How Chatgpt Can Empower Stock Market Research

How ChatGPT Can Empower Stock Market Research is transforming the way investors analyze markets, identify opportunities, and make data-driven decisions in real time. By leveraging advanced artificial intelligence, ChatGPT can quickly process vast amounts of financial data, summarize earnings reports, interpret market trends, and even generate actionable insights that would traditionally take hours of manual research. Whether you’re a beginner trying to understand key metrics or an experienced trader refining your strategy, ChatGPT enhances efficiency, reduces information overload, and helps uncover hidden patterns across stocks, sectors, and global markets—making it an indispensable tool for smarter, faster, and more informed stock market research.

1. The Stock Market Research Problem Nobody Talks About Enough

Let me paint you a picture. Every single day, thousands of news articles, earnings calls, SEC filings, analyst reports, Federal Reserve press releases, and social media posts are published about the stock market. You, as an individual investor or trader, are supposed to read all of that, synthesise it, compare it to historical data, factor in macroeconomic conditions, and then make a calm, rational decision about whether to buy or sell.

That’s basically asking you to be a superhero. Except instead of a cape, you’re wearing pyjamas. And instead of superpowers, you have a laptop that overheats.

The information overload problem is real and documented. The average retail investor is working at a structural disadvantage compared to institutional players who have entire research departments, Bloomberg terminals, and quant teams running automated analysis 24/7. For most of us, the idea of processing a fraction of the available market data is laughable. I once tried reading an entire 10-K annual report for a single company. I made it to page twelve before I started questioning every life decision I’d ever made.

This is precisely where ChatGPT and large language models (LLMs) are genuinely revolutionary for the everyday trader and investor. They don’t sleep. They don’t get emotional. They don’t panic-sell because their ex’s wedding announcement coincided with a market dip. They just process information, and they’re extraordinarily good at it.

2. The Academic Evidence: This Isn’t Just Hype

Look, I know what you’re thinking. “Another article telling me AI is going to change everything.” I hear you. I was sceptical too — until I started reading the actual peer-reviewed research. And what these papers reveal is genuinely exciting.

2.1 ChatGPT Can Predict Stock Price Movements

In one of the most widely cited papers in this space, researchers Alejandro Lopez-Lira and Yuehua Tang published groundbreaking findings in April 2023. Their paper, “Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Models”, available at https://arxiv.org/abs/2304.07619, demonstrated that ChatGPT could predict daily stock returns using news headlines — even without any direct financial training.

Their methodology involved feeding ChatGPT news headlines about U.S. common stocks listed on NYSE, NASDAQ, and AMEX, and asking it to classify each headline as positive, negative, or neutral. The results were staggering. A self-financing daily-rebalanced strategy that bought stocks with positive ChatGPT-4 recommendations and sold stocks with negative recommendations earned a daily average return of 38 basis points pre-transaction costs, compounding to a cumulative return of over 650% from October 2021 to December 2023. The strategy’s Sharpe ratio using ChatGPT-4 reached 3.8, compared to just 3.1 for ChatGPT-3.5, confirming that more sophisticated models deliver stronger predictive power.

Now, before you run off and quit your job, let me be very clear: this does not mean ChatGPT is a money-printing machine. Transaction costs, real-world slippage, and the fact that widespread adoption of these strategies reduces their edge over time are all serious considerations. The same paper noted that the annualised Sharpe ratio dropped from 6.54 in Q4 2021 to 2.33 in 2023, precisely because more investors started using similar tools. The market, as always, adapts. But the baseline signal is powerful and real.

2.2 ChatGPT and NASDAQ Sentiment

Researchers Lefort, Benhamou, and colleagues (2024) published work on “Sentiment Analysis of Bloomberg Markets Wrap Using ChatGPT: Application to the NASDAQ”, available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4780150. They developed a two-stage method where ChatGPT first identified key financial headlines from Bloomberg daily market summaries, then predicted whether each headline would lead to a rise, fall, or indecision in stock prices. By accumulating and detrending the resulting sentiment scores, they generated a practical NLP-driven investment strategy for the NASDAQ that demonstrably outperformed baseline approaches. The research confirmed a statistically significant positive correlation between ChatGPT’s sentiment scores and future short-to-medium-term equity market returns.

