Every strategy, fully explained
Understand exactly how a signal becomes a trade. No black boxes — every rule and default is documented here. Equity charts are illustrative and not a record of actual returns.
Social Momentum
Trending signalsDynamic discovery
Each cycle fetches the current StockTwits trending list — no watchlist required. The universe changes with the market's attention.
Multi-source enrichment
Every trending symbol is enriched with FinBERT sentiment, recent news headlines, SEC filings and Reddit discussion before being scored.
Risk controls
Hard stop-loss, max position size and max hold hours cap downside. Any breach exits the position on the next cycle.
LLM as final filter
Top-scoring candidates go to the LLM for a BUY/PASS verdict with reasoning — the composite score gets you in the room, the LLM decides.
// 90-day equity vs SPY · illustrative
Parameters
| Parameter | Default | Description |
|---|---|---|
| minCompositeScore | 0.40 | Minimum composite (sentiment × volume × freshness) to queue for LLM review |
| topN | 5 | Top-ranked symbols sent to the LLM per cycle |
| maxPositionPct | 15% | Max portfolio weight per position |
| stopLossPct | 4% | Hard stop-loss from entry price |
| maxHoldHours | 48 | Auto-exit positions held longer than this |
| extendedHoursTrading | false | Pre-market and after-hours limit orders |
Sentiment Momentum
EquityHow signals work
A local transformer scores news headlines. Stocks with rising sentiment velocity above the threshold become entry candidates.
LLM research
For each candidate the LLM reviews price action, RSI, volume and the top news items, then writes a BUY/PASS verdict with reasoning.
Risk controls
Hard stop-loss, max position size and max hold hours limit downside. Any breach exits the position on the next cycle.
Exit logic
Positions exit on stop, time limit, or an LLM-vetted reversal in sentiment — whichever comes first.
// 90-day equity vs SPY · illustrative
Parameters
| Parameter | Default | Description |
|---|---|---|
| minSentimentScore | 0.55 | Minimum bullish score to queue a signal |
| topN | 3 | Candidates sent to the LLM per cycle |
| maxPositionPct | 15% | Max portfolio weight per position |
| stopLossPct | 4% | Hard stop-loss from entry price |
| maxHoldHours | 48 | Auto-exit positions held longer than this |
| extendedHoursTrading | false | Pre-market and after-hours limit orders |
Hedge Momentum
Equity + HedgeHow signals work
Same sentiment scoring as Sentiment Momentum, but each long is paired with an inverse-ETF hedge sized off the long value.
LLM research
The LLM evaluates both the long candidate and the market regime; in high-volatility conditions it may recommend a larger hedge.
Risk controls
The hedge has its own size limit. In calm markets it is sized down to reduce drag on returns.
Regime aware
Hedge ratio adapts to volatility so downside protection scales with market stress, not a fixed rule.
// 90-day equity vs SPY · illustrative
Parameters
| Parameter | Default | Description |
|---|---|---|
| hedgeSymbol | SQQQ | Inverse ETF used as the hedge |
| hedgePct | 30% | Hedge size as % of long position value |
| minSentimentScore | 0.55 | Minimum score for long entries |
| maxPositionPct | 12% | Max portfolio weight per long position |
| stopLossPct | 5% | Hard stop-loss from entry price |
| extendedHoursTrading | false | Pre-market and after-hours limit orders |
Global Macro Rotation
Macro · ExperimentalRegime classification
Free FRED world-macro series are scored into a growth×inflation quadrant plus a risk-on/risk-off overlay, refreshed each cycle — no paid data feed required.
Rules-based rotation
Each quadrant maps to a liquid ETF sleeve (equities, cyclicals, real assets, or duration/quality). No LLM in the loop — the mapping is deterministic and backtest-safe.
Drift-based rebalance
Positions only rebalance once weight drifts past a configurable threshold, and a daily-loss kill-switch caps downside if the regime call is wrong.
Experimental
Unvalidated — paper trading only. No proven track record. Included for transparency while it accrues an out-of-sample history.
// 90-day equity vs SPY · illustrative
Parameters
| Parameter | Default | Description |
|---|---|---|
| maxPositions | 3 | Max sleeve ETFs held at once |
| deployFraction | 60% | Fraction of equity deployed across the active sleeve |
| rebalanceThresholdPct | 10% | Minimum weight drift before rebalancing a position |
| maxDailyLossPct | 5% | Kill-switch threshold for daily portfolio loss |
| minCoverage | 50% | Minimum FRED series coverage required before trading a regime |
Trading involves substantial risk of loss and is not suitable for every investor. Kairos is software, not investment advice. Charts and dashboards on this site are simulated and illustrative — they are not a record of actual returns and are not a promise of future performance.