TL;DR: Cross-asset trading signals use one market's state to decide trades in another: Bitcoin's trend to gate altcoin longs, the VIX to switch crypto risk on or off, or gold and the S&P 500 as macro context. They work best as slow regime filters, not as precise triggers, and they only mean something if the other market's data is aligned point-in-time. In Trigr, any node in a strategy graph can read a different asset, and the engine carries completed values forward without reading ahead.
What are cross-asset trading signals?
Most strategies look only at the market they trade: its price, volume and indicators. A cross-asset signal adds information from somewhere else. Traditional intermarket analysis has done this for decades with bonds, the dollar, commodities and equities. In crypto, the most important "other asset" is usually Bitcoin itself.
Typical examples:
- BTC as the market regime for altcoins. Most large altcoins move with BTC most of the time. Buying SOL breakouts while BTC is breaking down is a low-odds trade.
- The VIX as a risk-appetite gauge. Rising implied volatility in US equities often coincides with de-risking across asset classes, crypto included.
- Gold and equity indices as macro context. A falling S&P 500 or a sharp gold rally can mark a risk-off environment.
- ETH relative to BTC. Whether ETH is outperforming can color how you trade other smart-contract tokens.
The value is not that these markets "predict" crypto. It is that they describe the environment, and many strategies work in one environment and fail in another.
Why are cross-asset inputs better as filters than triggers?
Relationships between markets are real but unstable. Correlation between BTC and US equities has risen and fallen over the years. Gold's relationship with crypto has been weak or inconsistent. A lead-lag pattern you find on hourly data, such as "BTC moves first, alts follow an hour later," often disappears once you account for costs and fill timing.
So use the other asset to answer "is this a good environment for my idea?" rather than "should I enter right now?"
| Cross-asset input | Good use | Weak use | Why |
|---|---|---|---|
| BTC trend (EMA, SuperTrend) | Gate altcoin longs and shorts | Hourly lead-lag trigger | Alt-BTC co-movement is strong; timing edges are fragile |
| VIX level or direction | Risk-on/off filter for crypto | Intraday entry signal | Daily series; changes slowly and publishes on trading days |
| VIX term structure | Stress regime filter | Anything intraday | Also a daily series |
| Gold or SPX trend | Macro context filter | Direct trigger for crypto | Relationship with crypto shifts across regimes |
| BTC funding or OI | Crowding filter for alts | Standalone signal on another asset | Useful context, but noisy |
The macro data guide covers the slower series (FRED, COT, EIA) in more detail, including publication lags.
How do you align another market's data without look-ahead?
This is where most hand-built cross-asset backtests go wrong. Three specific traps:
Different bar schedules. If you trade SOL on 1H bars and filter on BTC's 4H trend, the 4H bar that covers 12:00 to 16:00 is not finished at 13:00. Using its value at 13:00 means using BTC's price from up to three hours in the future. The only correct value at 13:00 is the last completed 4H bar, the one that closed at 12:00.
Different trading calendars. Crypto trades 24/7. The VIX is published on US trading days. On a Saturday, the most recent VIX value is Friday's. A naive join on calendar dates can leave gaps, or worse, fill them with Monday's value.
Publication timing. A daily close is only known after the close. If your dataset stamps it with the start of the day, a join on that label leaks the whole day. FRED's VIX series (VIXCLS) is a daily close, and Cboe's VIX index page explains what the index measures.
Trigr's engine handles all three the same way. Every node uses only information available at the bar's close. When a cross-asset input is on a coarser timeframe, its last completed value is carried forward and never read ahead. For a cross-asset indicator such as BTC's 200 EMA, the engine computes the indicator on the reference asset's own history first, then aligns completed values onto your strategy's timeline. The point-in-time backtesting guide explains why this matters so much, and the same logic underpins multi-timeframe strategies.
Fills happen at the next bar's open after a signal, so even a correctly aligned cross-asset filter cannot trade at a price that was only visible in hindsight.
