This article was originally published on the Artificial Intelligence in Plain English blog. I’m reposting it here to share these ideas with a wider community. Please comment below and let’s start a discussion!

Pic: A screenshot of my Robinhood account](https://miro.medium.com/v2/resize:fit:1400/1*[email protected])

Warren Buffet said something that shocked the finance community — outperforming the S&P500 is easy. Because of the size of Berkshire Hathaway, Warren believes he’s at a massive structural disadvantage. He claims that if he was trading at a smaller scale, he could easily outperform the market.

“I think I could make you 50% a year on $1 million. No, I know I could. I guarantee that.” – Warren Buffet

At the time, Buffet relied on Moody’s Manual, a series of publications by a financial services company (named Moody’s) on publicly traded stocks. These publications provided detailed information on various industries, companies and securities.

Warren Buffet wasn’t sure what the modern-day equivalent of Moody’s Manual was, but I am confident that I found the answer…

It’s artificial intelligence.

My returns since incorporating AI

I’ve been incorporating automation and artificial intelligence into my investment strategy for around a year, and I have produced returns that have left the S&P500 in the dust. While some might say 1 year isn’t enough to make definitive conclusions, I have enough faith in my approach to share my process to the world.

It’s honestly easier than you think.

Pic: My returns year-to-date (YTD)

To start, let me show my Robinhood’s returns. The chart shows that I gained $12,006.92 YTD, increasing my account balance to $36,383.53, a 49.3% gain. In comparison, SPY, an ETF tracking the S&P 500, gained 21.6% in the same time period. This shows my investment strategy is significantly outperforming the broader market.

This trend continues even if we zoom out, even for the past 3 years (when this account was opened).

What’s even more amazing is that the true percent gain is actually much higher.

You see, my account balance started with less than $12,000, and after gaining more confidence in my strategy, I’ve deposited more cash into my portfolio. If you were to see my actual percent gain, you would be shocked.

You can even see that I’m managing to do so well despite having $29,888.88 in buying power. You might be wondering how I have so much buying power, and yet, I’m outperforming the market?

I trade options.

What is my investment strategy?

Let me preface this section by saying that I do not have a crystal ball. I do not offer financial advice to anybody, and my personal trading account is money I am 100% okay with losing. I’m not demanding that you should replicate my strategy to “get rich quick”, I’m merely explaining how I approach the market and incorporate AI into the process.

The basic premise of my strategy is this – if the S&P500 does approximately 10% per year, then around half of stocks will underperform the market, and another half will outperform.

When there are losers, there are winners.

The premise is simple — find one or two stocks that I believe will outperform the broader market. This year, I picked Tesla (near its bottom price), NVIDIA, Google, and Microsoft. Then, trade long-term (1 year-to-expiration) ATM options on the winning stocks.

If the stock happens to go down, the goal is to slowly accumulate more. “Slowly” is the keyword. You don’t want to spend all of your buying power in 2 weeks. The goal is to stretch out the purchases across weeks or even months, and slowly lower your cost basis.

If the stock goes up, sell off a few of your positions, but try to hold at least 1 for the long-term.

That’s really it. While I don’t have a backtesting platform for options, I’ve backtested very similar strategies using volatile assets like FNGU and TQQQ to validate my ideas.

Pic: Backtesting a leveraged strategy

The key part to this is finding great assets to trade. While QQQ is a winner, I’m a big fan of picking even more dominant winners.

And how do I find these winning stocks without a crystal ball?

I rely heavily on artificial intelligence.

Using AI to identify and evaluate fundamentally strong stocks

Pic: The characteristics for fundamentally strong AI stocks

I rely heavily on AI to both identify and evaluate fundamentally strong investment opportunities.

I built a free online platform, NexusTrade, that’s heavily integrated with large language models. Two of the key features that I use for financial analysis includes the AI-Powered Stock Screener and the AI-Powered Financial Analysis feature.

For example, in the article below, I described the process of how to find fundamentally strong AI stocks using AI.

Specifically, I used a natural language interface to query for stocks with the following characteristics. * AI or Semiconductor stocks * Have increased their gross profit margin over the past year * Their gross profit margin is 60% or more * They made over $5 billion in revenue in Q3 2023

The language model generated a query to find all stocks that fit this criteria, then sorted the list by profit margin descending.

Ultimately, I identified Broadcom, Microsoft, NVIDIA, Salesforce, and Meta.

Pic: Fundamentally strong AI stocks

These 5 stocks are crushing the market since I published the article. They are up 25% since February, compared to the S&P500’s percent gain of 15%.

But it’s more than just screening for stocks that makes AI so powerful. It can also do a holistic analysis.

Pic: Asking AI to analyze NVIDIA stock

For example, NVIDIA is one of my major investments for this year. One of the reasons that NVIDIA is a large percentage of my current portfolio is because its a fundamentally strong AI stock. Given the fundamentals of NVIDIA, the language model rated it a 4.5/5 as an investment.

Pic: AI’s recommendation of NVIDIA

I don’t just rely on one cherry-picked model. I use multiple, including Claude 3.5 Sonnet, GPT-4, and the newest GPT-o1-mini (“strawberry”) models.

Pic: The different LLMs within NexusTrade

Combining the consensus recommendation from the different LLMs with my human intuition that I’ve gained as a trader and investor, I can make much better decisions for my investments – decisions that have paid off massively.

And then, combining these fundamentally strong investments with a strategy validated by backtesting on historical data, I’ve ended up creating an extremely profitable trading strategy that has shown to outperform the market within the past few years, and doesn’t even use a significant portion of my buying power.

Concluding Thoughts

My strategy has been wildly profitable in the past couple of years, but I don’t pretend to have a crystal ball. Tomorrow, maybe NVIDIA goes directly to 0, and I’m only left with a 20% year instead of my market-destroying 50%. That can certainly happen.

But, its unlikely. While Warren Buffet relied on Moody’s Manual, there is a new opportunity to find strong, hidden gems thanks to artificial intelligence.

Not only is it easier to find novel investments, but we can also perform comprehensive analysis. The people wielding this technology are going to be far more effective than the people stuck in the stone age, reading newspapers and looking at 10K statements.

It’s time for the tech-savvy investors to usher in a new era for investment strategies. It’s time to incorporate AI into your process.

Or, get left behind. The choice is yours.