Think about what that means for a second. You’re a retail trader sitting at home, and you can now deploy a system that processes financial news headlines and generates market sentiment signals in seconds — the same type of analysis that hedge funds used to pay teams of analysts to approximate over weeks.

I’m not saying the playing field is level. But it just got a lot less uneven.

2.3 ChatGPT vs. DeepSeek: Who Predicts Markets Better?

A fascinating study by Chen, Tang, Zhou, and Zhu (2023, updated 2025), “ChatGPT and DeepSeek: Can They Predict the Stock Market and Macroeconomy?”, published as Olin Business School Research Paper No. 2023/18 and available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4660148, compared multiple large language models on their ability to extract predictive information from Wall Street Journal articles. Their findings revealed that ChatGPT had genuine predictive power for stock market returns, driven primarily by investors’ underreaction to positive news during periods of economic downturn and high uncertainty. DeepSeek underperformed ChatGPT, which the researchers attributed to ChatGPT’s more extensive training in English financial language. This study highlights something traders should keep in mind: not all AI models are created equal for financial research, and ChatGPT currently holds a meaningful edge in English-language market analysis.

2.4 ChatGPT-4o for Financial Data Analysis

A 2025 paper published in the Journal of Risk and Financial Management by researchers examining ChatGPT-4o’s capabilities (available at https://www.mdpi.com/1911-8074/18/2/99) found that the model’s performance in financial data analysis — including time series analysis, risk and return analysis, and ARMA-GARCH modelling — was generally comparable to traditional statistical software like Stata. For traders who don’t have a quantitative finance background, this is like being handed a PhD-level statistician who works for free and never complains about the hours. The researchers concluded that integrating ChatGPT-4o into financial research could lead to more efficient data processing and better-informed investment decisions. Honestly, at this point, the only reason NOT to use it is stubbornness. Which, by the way, I’ve been guilty of. I once spent three hours trying to manually calculate a moving average in a spreadsheet when I could have just… asked ChatGPT to do it.

3. Seven Powerful Ways ChatGPT Empowers Stock Market Research

Let me walk you through the specific, practical applications that can transform your research process. Each one represents hours of saved time and a meaningful upgrade to the quality of your analysis.

3.1 Earnings Call Analysis and Summarisation

Every quarter, publicly listed companies hold earnings calls. These are dense, jargon-heavy, and often deliberately vague — corporate speak is a whole language, and I still don’t know what “we’re cautiously optimistic about synergistic headwinds” means. I’ve read that sentence three times. I still don’t know.

ChatGPT can read an entire earnings call transcript and deliver a concise summary of the key financial metrics discussed, management tone, forward guidance, and risk factors — in seconds. You can ask it pointed questions: “Did management sound defensive about their debt levels?” or “What did the CFO say about Q3 inventory?” or “Compare this quarter’s guidance language to last quarter.” The model picks up on nuances in language that traditional keyword-based sentiment tools miss entirely.

This aligns with work by the CAIA Institute’s comparative NLP analysis (2024) available at https://caia.org/blog/2024/03/12/comparative-analysis-nlp-approaches-chatgpt-edition, which tested ChatGPT against other NLP approaches like FinBERT and Loughran-McDonald on earnings calls. ChatGPT showed the highest correlation with FinBERT in classifying sentiment at the aggregate security level, and the research confirmed that sentiment analysis on earnings calls can generate alpha — outperformance not explained by traditional risk and return factors.

