From Data to Decision: The Kairos Signal Pipeline
Manual options trading requires a trader to monitor dozens of tickers, track IV environments, evaluate fundamentals, check technical levels, and size positions — all simultaneously. Kairos automates that entire pipeline using a layered AI and quantitative signal system.
Layer 1 — Universe Screening (Magic Formula)
Before a single options contract is considered, Kairos filters the stock universe using a systematic fundamental screen inspired by Joel Greenblatt's Magic Formula: earnings yield and return on invested capital (ROIC). Only stocks that score well on both dimensions enter the eligible universe.
This ensures that any stock Kairos sells puts on is one you would rationally want to own — not just a high-IV lottery ticket.
Layer 2 — IV Rank Filter
Stocks passing the fundamental screen are then evaluated for their current implied volatility environment. Kairos calculates IV Rank across a rolling 52-week window and surfaces only tickers where IV Rank exceeds the configured threshold (typically 30+). Low-IV environments, where premium is thin and the risk-reward is unfavorable, are automatically excluded.
Layer 3 — Signal Scoring (EchoLoop)
Kairos's learning engine, EchoLoop, assigns a composite signal score to each candidate using a multi-factor model:
- IV weight — higher IV Rank scores more favorably.
- Delta weight — proximity of the strike to a target delta (typically 0.20–0.30) for optimal risk-reward.
- DTE weight — optimal days-to-expiration window (21–45 days) for theta decay acceleration.
- Regime weight — market regime (bull, bear, neutral, high-volatility) adjusts the overall aggressiveness of position sizing.
EchoLoop weights are not static. After each trade closes, the actual outcome (profit, loss, early close) is fed back into a Thompson Sampling bandit that slowly shifts weights toward the factor combinations that have produced the best risk-adjusted returns in live trading.
Layer 4 — Position Sizing (Modified Kelly)
Kairos uses a modified Kelly Criterion to size positions, capping individual trades at a fraction of the portfolio to prevent outsized losses from any single trade. The Kelly fraction is further scaled by signal confidence — high-confidence signals deploy more capital, marginal signals deploy less.
Layer 5 — Post-Entry Monitoring
Once a position is open, Kairos monitors it continuously for stop-loss triggers, take-profit targets (typically 50% of maximum profit), and DTE roll thresholds. The PortfolioGuardian runs 28 alert types spanning position risk, portfolio concentration, regime changes, and Greeks drift.
Transparency by Design
Every trade Kairos executes is logged with its signal score, reasoning narrative, and outcome. The public showcase portfolios on this site display live NAV, individual trades, and performance metrics — not hypothetical back-tests, but real paper and live executions with timestamped records.