By Keel Research Team · Updated May 12, 2026
Pick a basket of assets, choose equal-weight or inverse-volatility, optionally add a short basket for a long/short index, and click Trade. Your index opens in Keel as a real systematic strategy — ready to backtest, edit, and deploy on Hyperliquid.
Curated baskets from our research — customize anything after loading.
Assets to be long. Optionally add a short side for a long/short index.
How positions are weighted and when the pipeline runs.
The pipeline recomputes target weights once per bar at this cadence.
Further tuning — execution buffers, rebalance behavior, volatility windows — can be edited inside the Keel strategy editor after you click Trade.
Opens the Keel strategy editor with this index pre-built. Sign in on first visit — your config is preserved through the sign-in flow.
Every basket generates the same Keel pipeline: load price data, resample to your chosen timeframe, compute a baseline forecast for each asset, apply the weight allocator, then normalize to your target gross leverage. The strategy is a real systematic Keel pipeline — backtest, paper-trade, or deploy live, all from the same artifact.
Equal weight gives every asset 1/N of the target leverage, capped by the max-weight constraint. Simple, transparent, and the right default when you don’t have a strong volatility view. Inverse-vol (risk parity) sizes each asset inversely to its realized volatility — lower-vol names get more capital, higher-vol names get less. Use this when one asset (often the largest cap) dominates the basket’s return profile and you want a smoother ride.
Long/short mode wraps the pipeline in a parallel two-leg structure: long basket builds a positive forecast, short basket builds a negative one, and the legs combine into a single dollar-balanced position. The short side uses Hyperliquid perpetual futures — the differentiated feature vs long-only crypto-index ETFs.
Timeframe is the cadence the pipeline runs at — once per bar at the chosen interval (12h or 1d). Each run recomputes target weights from the latest data. The generated strategy uses buffered execution by default to keep turnover low; you can change the buffer width, execution mode, and other execution parameters inside the Keel strategy editor after deploy.
Keel is a Strategy OS for AI-assisted systematic trading on Hyperliquid. Build, backtest, and run live strategies with realistic fees, slippage, and funding modeled. Free to start — connect a Hyperliquid wallet when you’re ready to go live.
A crypto index is a basket of cryptocurrencies traded as a single position. You define which assets are in the basket and how they are weighted; the index gives you diversified exposure to a sector or narrative instead of betting on one token. Keel implements indexes as systematic strategies on Hyperliquid — they rebalance automatically and run as live trading positions, not as ETF wrappers.
Equal weight allocates 1/N to each asset. Inverse-volatility weight gives more capital to lower-volatility assets so each contributes roughly equal risk — useful when one asset (like HYPE) is much more volatile than the others. Custom weights are coming soon; for now the builder supports the two standard schemes that cover ~95% of index-fund methodology.
Yes — toggle "Add short side" and supply a separate basket of assets to short via Hyperliquid perpetual futures. The generated strategy combines both legs into a single dollar-balanced position with vol-targeted sizing. This is the differentiator vs traditional long-only crypto-index ETFs.
Your basket — assets, weighting, timeframe, leverage caps — opens directly in Keel as a real systematic strategy. You land in the strategy editor with everything pre-built, ready to backtest, modify, or deploy on Hyperliquid.
For the current version, the builder routes you to the Keel strategy editor for the full backtest. Inline preview metrics (Sharpe, return, max drawdown) are on the roadmap. The strategy ships with vol-targeted sizing and buffered execution to keep turnover low; everything is editable in the Keel editor.
Pre-built index of decentralized GPU + inference tokens (VVV, RENDER, IO, TAO). +67% YTD backtest, Sharpe 2.10.
Cross-sectional screener for HL perps. Filter by momentum, funding, volume, breakout, and regime metrics across 200+ pairs.
Live ranking of HL perps by current funding rate — annualized, with 7d persistence. Pairs well with a long-funding-short-no-funding index.