3.2 Financial News Sentiment Analysis

We’ve touched on this in the academic section, but let’s get practical. Every morning before markets open, you can paste ten to fifteen news articles about your watchlist companies into ChatGPT and ask: “Rank these by sentiment impact on stock price, from most positive to most negative. Explain your reasoning.” Within seconds, you have a prioritised reading list and a preliminary view on how markets might react.

The research by Fatouros et al. (2023), “Transforming Sentiment Analysis in the Financial Domain with ChatGPT”, published in Machine Learning with Applications (Vol. 14, 2023) and available at https://www.sciencedirect.com/science/article/pii/S2666827023000610, confirmed that ChatGPT-based sentiment analysis meaningfully outperforms earlier dictionary-based approaches in the financial domain. Older tools like the Loughran-McDonald lexicon were built for accounting filings and miss a lot of contextual meaning. ChatGPT understands context. It knows that “significant restructuring charges” is not a phrase that sends stock prices upward.

3.3 Fundamental Analysis Support

Want to understand a company’s business model, competitive moat, balance sheet health, or debt structure? ChatGPT can help you build a fundamental analysis framework in minutes. You can paste in financial statements and ask for key ratio calculations, trend analysis, and red flag identification. You can ask, “What are the biggest financial risks for a company with this level of debt-to-equity in a rising interest rate environment?” and receive a nuanced, well-reasoned answer.

Now, I want to be honest here. ChatGPT doesn’t have access to live financial data unless connected to external tools. You still need to pull the numbers from databases like Yahoo Finance, Macrotrends, or your broker’s platform. ChatGPT is the analyst — you supply the data. Think of it like a really smart intern who can do everything except remember to check their email. You give them the information; they do something brilliant with it.

3.4 SEC Filing and Annual Report Breakdown

10-K filings. 10-Q filings. Proxy statements. These documents can run to hundreds of pages of legalese so thick it reads like a terms and conditions agreement written by a committee of robots. I once tried to read a complete 10-K on my own. By page forty, I had forgotten what company I was researching. My eyes had glazed over. My soul had left my body.

ChatGPT can read these documents, identify the most material disclosures, flag unusual language in the risk factors section, extract management discussion and analysis insights, and summarise related-party transactions. If a company has buried a litigation risk or a change in accounting methodology in footnote thirty-seven, ChatGPT is the one tool that will find it and explain why it matters — in plain English.

3.5 Portfolio Screening and Stock Comparison

“Compare the revenue growth trajectories of Microsoft, Salesforce, and ServiceNow over the past three years and explain which has the strongest moat based on their competitive positioning.” That’s a question you could spend all afternoon researching manually. ChatGPT can synthesise your pre-gathered data and deliver a comparative assessment in moments. You can build screening criteria — “I’m looking for mid-cap tech companies with strong free cash flow, low debt, and exposure to AI infrastructure” — and ChatGPT will help you systematically work through whether candidate companies meet your criteria.

Research published in Finance Research Letters (referenced in the ScienceDirect portfolio management study at https://www.sciencedirect.com/science/article/pii/S1544612323011583) found that ChatGPT demonstrates superior financial advice capabilities for portfolio construction compared to traditional robo-advisors when investor risk appetite is factored in. Specifically, Oehler and Horn (2024) positioned generative AI as a significant advancement in personalised investment guidance.

3.6 Macroeconomic Research and Interpretation

When the Federal Reserve speaks, markets move. But parsing the Fed’s language is genuinely difficult. They speak in a dialect that normal humans were not designed to process. “The Committee seeks to achieve maximum employment and inflation at the rate of 2 percent over the longer run” — I know what those words mean individually. Together, though? Together they become a Rorschach test for economists.

Smales (2023) published research examining ChatGPT’s ability to interpret and respond to monetary policy announcements, revealing its potential to improve market efficiency by making central bank communication more accessible. The work is cited extensively in PMC’s review of LLMs in equity markets at https://pmc.ncbi.nlm.nih.gov/articles/PMC12421730/. In practical terms, you can paste a Fed statement into ChatGPT and ask: “How does this compare to the previous three statements? Is this more hawkish or dovish? What are the implications for rate-sensitive sectors?” You’ll get a professional-grade interpretation in seconds. It’s like having a former Fed economist in your pocket — one who doesn’t charge consulting fees and won’t judge you for not knowing what “quantitative tightening” means.

