Approach and data integrity

1. The 100, not a leaderboard

The tracked set is selected by a proprietary multi-signal process that evaluates performance context and observed trading behavior across multiple time horizons. Its central job is to distinguish accounts expressing useful directional information from activity that can create a misleading picture of conviction.

No single rank, position, or recent win is enough. The process is designed to reward durable signal and reduce the influence of accounts whose visible activity does not represent a consistent directional view.

What stays private: selection inputs, transformations, weights, thresholds, internal scores, and wallet identities. These are core HyperIndex intellectual property. We publish what the product measures, how to interpret it, and its limitations—not a recipe for reproducing the tracker.

2. Tracker integrity

The 100 are refreshed weekly and must pass automated checks for data sufficiency, set stability, structural integrity, and aggregate sanity. If a proposed refresh does not satisfy those controls, the previous accepted set stays live.

Accepted changes are deployed atomically. HyperIndex also pauses movement comparisons across a membership change so selection turnover is never presented as trader activity.

Latest tracker status

Loading the latest tracker status…

The public status record confirms the active trader count and refresh date. Selection diagnostics and wallet-level data remain private.

3. Metric definitions

Gross exposure is the sum of the tracked traders’ absolute long and short notionals. Net exposure is long notional minus short notional.

Asset tilt is asset net exposure divided by asset gross exposure. +1 is entirely long, −1 is entirely short, and values near zero are balanced.

Net leverage places directional exposure in the context of tracked capital. Conviction is a normalized measure of directional agreement among active positions.

HEAT Score is a normalized summary of aggregate direction and exposure. Positive readings indicate net-long positioning; negative readings indicate net-short positioning. Magnitude reflects positioning intensity—it is not a probability or price forecast. Exact transformations and weighting are proprietary.

4. Cadence and alerts

The public dashboard refreshes hourly. Free HyperIndex Signals publishes two scheduled public snapshots per day. Core evaluates the complete tracked asset universe after each successful hourly refresh and sends a private Telegram alert only after a regime, net exposure, tilt, or conviction change clears a materiality threshold.

A weekly refresh changes the measurement population. HyperIndex therefore pauses movement comparisons across that boundary instead of labeling selection turnover as trader activity.

5. Known limitations

• Hyperliquid-visible positions may be hedged on another venue or in spot.
• A strong historical record does not guarantee useful or persistent future signal.
• On-chain and API data can be delayed, incomplete, or erroneous.
• A screened trader or the aggregate read can be early, wrong, squeezed, or liquidated.
• The methodology and results have not been independently audited.