This question lands in my inbox a lot. “Kim, can AI help me pick stocks?”
The answer is yes. And the research got impossible to argue with. Let’s start at Stanford.
Researchers at the Stanford Graduate School of Business built an AI analyst and turned it loose on 30 years of real market data. It analyzed 3,300 actively managed mutual funds from 1990 to 2020, adjusting portfolios once per quarter using only public information. Nothing fancy. No insider tips. Patterns everyone could see but nobody bothered to read.
📊 Let the data speak
The AI beat 93% of human fund managers. By an average of 600%. Really.
The team spent 12 months trying to find their own mistake. They couldn’t. Human managers generated $2.8 million in quarterly returns. The AI generated $17.1 million. Same market. Same data. Completely different outcomes.
Then Harvard landed a study last week. A neural network could predict 71% of all mutual fund trading decisions, whether a manager would buy, sell or hold a given stock. Nearly 3 out of 4 calls. By a machine.
Dalbar’s 2025 report found the average equity investor earned 8.5% less than the S&P 500 last year. Not from picking bad stocks but rather from bad human psychology. Loss aversion. Panic selling. Chasing trends.
AI doesn’t do any of that. It doesn’t watch CNBC at midnight and make decisions it regrets by morning.
💬 3 prompts to try right now
Open ChatGPT, Claude, Gemini or Grok and try these:
1. “Act as a financial adviser for a [your age]-year-old saving for retirement. Recommend a diversified portfolio with U.S. and international exposure, moderate risk tolerance and a 10-year horizon. Explain why each stock was selected.”
2. “You are a stock analyst. Compare the investment potential of [list companies], using earnings reports, P/E ratios, market trends and analyst sentiment. Summarize the pros and cons of each.”
3. “I’m a [conservative/aggressive] investor with [$X]. Suggest five stocks across different sectors for long-term growth. For each pick, give a one-paragraph summary of recent performance and future outlook.”
⚠️ Here’s the reality check
AI doesn’t know your mortgage, your timeline or your stomach for watching your balance drop 20% in a single rough week. It sounds confident even when it’s wrong. Think of it as the world’s most well-read research assistant. Not your financial adviser.
Use it to get smarter, not reckless. And not as a reason to skip talking to a real adviser before moving serious money.
I guess you could say AI raised the bar on your portfolio. Or is it raised the bear? Too bullish.
🗣️ TEXT/POST THIS STAT Stanford found AI beat 93% of fund managers by 600% over 30 years. Harvard just proved it can predict 71% of fund manager trades. The person managing your money should be paying attention. GetKim.com
📩 Send this to someone who has any money in the stock market or is saving for retirement. Use those links below.