ChatGPT identified 5 stocks that DESTROYED the market in 2024. Here is what it says for 2025.
Last year, I used an AI-Powered stock screener to identify 5 fundamentally strong AI stocks. These stocks ended up DESTROYING the S&P 500, earning 44% since February 20th (the time I posted my article).
In contrast, the S&P 500 eeked out a modest 19%.
A backtest for these fundamentally strong AI stocksWas it luck? Or did I identify real factors that are useful for cherry-picking fundamentally strong stocks?
Let’s find out.
A Recap: What stocks did I pick last year and how?
Last year, I posted the following article.
In it, I used AI to find stocks at unprecedented speed. Unlike what you would find on TikTok and Instagram, these stocks are based on objective fundamental data.
More specifically, I said the following to an AI Assistant.
Find AI or semiconductor stocks that meet the following criteria:
* Increasing gross profit margin over the past year
* Profit margin of 60% or more
* over $5 billion in revenue Q3 2023
Sort by profit margin descending
These metrics are fundamental metrics – they give a glimpse into a company’s financial health.
For example, gross profit margin shows how much of each dollar of revenue is profit after paying for the cost of goods sold. Revenue tells us how much money we produced in a single quarter.
And thus, an increasing gross profit margin and high revenue show the company is making money and converting more of that money into profit (before considering operating expenses and other costs).
With this question to the model, Aurora identified the following 5 stocks:
- Broadcom (AVGO), up 88.2%
- Nvidia (NVDA), up 88.3%
- Meta (META), up 28.7%
- Salesforce (CRM), up 14.2%
- Microsoft (MSFT), up 4.7%
The performance of each stock since February 19thNow that it’s the beginning of the new year, I’m asking myself this question…
Were these stock picks lucky? Did I happen to pick 5 stocks that did extraordinarily well? Or, are these fundamental metrics truly indicative of future growth for AI stocks?
I backtested these rules in the past with outstanding results.
The results of these rules in the pastBut we all know that the past doesn’t predict the future.
And thus, I sought to test this more explicitly in real-time.
The changes in methodology from 2024 to 2025
We are going to repeat this experiment in 2025. In other words, we’re going to use the AI Assistant, Aurora, to query for fundamentally strong AI stocks in NexusTrade.
I will then transform the list of stocks into an algorithmic trading strategy.
To do this, I’m going to repeat what I did last year. However, while the request to the assistant will be the exact same, there are a few differences under the hood.
Querying for AI/semiconductor stocks in 2025The biggest difference is that the AI stock screener has had a massive overhaul. Some of the changes include:
In theory, this means the results should be more accurate than they were last year. Let’s use this model and see what stocks AI has picked in 2025.
Querying for the new list of fundamentally strong AI stocks
The response of the AI model in 2025This new list of stocks, retrieved using the O1 model, includes:
- Microsoft (MSFT)
- Meta (META)
- NVIDIA Corporation (NVDA)
- Cisco Systems Inc. (CSCO)
- Salesforce (CRM)
- Intuit Inc. (INTU)
- Adobe (ADBE)
Again, these stocks are derived directly from objective fundamental data. The question isn’t “whether or not I’m a good stock picker”; it becomes whether or not the fundamental metrics I look for in a stock leads to outsized gains.
To turn these stocks into a portfolio, I used the following command:
Create a strategy to buy and hold all these stocks
Creating a strategy automatically runs a backtest for these basket of stocks. As we can see, this set of stocks significantly outperformed the market last year.
But that’s not the goal of this article.
We want to see how it performs next year.
So, I have saved this portfolio and deployed it live to the market.
Deploying this portfolioWith this deployment, we can see how the portfolio performs at the end of 2025. I have an objective audit log with this conversation to Aurora, so we can truly see how effective these stock picks are.
Concluding Thoughts
While past performance is no guarantee of future results, the success of the 2024 stock picks demonstrates that using objective fundamental metrics with AI tools can yield impressive results. This year’s test is a continuation of last year’s experiment, and a deeper exploration into the potential of AI for making smarter investment decisions.
The value of using AI to streamline financial research cannot be overstated. With tools like Aurora in NexusTrade, investors can quickly and efficiently analyze vast amounts of data to uncover hidden opportunities. The new advancements in the AI stock screener, including the integration of cutting-edge models and a BigQuery backend, promise to further enhance the accuracy and reliability of these insights.
Are you ready to see how AI can transform your investing journey? Join me on NexusTrade to explore the power of AI-driven financial research and strategy creation. Let’s take the guesswork out of investing and embrace a data-driven approach to achieve our financial goals. Start now and discover what’s possible!
