Every AI-powered trading tool faces the same structural flaw: one model, one perspective, one blind spot. A model tuned for trend-following will miss mean-reversion setups. A conservative model will sit out the trades that matter. Ask a single AI whether to go long on BTC, and you get one opinion — shaped by whatever the model was optimised for.
Monetum Intelligence uses a different architecture. Instead of one AI, five independent archetypes analyse the same market data simultaneously, each with a distinct risk profile and analytical lens. Their votes are weighted, aggregated, and a consensus signal fires only when the evidence is overwhelming.
Each archetype is an independent reasoning agent that receives the same market state — price structure, regime, funding rates, on-chain context, CVD, volatility — and returns a directional vote: LONG, SHORT, or NO TRADE.
| Archetype | Weight | Profile |
|---|---|---|
| Conservative | 0.8 | Prioritises capital preservation. Requires multiple confirming factors — clean structure break, regime alignment, and funding neutrality. Fewest signals, highest hit rate. Underweighted because caution can suppress valid setups in fast-moving markets. |
| Moderate | 1.0 | The baseline archetype. Balanced risk/reward assessment. Uses a standard conviction threshold and considers all inputs without extreme bias. Carries neutral weight as the calibration reference point for the swarm. |
| Aggressive | 0.8 | Willing to enter on early signals before full confirmation. Accepts wider initial stops. Counterweighted to balance its tendency to generate noise — flags early-stage setups without pulling the swarm into low-conviction trades. |
| Mean-Reversion | 1.0 | Looks for exhaustion, overextension, and snapback setups. Thrives in ranging conditions and after violent moves. Acts as a counterbalance to trend-following archetypes — critical for avoiding momentum chasing at extremes. |
| Trend-Following | 1.2 | Highest-weighted archetype. Identifies sustained directional momentum and filters counter-trend noise. Given extra influence because the V8b signal system is fundamentally momentum-based — trend confirmation is the most reliable input in the current strategy architecture. |
Each archetype casts a weighted vote. The consensus percentage is calculated as:
At 100% consensus — all five archetypes in agreement — the maximum possible score is reached. The swarm fires a signal when weighted consensus reaches ≥ 60%, equivalent to roughly three archetypes in agreement when accounting for weights.
This threshold was chosen deliberately: it is high enough to filter noise but achievable often enough to generate actionable signals. A 100% reading represents the highest possible conviction state.
Swarm consensus reduces single-model risk. When the Aggressive archetype fires a LONG but Conservative and Moderate disagree, the signal is suppressed — even if a single-model system would have sent an alert. When all five agree despite their different risk profiles, the signal has been stress-tested against multiple analytical frameworks before it reaches you.
Swarm consensus is visible in real time on the Dashboard, with per-archetype votes and weights shown so you can see exactly why the consensus reached its level — not just what the conclusion is.