Methodology
The basket is chosen by data, every 30 days, with no discretionary picks.
1. Universe#
Every perpetual market on Hyperliquid's main dex and on every HIP-3 dex.
Exclusions#
- Stablecoins.
- Markets listed for less than 60 days.
- Markets that are delisted or in reduce-only mode.
2. Rank#
Each market is ranked by its notional volume over the last 30 days, summed from daily candles:
volume = Σ over 30 daily candles of v × (open + close) / 2
3. Select#
- HYPE5: the top 5 crypto markets that have a liquid HyperCore spot token.
- HYPE10: the top 10 of all markets.
- An existing member is kept until it drops below rank 7 (HYPE5) or 13 (HYPE10).
4. Weight and cap#
Weights are proportional to volume. Any asset above the cap (30% for HYPE5, 20% for HYPE10) is cut to the cap and the excess is shared out pro rata among the rest, repeating until nothing is above the cap.
raw weights BTC 41% ETH 33% HYPE 12% ZEC 9% SOL 5% after cap BTC 30% ETH 30% then the 14% excess is spread over HYPE, ZEC and SOL
5. Choose how each asset is held#
Spot if a HyperCore spot token exists and its order book within ±2% can absorb at least 10% of the target position. Otherwise a fully backed 1x long on the perp. See Spot and 1x long.
6. Publish, then execute#
The new composition is signed and written to the IndexRegistry contract three days before it takes effect, so anyone can see what will change. On the day, the fund sells what is overweight and buys what is underweight in small slices. The amounts held per token are recomputed at the current NAV, so the index price does not jump at a rebalance; only the mix changes.
Something unclear? Tell us.