Can ChatGPT Beat the Market?

Artificial intelligence has quickly become one of the biggest themes in investing, with traders increasingly using tools such as ChatGPT to analyse companies, interpret financial news and generate trading ideas.

But can ChatGPT actually pick winning stocks and beat the market?

That is exactly what we are going to find out in this live test.

Rather than asking ChatGPT for theoretical stock picks and then looking back to see how they would have performed, we are going to put its selections to the test in real time.

Every weekend, we will ask ChatGPT to select its “best trade of the week” from stocks listed on US exchanges.

There are a few simple rules.

The stock must have a market capitalisation of at least $500 million, which should eliminate very small and illiquid penny stocks. The selection must also be a normal US-listed common stock with sufficient liquidity for an ordinary trader to realistically buy and sell.

Most importantly, ChatGPT must make its selection before the market opens for the week.

Once the stock has been chosen, the selection is locked.

We will then record the stock’s official opening price on Monday — or the first trading day of the week if Monday is a US market holiday — and compare it with its official closing price on Friday.

That gives us a very simple measure of performance:

Monday opening price → Friday closing price

There is no opportunity for ChatGPT to change its mind halfway through the week if the trade starts going badly.

For every selection, we will also ask ChatGPT to explain why it has chosen the stock, looking at factors including:

• recent company news and financial results;
• upcoming earnings or other potential catalysts;
• analyst upgrades and downgrades;
• valuation;
• recent share-price momentum;
• trading volume and liquidity; and
• the main risks that could cause the trade to go wrong.

All of the selections and results will be recorded as the trial progresses, allowing us to calculate ChatGPT’s overall profit or loss, strike rate and average weekly return.

Why Are We Running This Test?

There is certainly no shortage of claims about AI transforming financial markets.

ChatGPT can process enormous amounts of information, summarise company results and identify potential trading catalysts remarkably quickly.

But identifying interesting stocks and predicting which direction their share price will move over the next five trading days are two very different things.

Markets are forward-looking. By the time an earnings report, analyst upgrade or new product announcement becomes public, professional investors and algorithmic trading systems may already have incorporated much of that information into the share price.

That makes this an interesting experiment.

Can ChatGPT sift through the available information and consistently identify stocks that outperform over the following week?

Or will its picks perform little better than random selection?

How the ChatGPT Stock-Picking Trial Works

We began the experiment in August 2026, asking ChatGPT to find one US stock each week that it believes has the strongest short-term trading opportunity.

The selections so far have included some of the world’s largest technology companies, alongside semiconductor and enterprise software stocks.

There have already been both winning and losing trades, including weeks where the underlying investment thesis appeared sensible but the share price still moved sharply in the opposite direction.

That in itself illustrates one of the fundamental challenges of short-term stock trading: being right about a company does not necessarily mean being right about its share price over the next five days.

We will continue running the test and updating the results each week.

There is no cherry-picking, deleting losing trades or retrospectively changing entry prices. Every selection is timestamped before the week’s trading begins and remains part of the final record.

Ultimately, we want to answer a simple question:

Can ChatGPT beat the stock market — or is AI still no substitute for a genuinely profitable trading strategy?

Follow the live results below to find out.

This experiment is for educational and entertainment purposes only and should not be considered personalised financial advice. Share prices can rise or fall substantially and past performance is not a reliable guide to future returns.