoveCollateral_event`, and `Pair_Withdraw_event`
- **Ethereum transaction records**: Transaction data containing user addresses (`from_address`) and block timestamps for correlating contract interactions with actual users
- **Blockchain timestamps**: Block-level timestamp data from the Ethereum blockchain for time-based aggregation and window calculations

## 3. Step-by-step calculation

### Component A: User identification and classification
- **Purpose**: Identify unique users and classify them by interaction type from Fraxlend contract events
- **Logic**: Extract user interactions from five specific Fraxlend contract events on Ethereum. Each event type corresponds to a different user classification: `BorrowAsset` events identify Borrowers, `RepayAsset` events identify Repayers, `AddCollateral` events identify Depositors, `RemoveCollateral` events identify Withdrawers, and `Withdraw` events also identify Withdrawers. The events are joined with Ethereum transaction records to obtain the actual user address from the `from_address` field and block timestamps. Each user interaction is tagged with metadata including chain (ethereum), market sector (lending), business segment (lending), and user type based on the specific action performed.
- **Key fields**: `block_timestamp` for time aggregation, `transaction_hash` for joining with transaction data, `from_address` as the user identifier, and event-specific parameters for user type classification

### Component B: Daily active user aggregation  
- **Purpose**: Calculate daily distinct active users with dimensional breakdowns
- **Logic**: Group the identified user interactions by day using timestamp truncation to the day level. Create distinct counts of users per day while maintaining dimensional attributes including chain, user type, market sector, business segment, and product identifiers. Each unique combination of user address and day represents one daily active user record. The output preserves all dimensional information needed for downstream WAU calculations.
- **Key fields**: `timestamp` truncated to day granularity, `user_address` as the distinct user identifier, `chain` for blockchain identification, `user_type` for interaction classification, and `product_id` for product-level breakdowns

### Component C: HyperLogLog initialization
- **Purpose**: Convert daily user sets into HyperLogLog sketches for efficient cardinality estimation
- **Logic**: Transform the daily active user data into HyperLogLog (HLL) probabilistic data structures using a precision parameter of 15. This allows for efficient approximate counting of distinct users across rolling windows without storing individual user addresses. Each HLL sketch represents the set of unique users for a specific day and dimensional combination. The sketches are grouped by timestamp, timestamp grain, chain, user type, and product identifiers.
- **Key fields**: `hll` sketches containing compressed user sets, dimensional grouping fields maintained from previous step

### Component D: Rolling window construction
- **Purpose**: Create a 7-day sliding window framework for WAU calculation
- **Logic**: Generate a comprehensive datespine covering all days from the earliest data date to yesterday. For each date in the datespine, create join relationships with the current day and the previous 6 days (total 7-day window). This creates a cartesian-style join structure where each target date is associated with all relevant dates in its 7-day lookback period. The window construction ensures that every day has complete coverage of its 7-day active user history, even if some days have no activity.
- **Key fields**: `timestamp` as the target date, `joinday` representing each day within the 7-day window for that target date

### Component E: Window aggregation with HLL merging
- **Purpose**: Calculate 7-day active user counts by merging HLL sketches across the rolling window
- **Logic**: Join the rolling window framework with the daily HLL sketches to collect all user activity within each 7-day period. For each target date and dimensional combination, merge the HLL sketches from all 7 days in the window using HLL merge operations. Extract the final cardinality estimate from the merged HLL sketch to get the count of distinct users active during the 7-day period. This process efficiently handles the union of user sets across multiple days while maintaining dimensional breakdowns.
- **Key fields**: Merged `hll` sketches representing 7-day user sets, `timestamp` for the target date, dimensional fields for grouping

### Component F: Dimensional breakdown generation
- **Purpose**: Create multiple aggregation levels for different analytical perspectives
- **Logic**: Generate four separate dimensional views of the WAU data. The project-level breakdown aggregates across all dimensions to show total protocol WAU. The chain-level breakdown maintains chain granularity while aggregating across user types and products. The user type breakdown focuses on different interaction patterns (borrowers, depositors, etc.) while aggregating across chains and products. The product-level breakdown enables product-specific analysis while aggregating across other dimensions. Each breakdown nullifies unused dimensional fields and applies HLL merging and extraction to calculate the appropriate user counts.
- **Key fields**: `dimensions` field indicating breakdown type, `user_wau` containing the distinct 7-day active user count, dimensional fields specific to each breakdown level

### Final aggregation
- **Combining components**: Union all four dimensional breakdowns (project, chain, user type, product) into a single result set, maintaining the dimensional structure and user count values from each breakdown
- **Currency conversion**: Not applicable - this metric counts users rather than monetary values
- **Time aggregation**: Daily boundary logic with 7-day sliding windows, where each day's metric represents unique users active during that day and the preceding 6 days
- **Final calculation**: The final WAU value for each day and dimensional combination represents the HyperLogLog cardinality estimate of distinct users who performed any Fraxlend interaction (borrowing, repaying, depositing collateral, removing collateral, or withdrawing) during the 7-day window ending on that date

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

Unique revenue-generating users over a rolling 7-day window.

Fraxlend

Active users (weekly)

Active users (weekly)

Competitive landscape