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

HMX is a non-custodial exchange for perpetual contracts built on Arbitrum. HMX was founded by Shu Wei, Rockman, and Taoree and was launched in 2022.

HMX tokenholders own and govern the protocol. They currently take a cut of the total trading fees paid by traders (revenue). HMX's investors include Coral DeFi, Compound Capital Partners, Cryptomind, etc.

github

twitter

discord

etherscan

website

Derivative exchanges

# HMX Active users (weekly) Methodology

## 1. Overview

Weekly active users (WAU) measures the number of distinct users who performed trading activities on the HMX protocol within any 7-day period. This metric uses a rolling window approach where each day's WAU represents the count of unique users who were active on that day or any of the previous 6 days. The calculation leverages HyperLogLog data structures for memory-efficient distinct user counting and provides breakdowns across multiple dimensions including project-level, chain, user type, and product.

## 2. Data sources

The metric is built from the following data sources:
- **Blockchain events**: Parsed Arbitrum blockchain event logs from the HMX protocol's TradeService contract
- **Trading events**: Specifically `LogIncreasePosition` and `LogDecreasePosition` events that capture position opening/increasing and closing/decreasing activities
- **User identification**: Primary account addresses extracted from trading events to identify individual traders
- **Chain configuration**: Chain availability filters that determine which blockchain networks are supported for the calculation

## 3. Step-by-step calculation

### Component A: User activity identification
- **Purpose**: Identify unique users performing trading activities on the HMX protocol
- **Logic**: Extract trader addresses from Arbitrum blockchain events for the HMX TradeService contract. The system captures two types of trading events: `LogIncreasePosition` events (when users open new positions or increase existing ones) and `LogDecreasePosition` events (when users close positions or reduce position sizes). Each event contains a `primaryAccount` parameter that identifies the trader's address. Users are classified as "Trader" type and associated with the derivatives market sector and perpetual business segment.
- **Key fields**: `block_timestamp` for timing, `transaction_hash` for deduplication, `chain` identifier (Arbitrum), `user_type` (Trader), `market_sector_id` (derivatives), `business_segment_id` (perpetual), and `user_address` from the `primaryAccount` parameter

### Component B: Daily active user aggregation
- **Purpose**: Create daily snapshots of unique active users with efficient counting structures
- **Logic**: Aggregate user activities by truncating blockchain timestamps to daily boundaries. For each day, collect all unique user addresses who performed any trading activity. The system creates HyperLogLog (HLL) sketches for each day, which are probabilistic data structures that enable memory-efficient distinct counting. Daily aggregations are grouped by timestamp, chain, user type, and product dimensions to support downstream breakdowns.
- **Key fields**: `timestamp` (daily truncated), `timestamp_grain` (day), `chain`, `user_type`, `product_id`, and `user_dau` containing the HLL sketch of active users

### Component C: Rolling window construction
- **Purpose**: Create 7-day rolling windows for calculating weekly active users
- **Logic**: Generate a comprehensive datespine covering all days from the earliest user activity to yesterday. For each day in this datespine, create joins to the previous 6 days plus the current day to form 7-day windows. This approach ensures that every day has a complete 7-day lookback period, even if some days had no user activity. The window construction uses a cross-join between the datespine and a generated array of the 7 days ending on each timestamp.
- **Key fields**: `timestamp` (the end date of each 7-day window), `joinday` (each of the 7 days within the window), and corresponding dimensional attributes

### Component D: Dimensional breakdowns
- **Purpose**: Calculate WAU across different analytical dimensions
- **Logic**: Merge HyperLogLog sketches across the 7-day windows to count unique users, with separate calculations for different dimensional breakdowns. The system creates four distinct views: project-level aggregation (all users regardless of other dimensions), chain-level breakdown (users by blockchain), user type breakdown (users by classification), and product breakdown (users by product category). Each breakdown merges the appropriate HLL sketches and extracts the final distinct user count.
- **Key fields**: `dimensions` (indicating breakdown type), `chain_id`, `product_id`, `user_type`, and `user_wau` containing the final count

### Final aggregation
- **Combining components**: Union all dimensional breakdowns into a single result set, with null values for non-applicable dimension columns in each breakdown
- **Time aggregation**: Daily grain timestamps represent the end date of each 7-day rolling window
- **Final calculation**: Extract distinct user counts from merged HyperLogLog structures using the `hll_count.extract()` function, then apply the `finalize_metric_user_wau` macro for standardization and gap-filling
- **Quality controls**: Apply chain availability filters to exclude data from unsupported networks and use incremental processing with lookback periods for efficient updates

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

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

HMXHMX

Active users (weekly)

Active users (weekly)

Competitive landscape

Chain breakdown