Python Trading Toolkit

The Problem

You ask your AI agent to "check my portfolio" and it hallucinates stock prices, mixes up ticker formats between US and HK exchanges, or gives you a Sharpe ratio with no context about the underlying calculation. Without a structured toolkit, agents improvise with web searches that return stale data, miss A-share tickers entirely, or produce analysis that looks professional but is mathematically wrong. The result: decisions based on fabricated numbers.

What You Get

• Pull live prices and fundamentals for any US (NYSE/NASDAQ), HK (HKEX), or A-share (SSE/SZSE) ticker using Yahoo Finance and Sina Finance APIs with correct ticker format handling (sh600519, 0700.HK, AAPL)
• Calculate portfolio-level risk metrics — Sharpe ratio, Sortino ratio, max drawdown, VaR, beta, and correlation matrices from position-weighted return series
• Run full technical analysis with SMA/EMA crossovers, RSI divergence, MACD histograms, Bollinger Band breakouts, and volume anomaly detection (Z-score based)
• Scan options chains for put/call ratios, unusual activity (volume > 2× open interest), implied volatility surface, and Greeks estimation
• Screen stocks by multi-factor criteria — filter by P/E, revenue growth, volume thresholds, and market cap with automated alert triggers for gap moves
• Generate attribution analysis — decompose portfolio returns into alpha, beta, tracking error, and information ratio versus any benchmark

Why This Beats Prompting It Yourself

When you ask an agent to "analyze AAPL," it hallucinates the current P/E ratio because it doesn't have a data pipeline — it generates plausible-sounding numbers. This skill wires up yfinance and Sina Finance with proper error handling, caching awareness, and multi-market format detection. The agent writes and executes real Python, not fabricated analysis.

Example Session

You: Analyze my portfolio — AAPL, 0700.HK, sh600519, with weights 40%, 30%, 30%

Agent:

Portfolio Summary
   Total Return: 14.23%
   Annualized: 18.67%
   Sharpe Ratio: 1.42
   Max Drawdown: -8.31%
   Volatility: 13.15%

Technical Signals
   AAPL: BULLISH — RSI oversold at 28.4, Golden Cross forming
   0700.HK: NEUTRAL — Price at Bollinger midline, volume normal
   sh600519: BEARISH — Death Cross confirmed, volume spike (Z: 2.7)

Risk Attribution
   Alpha: +3.2% | Beta: 0.87 | Tracking Error: 4.1%
   Information Ratio: 0.78

Use Cases

• Morning market briefing: pull overnight HK/A-share moves before US pre-market
• Portfolio rebalancing: run Sharpe optimization across multi-market positions
• Earnings season screening: filter watchlist by P/E and volume before reports
• Options strategy setup: scan unusual activity before position entry
• Technical signal review: identify golden crosses and RSI divergences across watchlist

Known Limitations

Yahoo Finance rate-limits at ~2000 requests/hour and may return stale data for HK/A-share tickers. Sina Finance API requires Referer headers and may be inaccessible outside mainland China. Options data availability varies by exchange and may be delayed 15+ minutes.

Upgrade to Pro

Free covers data pulls and portfolio snapshots. Python Trading Toolkit Pro ($5) adds the strategy layer: walk-forward backtesting with cost modeling, Sharpe/drawdown/alpha attribution, position-sizing calculators, and rebalancing workflows. Upgrade when you start making position decisions from the output — Pro version.