Conviction Trading Strategy

A multi-factor conviction-based trading strategy engine with backtesting, technical indicators, sentiment analysis, and AI thesis validation.

Comment

Conviction Trading Strategy

The goal of Conviction Trading Strategy is to test, validate, and optimize market trading ideas using multi-factor conviction scoring. Instead of relying on just one indicator (like only RSI or only earnings news), institutional traders combine multiple independent signals—technical indicators, earnings surprises, insider buying, news sentiment, and macro market regimes. This application turns any market thesis into an automated, mathematically testable strategy and simulates how it would have performed historically (backtesting) before risking real capital.

What Does the Application Do? The program continuously calculates a Conviction Score from 0 to 100 for each market asset by evaluating five core pillars: Technical Momentum: Evaluates RSI oversold rebounds, 50-day and 200-day moving averages, MACD crossovers, and volume surges.

Earnings & Surprises: Quantifies quarterly earnings beats or misses and guidance momentum.

News & Social Sentiment: Measures market sentiment and breaking media catalysts.

Insider & Fundamentals: Tracks corporate executive (C-Suite) open-market buying and selling activity. Macro Regime: Measures overall market volatility (VIX) and risk-on vs. risk-off conditions. Step-by-Step Instructions: How to Use It

  1. The Backtest Terminal (Default View) Select an Asset & Timeframe: Use the top bar to switch between assets (such as NVDA, TSLA, AAPL, SPY, BTC-USD) and periods (6M, 1Y, 2Y).

Inspect Key Performance Metrics:

Strategy Return vs. Benchmark: Shows whether the conviction strategy beat standard buy-and-hold investing (Alpha). Sharpe & Sortino Ratios: Measure risk-adjusted returns (how much return you get per unit of volatility). Win Rate & Profit Factor: Shows the percentage of profitable trades and the ratio of gross profits to gross losses.

Maximum Drawdown: Measures the worst peak-to-trough portfolio loss. Review the Interactive Charts: Price Action: Candlestick chart showing the exact entry (BUY) and exit points on the chart with moving averages.

Equity Curve: Visualizes how your account balance grew over time compared to the underlying stock.

Conviction Oscillator: Shows daily historical conviction scores, with green entry lines () and red decay exit lines ().

  1. Conviction Weights & Rules Tab

Pick an Institutional Preset: Quick-start with pre-configured hypotheses:

Oversold Institutional Accumulation (RSI < 30 + Insider buying)

Earnings Beat & Volume Breakout (Earnings surprise + MACD momentum)

Breaking News & Sentiment Surfer (Catalyst-driven momentum)

Multi-Indicator Technical Confluence (RSI, MACD, 50/200 SMA)

Adjust Factor Sliders: Customize the weight of each factor (Technicals, Sentiment, Earnings, Insider Buying, Macro).

Set Risk Gates:

Entry Conviction Gate: Minimum composite score needed to buy (e.g., ).

Exit Conviction Decay: Threshold to automatically close the trade if thesis weakens (e.g., ).

Stop-Loss & Take-Profit: Hard percentage limits to protect capital and lock in gains.

Position Sizing Mode: Choose between Conviction-Weighted (allocating more capital to trades with higher confidence scores), Fixed %, or Full Capital.

  1. AI Thesis Architect Tab Formulate Any Strategy in Plain English: Type your thesis into the prompt box (e.g., "Buy when earnings beat by 10% and RSI is recovering from below 30, but exit quickly if market volatility spikes").

Click "Evaluate & Synthesize Strategy": The integrated Gemini AI quant model analyzes your hypothesis, calculates the optimal factor weights, sets protective stop-losses, highlights risks, and automatically updates the backtest engine.

  1. Trade Log & Forensics Tab

Audit Every Trade: View a complete chronological record of every trade taken during the backtest.

Forensics Modal: Click any trade row to see the exact breakdown of all five conviction pillar scores on the day the trade was opened, along with why it was closed (Take Profit, Stop Loss, or Conviction Decay).

  1. Live Market Conviction Radar Tab Real-Time Watchlist: View live conviction scores, signal states (BUY, HOLD, EXIT), and primary driving factors across multiple assets.

Interactive Stress-Testing: Click the simulation buttons (Earnings Beat, Viral Sentiment Surge,

Insider Buying, or Flash Liquidity Crash) to see how the agent adapts its conviction scores and signals to unexpected market shocks in real time.

Share this project:

Updates

Submission history