TL;DR: Hyperliquid's HIP-3 lets builders deploy perpetual markets on assets like gold, the S&P 500 and individual stocks such as NVDA. With Trigr you can build strategies on eleven of these TradFi perps, backtest them on point-in-time data and forward-test them on paper agents today. Live TradFi execution is not available in the current beta, so treat this as a research and paper-trading workflow for now.
What is Hyperliquid HIP-3?
HIP-3 is a Hyperliquid improvement proposal for builder-deployed perpetuals. It allows a deployer that meets a staking requirement to launch its own perp DEX on Hyperliquid, with its own markets, margining and order books.
According to Hyperliquid's HIP-3 documentation, mainnet deployers must stake 500,000 HYPE and keep it staked for at least 183 days after deploying. The deployer is responsible for market definition, including the oracle and contract specifications, and for market operation, including setting oracle prices and leverage limits. Validators can slash the stake through a stake-weighted vote in the event of malicious market operation.
The result is that assets outside crypto, such as metals, oil, equity indices and single stocks, can trade as perps on Hyperliquid's infrastructure. Each HIP-3 DEX also chooses its own collateral asset; Hyperliquid's documentation says any quote asset can be used, so check the specifications of the market you trade.
Are HIP-3 stock perps the same as owning stocks?
No. A stock perp is a derivative contract that tracks a price feed. You do not own shares, you receive no dividends, and you have no voting rights. What you get is leveraged long or short exposure to the price, with margin, liquidation and funding like any other perp.
Some differences from crypto perps are worth keeping in mind:
- Underlying market hours. Stocks and indices have official trading sessions. Hyperliquid's HIP-3 documentation leaves oracle and contract specifications to the deployer, so how a market is priced while its underlying is closed depends on that deployer's design. Read it before trading around session boundaries.
- Oracle design. The deployer defines the oracle. Read how each market is priced before trading it.
- Deployer-set parameters. Leverage limits and other parameters are set per HIP-3 DEX, so they can differ from core Hyperliquid markets.
- Liquidity. Newer markets can have thinner books than major crypto perps, which makes slippage assumptions more important.
Event risk also looks different. Single stocks can move sharply on earnings, and indices and commodities react to scheduled releases such as inflation data or oil inventory reports. If news breaks while the underlying market is closed, the perp may reprice on its own, and any gap between the perp and the reference market can close abruptly when the session reopens. A stop that looks sensible on continuous price data may fill well away from its level around these events.
Which TradFi markets can you research in Trigr?
Trigr currently supports eleven HIP-3 TradFi markets alongside its 43 certified crypto perps:
| Ticker | Market | Type |
|---|---|---|
| GOLD | Gold | Metal |
| SILVER | Silver | Metal |
| OIL | Crude oil (WTI) | Energy |
| BRENTOIL | Crude oil (Brent) | Energy |
| SPX | S&P 500 | Equity index |
| NVDA | NVIDIA | Single stock |
| TSLA | Tesla | Single stock |
| AAPL | Apple | Single stock |
| GOOGL | Alphabet (Class A) | Single stock |
| MU | Micron Technology | Single stock |
| INTC | Intel | Single stock |
What you can do with them today:
| Capability | Crypto perps | TradFi (HIP-3) perps |
|---|---|---|
| Build in the no-code Studio | Yes | Yes |
| Backtest on point-in-time data | Yes | Yes |
| Paper agents | Yes | Yes |
| Live trading on Hyperliquid | Yes | Not in the current beta |
| Strategy inputs | Indicators plus API data (funding, OI, macro and more) | Price-based indicators, per-symbol implied vol, global macro series |
| Timeframes | 5m to 1M | 5m to 1M, except 30m |
The capability catalog is the authoritative, current list of markets and what each supports, so check it before designing a strategy.
How do you build a TradFi strategy in Trigr?
