Beating the Market with AI?: Finding Buffett-Style Bargains in 2025

Welcome back to AI in Investment Research and Finance, your trusted resource for blending cutting-edge Artificial Intelligence (AI) tools with time-honoured investment strategies.

Fundamental analysis remains the backbone of successful investing, providing vital insights into a company’s financial health, intrinsic value, and future potential. Now, with the rise of AI models, this essential process has evolved. AI-based tools enable investors to analyze vast datasets with unmatched speed and precision, potentially uncovering undervalued opportunities—stocks trading below their intrinsic value—and hidden gems in a fraction of the time compared to traditional methods.

Value stocks are enjoying renewed momentum in 2025. So far in the year, companies like JPMorgan Chase, IBM, and Ford have exceeded analysts' expectations, while the Vanguard Value ETF delivered a 4.8% return, outpacing the broader market's 3.5%. (Source: Barron's)

This week, we explore how AI tools, coupled with Warren Buffett's principles, can help you find undervalued stocks during earnings season. By prompting AI effectively, you can identify mismatches and hidden gems that align with Buffett's wisdom. “Price is what you pay, value is what you get.”

Warren Buffett's Philosophy on Value Investing

Warren Buffett, the legendary “Oracle of Omaha,” built his fortune by identifying undervalued companies with strong fundamentals and holding them for the long term. His approach revolves around:

1. Intrinsic Value: Buying stocks priced below their true worth.

2. Margin of Safety: Ensuring a cushion to protect against errors in valuation.

3. Long-Term Focus: Prioritizing quality businesses with sustainable growth potential.

Buffett’s philosophy reminds us that the market often undervalues companies during uncertain times, especially around earnings season. With AI, we can automate the process of identifying these opportunities by analyzing key financial metrics and market trends.

Two Effective Prompts for Finding Undervalued Stocks

Here are two prompts to help you uncover undervalued stocks during this earnings season:

Prompt 1: Earnings Report Valuation Check

“I want to analyze the most recent earnings reports of [Sector/Industry] companies listed on the S&P 500. My aim is to compare key valuation metrics, including Price-to-Earnings (P/E) ratio, Price-to-Book (P/B) ratio, and Enterprise Value to EBITDA (EV/EBITDA), to the industry median, using data from publicly available sources. Identify three S&P 500 companies in the [Sector/Industry] with valuation metrics below the industry median and earnings growth exceeding 10% over the last 12 months. For each company, include the company name, key valuation metrics, earnings growth rate, and a brief explanation of why the company may be undervalued based on its financial performance and market positioning. If direct data is unavailable, identify three companies with [revenue growth above 5%] based on recent earnings trends. Where possible, include references to validate findings."

Why it works:
This prompt helps you target undervalued stocks by comparing their current valuations with historical benchmarks while factoring in earnings growth—aligning with Buffett’s principles.

Prompt 2: Intrinsic Value vs. Market Price

“Evaluate [Company Name] stock using a discounted cash flow (DCF) model to estimate its intrinsic value. Compare this intrinsic value to the current market price and determine whether the stock is trading at a discount or premium. Highlight the key assumptions driving the valuation, such as revenue growth, discount rate, and terminal value. Additionally, summarize any risks or qualitative factors that may impact the DCF assumptions or valuation outcome. Provide references where available."

Why it works:
This prompt applies a fundamental valuation approach, such as DCF, to determine whether a stock is undervalued compared to its intrinsic value, giving you actionable insights.

Pro Tips Corner: Automate Your Stock Research

Finding undervalued stocks can be time-intensive, but ChatGPT’s Tasks feature simplifies the process with scheduled prompts.

Note: The "GPT-4 with Scheduled Tasks" feature is in beta and available to Plus, Pro, and Team subscribers. Free-tier users currently do not have access.

Here’s how you can automate your stock research:

1. Select the Right Model: Open ChatGPT and choose “GPT-4 with Scheduled Tasks” in the dropdown menu.

2. Set Up Scheduled Reports: Prompt the AI to deliver a daily analysis of undervalued stocks in specific sectors at a specific time.

3. Incorporate Earnings Season Data: Request an analysis of recent earnings surprises, focusing on companies with strong fundamentals and low valuation multiples.

4. Highlight Opportunities and Risks: Ask for an overview of market sentiment and potential red flags for undervalued stocks.

Prompt Example:

“Every morning at 6:30 AM EST, provide a concise analysis of the top 3 undervalued stocks within the [Sector, e.g., technology, energy, etc.] from major market indices [e.g. S&P 500, Nasdaq 100, Dow Jones Industrial Average (DJIA), etc.]. Identify top undervalued companies based on strong fundamentals and low valuation multiples, such as P/E and P/B ratios. Include recent earnings insights, highlighting companies with notable earnings surprises or forecast upgrades. Summarize the potential growth opportunities, market sentiment, and any associated risks (Beta) that may affect these stocks. Incorporate relevant metrics like revenue growth, profit margins, and debt levels to ensure the report is both actionable and insightful, presenting the most compelling trends and emerging opportunities in the market.”

Pro Tip: Always validate AI-driven insights with your own analysis, using tools like SEC Edgar, and cross-referencing with reputable financial news sources to ensure accuracy and avoid over-reliance on automated recommendations.

The Bottom Line

Uncovering undervalued stocks is a cornerstone of value investing. By combining Buffett’s principles with the analytical power of AI, you can streamline your research process and make more informed decisions during earnings season.

Try using these AI prompts this week and share your findings with us! Your strategies might inspire our next newsletter!

(Please add our email address to your contact list so that the next newsletter goes into your Inbox folder: [email protected])

This Week in History:

📌 February 19, 1946 – Alan Turing proposed the design for one of the first stored-program computers.
📌 February 19, 1986 – The Soviet Union launched Mir, the first modular space station, which remained in orbit for 15 years.
📌 February 25, 1959 – Demonstration of the Automatically Programmed Tool (APT) language, an early innovation in automation.