The Beta Threshold rule measures how sensitive Market A’s simple bar returns are to Market B’s over a rolling window (OLS β of A on B on the strategy timeframe). When that β meets your compare condition against a threshold, it fires a configured buy or sell.
Use it when relative sensitivity matters—for example, only enter when ETH moves more than 1:1 with BTC (β > 1), or exit when the hedge ratio collapses.
When to use it
- Confirm a pairs or relative-value idea when A is “hotter” or “cooler” than B
- Avoid entries when β is outside your risk band
- As a filter: allow a pending signal only when β agrees with the pending side
This is different from Correlation Threshold, which measures co-movement strength in [−1, 1]. Beta is a slope and is not bounded to that range.
What you’ll set in the app
Add Beta threshold under entry, exit, increase, decrease, or reversal—or under filters for the same sections.
- Market A — symbol / exchange / timeframe (empty = strategy primary)
- Market B — second market (symbol required)
- Window — number of return bars in the sample (default 20; needs
window + 1closes) - Compare —
beta >,<,≥,≤, or≈(equality uses a small tolerance) - Threshold — e.g.
1.0for unit beta - Side when true — buy or sell when the condition holds
Until enough history exists, Market B is missing, or B’s return variance is zero, the rule does not fire (as a filter, that blocks agreement for that instance).
Tips
- Beta is on returns, not raw prices.
- The window is in bars of your chosen timeframe.
- As a filter, set side to match the signals you want to allow.
Example
Entry rules: Beta threshold, Market A = strategy ETHUSDT, Market B = BTCUSDT, window 20, beta > 1.0, side buy, combined with RSI oversold. Entry only when ETH’s recent returns are more sensitive than 1:1 to BTC and RSI agrees.
Entry filters: Same rule with side buy so a MACD long only passes when β is elevated.