SparkTrade Low-Vol Equity Signals (Long-Short, VIX < 25)

## Daily long-short scores engineered for calm-water markets.

Daily 0-100 prediction ranks on **2 000+** U.S. equities. Model is trained exclusively on regimes where the Cboe VIX closes **below 25** to exploit gentle, trend-friendly tapes. Orthogonal to Fama–French factors (R² ≈ 0.12).

## **Detailed Description**

*SparkTrade Low-Vol Equity Signals* is an AI-driven data feed that identifies the stocks most likely to drift upward (long) or downward (short) during relatively **low-volatility market environments**—those tranquil stretches between macro shocks when trends tend to persist and mean-reversion opportunities abound.

Where our High-Vol sleeve hunts for sharp tail moves, this product specializes in **orderly markets dominated by fundamentals, steady sentiment, and incremental rotation.** It therefore functions as a **core alpha layer** that can sit quietly inside quantitative or discretionary books without spiking portfolio VaR.

## What makes it different 

1. **Regime-Specific Training** – The model ingests only history slices with VIX < 25, allowing it to learn price behaviors, factor sensitivities, and liquidity patterns that dominate in “boring” markets rather than crisis tapes. 

2. **Multi-Factor Ensemble** – Gradient-boosted trees (XGBoost-style) and random-forest learners trained separately for each regime (Low-Vol, High-Vol). Inputs span 20 yrs of price/volume, point-in-time fundamentals, micro-structure metrics, options skew, ETF flow, and sentiment. No RNNs or black-box sequence models; everything stays in a transparent, supervised-learning stack.

3. **Orthogonality** – Weighted-least-squares regression versus the Fama-French 3-factor model shows **adjusted R² ≈ 0.12**, meaning **88 %** of this sleeve’s variance is unexplained by classic market, size, or value betas. For allocators, that translates to **true diversification** instead of re-packaged “smart beta.”

## Coverage & History 

## **Universe**

>2 000 U.S. equities with ≥ $300 M float-adjusted cap and persistent liquidity.

## **Historical depth**

**Jan-1999 → present** (T-1 close).

## **Delivery cadence**

Secure-Share file publishes in Snowflake **by 20:00 ET every trading day** (≈ 4 hours after the closing bell).

## Data Dictionary

NAME TYPE DESCRIPTION

DATE DATE Date of prediction

TICKER STRING Stock ticker symbol (e.g., AAPL)

LONG_INDEX_SCORE INTEGER 0–100 scaled score (higher = stronger long)

SHORT_INDEX_SCORE INTEGER 0–100 scaled score (higher = stronger short)

LONG_RAW_SCORE FLOAT 0 to 1.0 unscaled signal before index conversion

SHORT_RAW_SCORE FLOAT 0 to 1.0 unscaled signal before index conversion

VOLATILITY_REGIME STRING "low", "high", or "auto" – or for ETF model "all weather"

MODEL_NAME STRING Source model name (e.g., "Low Vol", "AutoSwitch")

MODEL_VERSION STRING Version or release ID (for audit/debug)

## **Update Frequency**

Trading days, file posted **by 20:00 ET**