How to Prompt AI for Financial Advice

by | Jul 17, 2025 | The Doctor's Dose

How to Prompt AI for Better Financial Advice: What Investors Need to Know

Artificial intelligence is reshaping the way we research stocks, analyze sectors, and make investment decisions. However, when it comes to prompting AI for financial advice, many retail investors are still learning that how you ask is just as important as what you ask.

In recent months, there’s been a surge in people turning to chatbots like ChatGPT, Claude, and Gemini to help with financial queries—from building portfolios to analyzing earnings reports. According to Sensor Tower data, questions related to economics, finance, and taxes accounted for 13% of all chatbot prompts between March and April 2025, up from 4% a year earlier.

The takeaway? More people are leaning on AI for money decisions—but not all are getting answers they can trust.

Why Prompting Matters

Large language models (LLMs), like the ones behind ChatGPT, are incredibly powerful—but they aren’t mind readers. They work best when users provide layered, specific instructions. Think of prompting like giving directions to an intern: the more clarity and structure you provide, the better the results.

“LLMs are very smart, but they need the right context,” says Chris Ackerson, SVP of Product at AlphaSense, an AI platform used by hedge funds and institutional investors. “You have to walk them through what you want.”

AlphaSense, for example, customizes public LLMs with proprietary data—like SEC filings, earnings call transcripts, and sell-side research—to deliver deeper insights. Retail investors can’t fully replicate that, but with the right prompting techniques, they can come close.

A Five-Step Prompting Process for Investors

Dave Wang, a former fund manager turned retail trader and AI researcher, shared his approach to prompting chatbots for financial analysis. Here’s his five-part method for getting more accurate, actionable answers:

1. Assign a Persona

Begin by telling the AI what kind of expert to be. For example:

“Act as a senior equity research analyst covering the technology sector.”

This helps the model prioritize relevant data and tailor the tone and depth of the answer.

2. Define the Task

Be clear about what you’re asking the AI to do. For instance:

“Analyze the financial health of Apple (AAPL) using its most recent 10-K filing and Q2 earnings report.”

3. Provide Context

Add relevant data or a brief background to guide the model. Example:

“Apple just posted stronger-than-expected iPhone sales, but services revenue missed. The stock is up 5% after earnings.”

4. Request Specific Output

Guide the format and type of analysis you want:

“Summarize in bullet points the growth trends, valuation risks, and competitive advantages.”

5. Ask for Validation

Prompt the AI to list sources, offer citations, or explain how it reached its conclusions:

“Reference specific line items from Apple’s financial statements or earnings call transcript.”

This helps identify whether the response is grounded in verifiable data—or if it’s just plausible-sounding fluff.

Enterprise AI vs. Public Chatbots

Institutional investors often use enterprise-grade AI tools that integrate proprietary databases and have stricter data validation protocols. These tools—like those from OpenAI, Anthropic, or AlphaSense—cost more but offer added layers of control, privacy, and performance.

Retail investors using public versions of AI should compensate by:

  • Double-checking facts.

  • Cross-referencing results with primary sources.

  • Using AI to speed up research—not replace it.

Where It’s Going

According to a Natixis survey of 520 investment professionals, 58% of firms already use AI in their investment processes. As AI grows more advanced, prompting will become an essential skill—much like reading a balance sheet or spotting technical setups on a chart.

Final Takeaway

AI isn’t a crystal ball. It’s a tool—and one that requires direction. When used properly, it can speed up analysis, surface ideas, and give traders and investors a competitive edge.

But if your only prompt is: “What stocks should I buy?”—don’t be surprised if the results miss the mark.

📈 Want to see how we use AI-powered research in real-time trading? Join our daily War Room sessions at TraderInsight.com/live-stream.