3.7 Risk Assessment and Scenario Planning

“What are the main risks facing the semiconductor sector if U.S.-China trade tensions escalate further?” “How would a 100 basis point rate hike affect the real estate investment trust sector?” “Build me a bear case, base case, and bull case for this position.” These are scenario planning questions that ChatGPT handles brilliantly. It synthesises geopolitical, macroeconomic, sector-specific, and company-level risks into coherent frameworks — a task that would normally require hours of reading and analysis.

4. Case Studies: Real-World Applications

Theory is great. Evidence is better. Let’s look at specific use cases that illustrate how traders and researchers are leveraging ChatGPT in practice.

Case Study 1: The Retail Trader Who Outperformed His Gut

Marcus, a part-time trader and full-time accountant based in London, had been investing in individual stocks for six years with moderate results. His process was largely intuition-based — he read the financial news, watched a few YouTube channels, and made calls based on vibes. After incorporating ChatGPT into his workflow in late 2023, he developed a structured process: every Sunday evening, he feeds ChatGPT the week’s most significant news articles for his ten watchlist stocks, asks for a sentiment summary and key risk flags, and uses the output to plan his entry and exit decisions for the week ahead. In his own words: “It’s like having a research associate who did a finance degree. Except they never get tired and never tell me my ideas are bad — they just show me why they might be.”

In the twelve months following his process change, Marcus reported a meaningful improvement in his decision consistency and a reduction in reactive trading. He stopped buying stocks because of a headline he half-read at 11 p.m. That alone, he says, was worth everything.

Case Study 2: A Hedge Fund Research Team Saves 200 Hours Monthly

A mid-sized European hedge fund — which asked not to be named — integrated ChatGPT-based tools into their earnings analysis workflow in early 2024. Their research team was previously spending approximately forty hours per analyst per month reading and summarising earnings call transcripts for their coverage universe of eighty companies. After deploying a ChatGPT-powered system that automatically summarised each transcript and flagged management language deviations from prior quarters, the team reduced that time to under five hours per analyst monthly. The saved time was redirected toward higher-value analysis — building financial models, conducting channel checks, and developing proprietary investment theses. The fund’s research director described the tool as “the most impactful change to our process in a decade.”

Now, I want to make a joke here about how the analysts whose jobs got easier are probably still complaining, but I’ll let that one go. Barely.

Case Study 3: Sentiment-Driven NASDAQ Strategy

This one comes directly from academic research rather than an anecdotal case, and it’s worth dwelling on. Lefort et al. (2024), in their SSRN working paper https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4780150, constructed a systematic strategy using ChatGPT’s analysis of Bloomberg Markets Wrap daily news summaries. By asking ChatGPT to extract headlines and then classify each headline’s directional impact on stock prices — bullish, bearish, or neutral — they built a daily sentiment score. After detrending and accumulating this score, they translated it into a trading signal for the NASDAQ index. The resulting strategy demonstrated improved Sharpe ratios and reduced maximum drawdowns compared to strategies that ignored news sentiment entirely. The findings were validated across multiple equity markets including the S&P 500, Nikkei, and Euro Stoxx, confirming the approach generalises globally.

For the average trader, the lesson here is simple: news sentiment matters, and ChatGPT can help you process it at scale, faster and more consistently than any human analyst.