The graph structure is the same as for crypto: one trigger, optional filters, a signal, and a risk node. The no-code strategy builder guide covers each step. The main constraint is that each TradFi market's own data is price-only. That still leaves a wide toolbox: moving averages, RSI, Bollinger bands, ATR, SuperTrend, Donchian channels and more. Avoid volume-based indicators such as OBV or VWAP here, because the volume in this history is tick volume from the reference price feed rather than traded volume. Beyond price, a strategy can read per-symbol implied volatility from ORATS (ETF-proxied for gold, silver, oil, Brent and the S&P 500) and use global macro series such as the VIX or Treasury yields as filters.
Some illustrative ideas to test, not recommendations:
- Gold trend follower: GOLD 1D close above its 100-day EMA triggers a long, filtered by ATR as a percentage of price staying below a ceiling, with an ATR trailing stop.
- Index pullback: SPX 4H RSI(14) crosses up through 35 while the 1D close is above its 200-day moving average; long only, with a time stop so trades do not linger.
- Single-stock breakout: NVDA 1D close above a 20-day Donchian high; long only, modest size, with a stop-loss sized to recent volatility.
- Oil mean reversion: OIL 4H close below the lower Bollinger band triggers a long, with a short time stop.
Because a HIP-3 market's own history is short, long backtests for these assets rely on history from the underlying market rather than from the perp's own trading record. That is a reasonable research proxy, but it cannot capture everything about how the perp trades, especially outside the underlying's regular session. It is one more reason to forward-test on paper.
How should you backtest TradFi perps honestly?
The same engine rules apply to TradFi as to crypto. Every feature uses only information available at the bar's close, orders fill at the next bar's open after the signal, and if a take-profit and stop-loss are both touched within one bar, the stop wins. Trading fees and the builder fee are always applied.
Two costs deserve extra attention:
- Slippage. It is opt-in in Trigr and applied as a flat basis-point assumption per fill against you. It does not model order-book depth or market impact, so on thinner TradFi markets choose a conservative number.
- Funding. Perps pay or charge funding while you hold. Hyperliquid's funding documentation says the funding cap and interval do not depend on the asset, and that HIP-3 perps use a more responsive premium formula. Trigr's backtests do not apply funding to TradFi markets at all, not even as a flat assumption, and paper agents do not accrue funding either. Treat every TradFi result as gross of funding and estimate holding costs yourself for longer holds.
The engine also does not model exchange liquidation on TradFi markets, so a highly leveraged TradFi backtest can look safer than the same position would be on a live venue. With slippage off, results are labeled gross. The backtesting docs explain what gross means and what is not modeled.
Treat raw Sharpe with care: the Deflated Sharpe Ratio, which adjusts for the number of variants tried, comes from an ML optimization run, and a standard backtest marks it as not yet computed. Studio's out-of-sample split (the last 30% by default) is unseen only until you act on it, so do not treat a good-looking out-of-sample stretch as proof.
Why forward-test TradFi strategies on paper agents?
Paper agents are how you gather real forward evidence on TradFi strategies today. A paper agent runs your frozen strategy on live Hyperliquid prices, filling at the current price with a base-tier trading fee and the builder fee deducted, but no slippage or funding modeled. The paper trading guide covers how to read a forward record.
When comparing paper results with the backtest, look at more than profit and loss. Check whether trade frequency is similar, whether entries cluster around the underlying market's session opens and closes, and whether stops are being hit more often than the backtest suggested. Differences in any of these can point to behavior of the perp that the underlying history did not capture.
An agent can hold up to 52 strategy slots, one per market, each with its own allocation. That makes it possible to run a small cross-asset book on paper, for example a gold trend strategy, an S&P 500 pullback strategy and a BTC momentum strategy side by side, and watch how they behave together. The setup screen shows a combined backtest and a risk panel, and a frontier marker separates backtest from forward results.
Paper agents are available on every plan: the Free plan includes five, Trader includes 25 and Pro has no fixed limit. See pricing for details.
Backtests and paper results are not guarantees; perps are leveraged and can lose more than expected.
Next steps
Pick one TradFi market from the capability catalog, build a simple price-based strategy, backtest it with slippage turned on, and then run it on a paper agent to build a forward record.