Systematic · Data-Driven · AI-Powered
Strategy
AceAlpha is a systematic, AI-driven stock selection strategy that rebalances monthly. Each month, a proprietary model scores every large-cap S&P 500 stock across news attention, price momentum, options signals, and macro conditions — then constructs an equal-weight portfolio of the top-ranked names. No gut feel, no discretion.
The result is a disciplined, repeatable process with a verified edge.
Key Features
Designed to be simple to follow and execute.
Monthly rebalance
Low time commitment — one decision per month.
Fully systematic
No discretionary overrides. Data in, portfolio out.
Equal-weight positions
Simple to execute across any account size.
No intraday trading
All entries and exits at the monthly closing auction.
Dynamic position count
Scales with the number of high-conviction candidates.
S&P 500 universe
Large-cap, liquid stocks only — top 200 by market cap.
How the Model Works
Six steps, fully automated, every month.
Universe screening
We start with the largest, most liquid S&P 500 stocks by market capitalisation. Illiquid or thinly-covered names are excluded before any scoring begins.
Multi-source data collection
Each month the model ingests data across multiple categories: news and text sources, price and momentum data, options market signals, and macro regime indicators. All data is point-in-time — no look-ahead.
AI-powered scoring
A proprietary AI model processes the collected data and produces a composite score for each stock. The model combines several signal categories, each weighted and adjusted for current market conditions.
Quantitative risk filtering
Stocks failing minimum quality thresholds are removed automatically. Additional risk factors reduce a stock's score without excluding it entirely. The result is a clean, risk-aware ranked list.
Portfolio construction
The top-ranked stocks are selected into an equal-weight portfolio. Sector concentration controls prevent over-exposure to any single part of the market. Position count adjusts dynamically based on conviction.
Monthly rebalance
Positions are entered at market close on the last trading day of the month and held through the following month-end close. Subscribers can also execute at market open on the first trading day of the new month — either window produces similar results. The process repeats in full each cycle.
Signal Categories
Four independent data streams feed the scoring model.
News Attention
The primary signal. Measures the composition of a stock's news coverage — specifically, what fraction focuses on broader market forces rather than company-specific events. Higher scores have historically preceded outperformance.
Price & Momentum
Short and medium-term price trends, Bollinger Band positioning, 60-day moving average, and 10-day normalised momentum. Used primarily as a filter to avoid entries into deteriorating price action.
Options Market
Put/call ratio signals from the options market gauge directional risk and near-term market expectations. Elevated put/call ratios trigger automatic exclusion.
Macro Regime
VIX and Fear & Greed indicators scale overall portfolio exposure. In high-fear environments the model deploys less capital, reducing drawdown risk during broad market stress.
Risk Controls
Two layers of protection are applied before any stock enters the portfolio.
Hard filters
Stocks failing minimum criteria are automatically excluded — no exceptions. Targets clear signs of near-term risk: adverse price trends, unfavourable options positioning, net-negative sentiment.
Sector caps
Maximum position limits per sector prevent concentration in any single industry. Certain sectors are excluded entirely where the signal has no demonstrated predictive edge.
Score penalties
Elevated risk factors reduce a stock's composite score without removing it entirely. Borderline candidates are ranked lower, not blindly excluded — preserving signal from imperfect but viable names.
Regime scaling
In high-volatility or high-fear market regimes overall position sizing is reduced. The model deploys less capital when macro conditions are unfavourable.
Backtest Integrity
Strict walk forward process — no look-ahead, no curve-fitting. Each month is an out of sample test.
Signal date
Last trading day of the prior month
All data inputs are sourced strictly as of this date. No future information is used at any step.
Entry date
Last trading day of the prior month
Positions are entered at market close on the last trading day of the prior month.
Exit date
Last trading day of the current month
All positions close simultaneously. The cycle then restarts with fresh signals.