Case Study 4: The Solo Investor Who Built a Research Process From Scratch

Priya, a software engineer in Birmingham with no formal finance training, wanted to start investing seriously but felt overwhelmed by the complexity of stock market research. She had tried reading investing books, watching tutorials, and using stock screeners — but she always felt like she was missing context. She started using ChatGPT as a teacher first: asking it to explain P/E ratios, free cash flow, capital allocation frameworks, and sector dynamics. Then she graduated to using it as an analyst: pasting in company filings, asking for ratio analyses, and requesting plain-English explanations of anything she didn’t understand. Within eight months, she had developed a coherent, repeatable research process for analysing companies before investing. “I didn’t go to business school,” she says. “ChatGPT was my business school. Except cheaper. And it never gave me homework.”

Honestly, Priya is out here winning. The rest of us paid tuition.

5. How to Use ChatGPT Effectively for Stock Market Research

Having the tool is one thing. Using it well is another. I have made every possible mistake, so you don’t have to. Let me summarise what actually works.

5.1 Be Specific and Detailed in Your Prompts

The quality of ChatGPT’s output is directly proportional to the quality of your prompt. Vague questions get vague answers. “Tell me about Apple” will give you a Wikipedia-style summary. “Analyse Apple’s gross margin trajectory over the last four quarters, identify the key drivers of any change, and assess whether these trends are sustainable given their current product mix and competitive dynamics in the premium smartphone market” — that gives you something useful. Think like an analyst when you ask questions.

5.2 Provide Context and Data

Remember: ChatGPT’s training data has a cutoff and it doesn’t have live market access by default. If you want current analysis, you need to bring the current data. Paste in the most recent earnings report. Copy the most recent news articles. Share the key financial metrics. The more relevant context you provide, the more targeted and accurate the analysis will be. ChatGPT is brilliant at synthesis. You need to give it something to synthesise.

5.3 Use It Iteratively

The best research sessions with ChatGPT are conversations, not single questions. Start with a broad question, get an answer, then drill down into the most interesting or concerning points. “You mentioned their debt maturity profile could be an issue — can you elaborate on what scenarios would make that a serious risk?” Follow the thread. Push back. Ask for counter-arguments. This iterative process generates genuinely deep analysis that would take hours to replicate through traditional research methods.

5.4 Verify Important Claims

ChatGPT can occasionally confuse data points or reference figures that need to be double-checked against primary sources. For any analysis that will drive a real investment decision, always verify key numbers against original filings, official databases, or your broker’s research tools. Use ChatGPT as your analytical engine — but keep your own verification process for the facts it builds on. Consider it a first draft, not a signed document. I once had ChatGPT give me a revenue figure that was off by a rounding convention. My own fault for not checking. Lesson learned. Painfully. But lesson learned.

5.5 Use It for Continuous Learning

Beyond research, ChatGPT is an extraordinary financial education tool. Ask it to teach you about concepts you don’t fully understand. Ask it to explain the mechanics of options pricing, or why inverted yield curves historically precede recessions, or how to read a cash flow statement. The faster you develop your own financial literacy, the more effectively you’ll be able to use and verify the AI’s output. The goal is to become a better investor — not just to outsource your thinking.

6. The Honest Limitations: What ChatGPT Cannot Do

I would be doing you a disservice if I only told you the good stuff. ChatGPT is extraordinary but it is not omniscient. And as someone who has made enough trading mistakes to write a separate article entirely, I respect the importance of knowing what your tools can’t do.

It doesn’t have live data. Unless you’re using a version connected to real-time financial data tools, ChatGPT’s knowledge has a training cutoff. It cannot tell you what Apple’s stock price is right now, or what the Fed said in this morning’s press conference. You need to bring current data to it.

It can’t predict the future. The academic research showing predictive power is based on probabilistic signals derived from information processing — not clairvoyance. The market can and does do irrational things. ChatGPT doesn’t know that a CEO is about to resign for personal reasons, or that a geopolitical event will spike oil prices overnight. Nobody does.

It can occasionally hallucinate. In AI language modelling, “hallucination” refers to the model generating plausible-sounding but factually incorrect information. For financial research, this is a genuine risk. Always verify specific data points, figures, and citations from primary sources. Treat ChatGPT output the way you’d treat the first draft of an analyst report: intelligent and useful as a starting point, but requiring review before you act on it.

