Quant: Deterministic signal computations

Deterministic signal computations — pure Haskell math, no LLM.

Files: Omni/Fund/Quant/Signal.hs

Deliverables:

• Signal type (asset, signal type, value [-1,1], confidence, source, timestamp)
• SignalType enum (Momentum, MeanReversion, Volatility, MacroRegime, InsiderSentiment, EarningsSurprise, CrossSignal)
• SignalBundle type (timestamp, signals, correlation matrix, alpha scores)

Deterministic signal functions:

• momentumSignal :: Int -> Map Text [DailyBar] -> Map Text Signal — cross-sectional z-score of trailing returns
• meanReversionSignal :: Int -> Map Text [DailyBar] -> Map Text Signal — z-score vs SMA
• volRegimeSignal :: Map Text [DailyBar] -> Map Text Signal — short/long vol ratio
• macroRegimeSignal :: Map Text [FredObs] -> Signal — composite macro indicator

Utility functions:

• crossSectionalZScore :: Map Text Double -> Map Text Double
• rank :: [Double] -> [Double] — with tie handling

Properties (QuickCheck):

• Signal values bounded in [-10, 10] for z-scores
• Null input → empty signals
• Rank output is a permutation of [1..n]