AI Prompts for Trading Analysis

How to Use These Prompts

Copy each prompt and paste it into Claude, ChatGPT, or your preferred AI platform. Replace the bracketed placeholders [TICKER], [DATA], etc. with your actual data. For best results, always include the raw data (earnings transcript, price data, financial statements) rather than asking the AI to recall it from memory.

Earnings & Fundamental Analysis

Prompt: Earnings Call Analysis

Best on: Claude (long context handles full transcripts)

I'm pasting the full earnings call transcript for [TICKER] from [QUARTER/YEAR]. Please analyse it and provide:

1. Three most important forward-looking statements from management
2. Any guidance changes compared to the previous quarter (if I provide that too)
3. Key risks or concerns raised by analysts during Q&A
4. Tone assessment: is management more confident or cautious vs last quarter?
5. One thing management didn't address that they should have

[PASTE FULL TRANSCRIPT HERE]

Prompt: Financial Health Check

Best on: Claude or ChatGPT

Here are the last 4 quarters of financial data for [TICKER]:

[PASTE INCOME STATEMENT / BALANCE SHEET / CASH FLOW DATA]

Analyse the trend in: revenue growth, margins (gross and operating), free cash flow, debt-to-equity, and current ratio. Flag anything that is deteriorating or improving significantly. Are there any red flags that a fundamental investor should investigate further?

Strategy & Backtesting

Prompt: Backtest a Trend Following Strategy

Best on: Claude (code quality) or ChatGPT Code Interpreter (instant execution)

Write a Python script that backtests a trend following strategy on [TICKER] using the following rules:

- Entry: Buy when the 50-day EMA crosses above the 200-day EMA
- Exit: Sell when the 50-day EMA crosses below the 200-day EMA
- Position size: 100% of capital (simple version)
- Data: Use yfinance to pull daily data from 2016 to present

Output: total return, max drawdown, Sharpe ratio, win rate, and an equity curve chart. Compare against buy-and-hold over the same period.

Prompt: Turtle Trading Backtest

Best on: Claude Code (can iterate and refine the script)

Write a Python script that implements the Turtle Trading System 1 rules:

- Entry: Buy on 20-day Donchian channel breakout (new 20-day high)
- Exit: Exit on 10-day low
- Position sizing: Risk 1% of equity per trade, using 20-day ATR
- Pyramiding: Add up to 3 units at 0.5 ATR intervals above entry
- Test on: SPY, QQQ, and GLD from 2010 to present using yfinance

Output: per-symbol results, combined portfolio return, max drawdown, and equity curve. Also show the distribution of trade returns (histogram).

Risk & Position Analysis

Prompt: Devil's Advocate

Best on: Claude (careful, structured reasoning)

I'm considering going long [TICKER] at [PRICE]. My thesis is: [DESCRIBE YOUR THESIS IN 2-3 SENTENCES].

Please argue against this position as aggressively as possible. What are the three strongest reasons this trade could fail? What market conditions would invalidate my thesis? What am I most likely overlooking? Be specific — I want to stress-test this, not get reassurance.

Prompt: Portfolio Risk Scenario

Best on: Claude

Here is my current portfolio:

[LIST HOLDINGS: ticker, weight, sector]

Analyse three scenarios:
1. A 20% market correction over 3 months (broad risk-off)
2. Interest rates rise 100bps over 6 months
3. A sector rotation from growth to value

For each scenario: which positions are most at risk, which might benefit, and what hedges or adjustments would reduce overall portfolio drawdown?

Market Research

Prompt: Market Regime Assessment