What Research is
Research is where you read the market before you act on it. Instead of reacting to a single indicator, you assemble the market’s structure into a picture and name the regime — is this a trend, a range, or a volatility spike? The right system for a range is the wrong system for a trend, so naming the regime comes first.
What it reads
- •Funding rate — what perpetual longs pay shorts (or vice versa): crowd positioning and its cost.
- •Open interest — how much leverage is in the market; rising OI into a move means conviction, falling OI means unwinding.
- •Long/short ratio — how the crowd is leaning.
- •Fear & Greed — sentiment extremes that often precede reversals.
- •Dominance — where capital sits across the market.
- •Technical consensus — a rating built from ~25 indicators plus confluence checks, condensed into one read of trend/momentum.
From observation to hypothesis
The output is not “buy” or “sell.” It is a regime label and the evidence behind it, following the platform’s discipline: Observation → Reasoning → Risk → Recommendation → Next step. You leave with a hypothesis to test, not an order to place.
In-product example
Positive but stretched funding, rising OI, and a greedy sentiment reading together describe a crowded long — a regime where a range or mean-reversion system fits better than chasing the trend. That is a hypothesis you then validate in Test Lab.
The pitfall
Treating a regime read as a prediction. Structure describes the present balance of the market, not the future — funding can stay stretched for weeks, greed can get greedier. Research sharpens *which kind of system fits now*; it never promises *what price will do next*. Confirm the idea in Test Lab and let the Risk Gate guard the live step.
Next
Turn a regime hypothesis into a designed system with the AI Architect, then prove it in Test Lab & Proof.