Which cross-asset inputs does Trigr support?
In Trigr's Studio, any node can carry an asset different from the one the strategy trades. The strategy still trades its primary market; the other asset only feeds the condition. (The no-code strategy builder guide covers the full trigger, filter, signal and risk structure.) Options include:
- Any certified crypto perp as a reference. Read BTC's EMA, SuperTrend, RSI or ADX while trading an altcoin. Crypto references can also use derivatives data such as BTC funding.
- TradFi markets via Hyperliquid HIP-3. GOLD, SILVER, OIL, SPX and single stocks such as NVDA are available as reference assets with OHLCV-only inputs. For a trend read, use a price-based indicator, for example SPX price above its 200 EMA. Funding, open interest and similar derivatives data do not exist for these markets in the current beta.
- The VIX and its term structure. VIX rising or falling over a lookback, plus 9-day, 3-month and term-structure ratio series. These are global series usable with any traded market.
- Rates and the dollar. US Treasury yields, credit spreads and financial-conditions indices from FRED. One caveat: the broad dollar index publishes with a multi-day lag, so in Trigr it can be backtested but is not eligible for live agents. That is an honest consequence of point-in-time data rather than a limitation to work around.
The capability catalog lists every reference asset, timeframe and data source with its exact parameters, and which feeds exist for which asset.
What does an example cross-asset strategy look like?
An idea to test, not a recommendation. It trades AVAX on 4H bars.
- Trigger: AVAX 4H Donchian-channel breakout to the upside.
- Filter: BTC 1D price above its 100 EMA (cross-asset: crypto regime).
- Filter: VIX falling over 30 days (cross-asset: risk appetite).
- Signal: long when all three hold.
- Risk: 8% size, 2x leverage, long only, ATR trailing stop, 5% stop-loss, time stop at 30 bars.
Each filter has a reason to exist that you can state before seeing a backtest. That is the test for a good cross-asset filter. "Only buy altcoin breakouts when BTC is in an uptrend and equity volatility is subsiding" is a hypothesis. "Only buy when gold's 13-period RSI is between 41 and 58" is curve fitting.
Two practical checks. First, look at how many trades the filters remove. A filter that removes 90% of trades leaves a sample too small to judge. Second, look at which trades were removed. If the filter mostly removed losing trades in one crash, you have fitted one event.
How should you judge whether the cross-asset filter helps?
Compare the strategy with and without the filter, net of costs, and read the difference skeptically.
- Net, not gross. Fees and the builder fee are always applied in Trigr. Slippage and funding are opt-in; turn them on before comparing, since filters change trade count and holding time, which changes costs.
- Monthly returns. A filter that helps in one year and hurts in two is a coin flip. Look for consistency, not a single saved drawdown.
- Trial count. Trying five reference assets across four timeframes and three lookbacks is sixty variants. Run them as labeled experiments inside one strategy so the count stays visible in its version history, rather than saving sixty strategies and keeping the prettiest.
- Stronger evidence. A plain backtest does not compute the Deflated Sharpe Ratio; an ML optimization run reports DSR, PBO and feature importance, which can show whether a cross-asset feature carries information at all. The shaded trailing 25% in Studio is a recent-period diagnostic, not an out-of-sample holdout.
What this means for you
You can express intermarket ideas in a few nodes, without building your own data pipeline to align crypto, equity-volatility and macro calendars. The alignment is done the same way in backtests, paper agents and live agents, so a filter behaves identically across all three. For TradFi markets themselves, note that they are backtest and paper only in the current beta; see the HIP-3 guide for details and Hyperliquid's HIP-3 documentation for how those markets work.
Backtests are not guarantees; perps are leveraged and can lose more than expected, and cross-market relationships can change without warning.
Next steps
Add one cross-asset filter with a stated reason to a strategy you already have, then compare net results with and without it. The strategy builder docs show how to set a node's asset; plan limits for saved strategies and credits are on the pricing page.