It is not personalised financial advice. ChatGPT can inform your research process. It cannot know your personal financial situation, tax position, risk tolerance, investment horizon, or the myriad other factors that should drive individual investment decisions. For personalised advice, consult a qualified financial adviser. Nothing in this article — and nothing ChatGPT tells you — constitutes personalised investment advice.

7. The Future of AI in Stock Market Research

The trajectory here is clear. The research community is producing dozens of papers every year exploring LLM applications in finance. BloombergGPT, a large language model specifically trained on financial data, was developed and published by Bloomberg LP’s AI team (Wu et al., 2023) and represents the logical next step: LLMs trained not just on general language but on decades of proprietary financial text. As these specialised models mature, their analytical power for stock market research will only grow.

The PMC review of LLMs in equity markets at https://pmc.ncbi.nlm.nih.gov/articles/PMC12421730/ documents an accelerating wave of research across multi-agent trading frameworks, reinforcement learning-based portfolio management, and LLM-driven alpha generation strategies. This is not a passing trend. This is the new infrastructure of financial research, and retail investors who learn to work with these tools effectively will have a structural advantage over those who don’t.

The study by Lopez-Lira and Tang also made a profoundly important point: widespread adoption of LLMs can enhance market efficiency overall. As more market participants use AI to process information faster and more accurately, the market’s ability to price in news improves. The edge from any individual AI-based signal will diminish over time — but the baseline quality of investment decision-making across the market will rise. In that sense, ChatGPT isn’t just empowering individual traders. It’s making markets better.

And look — I know that sounds like the kind of thing an AI company would put in a press release. But the evidence supports it. And unlike my gut feeling about that barbecue stock tip, this time the evidence is peer-reviewed.

8. A Practical Daily Research Workflow Using ChatGPT

Here’s a concrete morning routine I use that takes about thirty to forty-five minutes and covers the key bases:

Step 1 — Market Briefing (5 minutes). Paste the three to five most significant overnight financial news stories into ChatGPT and ask: “Summarise the key market themes from these articles and identify which sectors or individual stocks are most likely to be affected today, and in what direction.” You now have a prioritised agenda for the morning.

Step 2 — Watchlist News Scan (10 minutes). For any companies you hold or are watching, paste relevant headlines and ask for a quick sentiment assessment. Flag anything that warrants deeper investigation.

Step 3 — Deep Dive on Any New Developments (15 minutes). If there’s an earnings release, major corporate announcement, or economic data release relevant to your positions, use this time to paste the full text and ask for a comprehensive breakdown. What surprised? What was expected? What do you need to monitor going forward?

Step 4 — Risk Check (5 minutes). Ask ChatGPT to identify the two or three most significant tail risks to your current positions given the week’s news flow. Not because you’ll act on all of them — but because knowing your risks is the foundation of surviving long enough in this game to make money.

Step 5 — Learning (10 minutes, optional but recommended). Use any concept you encountered during research that you didn’t fully understand as a learning prompt. “Explain how rising Treasury yields affect tech stock valuations and the mechanism behind that relationship.” Compound your knowledge every single day. It adds up.

9. Conclusion: Your Competitive Edge Is Sitting Right There

Here’s the truth about stock market research: most retail investors are working harder than they need to and smarter than they think they can. The information has always been out there. The analysis framework has always been learnable. What was missing was the tool to make it all accessible, affordable, and fast.

ChatGPT for stock market research is not a magic formula. It will not tell you what to buy tomorrow. It will not guarantee returns. If you’re looking for a sure thing, I’m sorry — the stock market doesn’t do those. I’ve looked everywhere. They’re not there. Trust me.

