Lyceum

Multiple minds. One market.

Lyceum trades the uncertainty.

Five probabilistic minds → entropy and JSD → option or NO_TRADE → deterministic risk → Alpaca paper.

PAPER ONLYREAD_ONLY SAMPLE

DEMO / SANITIZED SAMPLE DATA · Synthetic SPY market snapshot · No order, live account access, or P&L claim

Council

NEUTRAL

JSD disagreement

0.106

Entropy

0.813

Final risk

REJECTED

Five-agent market council

QUANT

TechnicalQuant

44% confidence · FLAT

SD1%D11%F45%U37%SU6%

Momentum across one-hour and one-day horizons.

OPTIONS

OptionsMarket

42% confidence · FLAT

SD2%D19%F49%U27%SU3%

Implied volatility relative to realized movement.

CATALYST

NewsCatalyst

30% confidence · FLAT

SD4%D24%F45%U24%SU4%

Catalysts evaluated without directional advocacy.

ADVOCATE

BullAdvocate

49% confidence · UP

SD0%D3%F31%U51%SU15%

Strongest evidence-supported upside interpretation.

ADVOCATE

BearAdvocate

35% confidence · DOWN

SD10%D40%F40%U10%SU1%

Strongest evidence-supported downside interpretation.

Consensus and decision

Strong down

2.8%

Down

17.8%

Flat

41.5%

31.7%

Strong up

6.2%

Selected option candidate

NO_TRADE

Consensus and volatility do not justify defined-risk exposure.

Skeptic: the council signal may be noise rather than a stable short-horizon edge.

Risk gate: REJECTED
NO_TRADE_SELECTED · SKEPTIC_VETO

Alpaca market data→five beliefs→consensus + JSD→option construction→skeptic→deterministic risk→Alpaca paper execution

Research discipline

33.58 yearsof SPY history

8,453SPY sessions

19registered hypotheses

361,439five-minute observations

9,627real option structures

Fail closedwhen evidence is missing

What the evidence says: close-to-open SPY drift and HAR-style volatility forecasting are statistically supported. The full LLM council has not demonstrated directional value-add, and ordinary expected moves do not clear observed option costs. No profitable executable edge is claimed; NO_TRADE is the default.