The Consensus is the most unusual bot in the competition — it doesn't analyze stocks directly. Instead it watches the other three bots and asks a single question: where do they agree? When strategies as different as quality compounding, catalyst value, and options trading all independently arrive at the same underlying, that convergence carries more information than any individual recommendation.

The logic is rooted in a well-established statistical principle: independent sources of error are less correlated than dependent ones. If three people with completely different reasoning processes all arrive at the same answer, the probability that they're all wrong for the same reason is much lower than if they were following the same methodology. The Consensus exploits this by treating each bot as an independent signal and amplifying their areas of agreement.

Every position The Consensus takes is sized according to how many bots independently agree on the same underlying. More agreement means higher conviction — and higher conviction means a larger allocation. The tier system makes this explicit and removes discretion from the sizing decision.

The Consensus has a hard constraint that distinguishes it from The Risk Taker: it only buys options on names it already holds as stock. Options amplify conviction on existing positions — they are never used as standalone bets on names not in the portfolio.

Cash management follows from this rule: stock allocations and option premiums draw from the same cash pool. The Consensus must reserve enough cash in its stock allocation to cover any options positions it intends to open in the same session. This prevents over-leverage and keeps total exposure within the convergence tier's intended sizing.

The Consensus tracks which convergence tier produced the best outcomes over time and which market conditions amplify or negate each bot's edge. Over weeks and months, this builds a picture of when Tier 1 signals are genuinely more reliable than Tier 2 signals — and when market regimes cause individual bots to be systematically right or wrong in ways that affect how much weight the convergence should carry.

The most important distinction the learning loop maintains is between genuine and coincidental convergence. Genuine convergence is three bots independently reaching the same conclusion through different logic. Coincidental overlap is three bots all responding to the same headline or news event — and arriving at the same name for the same reason. The latter is not a stronger signal. It is the same signal counted three times. The Consensus is specifically designed to identify and discount this pattern.