But ChatGPT will dramatically improve the quality and depth of your research. It will help you read earnings calls like an analyst, process news sentiment like a quant, understand SEC filings like a lawyer, and assess macroeconomic risk like an economist — all without the four degrees those professionals spent years earning. It is, genuinely, one of the most democratising tools ever introduced into the world of investing.

The academic evidence is clear. Lopez-Lira and Tang showed it can predict stock returns. Lefort et al. showed it can generate tradeable NASDAQ signals. Chen et al. showed it outperforms competing models for market prediction. Fatouros et al. showed it transforms financial sentiment analysis. The research community is in broad agreement: LLMs like ChatGPT represent a genuine advancement in financial research capability.

You’ve got the tool. You’ve got the evidence. You’ve got a workflow you can start tomorrow morning. The only thing left is to actually use it. Because the traders who are sleeping on this right now? They’re going to look back in three years the same way people look back at not buying index funds in 2010.

And that is not a position you want to be in. I know. Because I’ve been in that position. Multiple times. About multiple things. Don’t be me. Be the version of you who read this article, opened ChatGPT, and started researching like a professional.

Your portfolio will thank you. Your 2 a.m. stress sessions will thank you. And frankly, your gut feeling will thank you too — because once it’s been trained by actually good research, it’ll start being right more often.

Now go make some money. And maybe read a 10-K while you’re at it. Just… let ChatGPT read it first.

References

Lopez-Lira, A., & Tang, Y. (2023, updated 2024). Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Models. arXiv. https://arxiv.org/abs/2304.07619
Lefort, B., Benhamou, E., Ohana, J-J., Saltiel, D., Guez, B., & Jacquot, T. (2024). Sentiment Analysis of Bloomberg Markets Wrap Using ChatGPT: Application to the NASDAQ. SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4780150
Chen, J., Tang, G., Zhou, G., & Zhu, W. (2023). ChatGPT and DeepSeek: Can They Predict the Stock Market and Macroeconomy? Olin Business School Center for Finance & Accounting Research Paper No. 2023/18. SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4660148
Fatouros, G., Soldatos, J., Kouroumali, K., Makridis, G., & Kyriazis, D. (2023). Transforming Sentiment Analysis in the Financial Domain with ChatGPT. Machine Learning with Applications, Vol. 14. https://www.sciencedirect.com/science/article/pii/S2666827023000610
Oehler, A., & Horn, M. (2024). ChatGPT provides superior financial advice for one-time investments compared to robo-advisors. Referenced in ScienceDirect: https://www.sciencedirect.com/science/article/pii/S1544612323011583
CAIA Institute. (2024). Comparative Analysis of NLP Approaches – ChatGPT Edition. Portfolio for the Future. https://caia.org/blog/2024/03/12/comparative-analysis-nlp-approaches-chatgpt-edition
PMC Review. (2025). Large Language Models in Equity Markets: Applications, Techniques, and Insights. National Library of Medicine. https://pmc.ncbi.nlm.nih.gov/articles/PMC12421730/
MDPI Journal of Risk and Financial Management. (2025). A First Look at Financial Data Analysis Using ChatGPT-4o. https://www.mdpi.com/1911-8074/18/2/99
Lefort, B., Benhamou, E., et al. (2024). Can ChatGPT Compute Trustworthy Sentiment Scores from Bloomberg Market Wraps? arXiv. https://arxiv.org/html/2401.05447v1
Lefort, B., et al. (2024). Stress Index Strategy Enhanced with Financial News Sentiment Analysis for Equity Markets. arXiv. https://arxiv.org/html/2404.00012v1

One last thing — if you found this article useful, share it with a fellow trader. Because the best thing about levelling up your research game is watching your whole community get better at the same time. Rising tides, ships, all that good stuff.

Disclaimer: This article is for informational and educational purposes only. Nothing contained herein constitutes personalised financial or investment advice. All investment activities carry risk. Past performance of any strategy, including AI-assisted approaches, does not guarantee future results. Always consult a qualified financial adviser before making investment decisions.

Market Investigation