AI Investing Conclusion and a Surprise Call from the CEO of Danelfin

My three month experiment is over. I’ve sold all my stocks, applied salve to my wounds, and it’s time to report on the findings. For new readers, I used an AI Stock Picking tool called Danelfin AI back on September 12, 2023, to give me its 10 best stock picks for a three month experiment.

I then invested $10k in each of these stocks, for a total of $100k invested. I sold whenever the AI dropped its rating from a strong buy to a weak buy.

Here are the final numbers over the 3 months of the experience:

1. S&P 500 started at 4450, ended at 4626, +4%

2. Using Danelfin AI I started with $100,000, and ended with $87,000, -13%

3. If I had just held on to the portfolio as recommended, for three months, my final total would have been $82,000, -18%

So depending how you want to look at it, using AI instead of just investing in the S&P 500 cost me 17% or 22% in returns. Boo. AI bad. But wait!

In a surprise move after my last update on this experiment, the CEO of Danelfin AI, Tomas Diago reached out to me on LinkedIn asking if we could connect via Zoom to talk about how I was using the platform and how to best leverage his company’s software.

I have enormous respect for any CEO who pays attention to criticism of their product, and so when we had our Zoom last week, I wasn’t surprised to find that Tomas was thoughtful and had great insight into the stock market as well as on how best to use AI for investing.

His central point was that AI tools are best used to do research on markets, and find out more about stocks you’re interested in buying. One way he suggested I could use Danelfin to make purchase selections was to take stocks highly rated by the AI, filter out the highest risk ones, and then create a diversified portfolio across five major sectors to have about 30 stocks, all with low to moderate risk (6 or higher).

Danelfin has published a white paper with the results of such a strategy since 2017 and the results are about 20% better than the S&P over the past six years. Take it with a grain of salt, it’s a backtest after all, but it’s certainly a different strategy than the one I was using, which was just picking the 10 highest rated stocks overall.

I briefly considered running another 3 month experiment, this time using his suggested strategy and a much lower starting investment, but then decided I had solved my actual problem, which was the worry that I might be missing out on some great investment strategy using AI. That is definitely not a thing.

AI is a useful research tool. So is Yahoo Finance, talking to your friends, and reading annual reports. Add it to your portfolio of tools if you’re inclined, but don’t use it to actually pick stocks. My advice to myself and to you — if picking stocks is anything less than a full time job for you, stick to investing in indexes and go worry about something else, like what to wear at the dinner tonight that you can afford because you didn’t lose $13,000 on a dumb stock market experiment.

That’s a hell of a lot less than I’ve lost on other experiments, including getting an MBA, and if it convinces me to trust the professionals like my own advisor, Dean Colling, and stay out of investing in individual public stocks, then it was money well spent.

There are indeed plenty of firms out there who have developed software to help them make better investment decisions. That proprietary software often relies on being able to execute trades faster than other firms, and certainly faster than individuals.

Investors like Warren Buffet can move markets with their trades, and have access to information about their investments via board positions and their network that you, as an individual, do not have.

For an individual, the stock market is a casino. You can invest in the house by buying an index fund, or you can sit on the other side having fun and saying “hit me”. But if you choose to play in the casino, don’t kid yourself and say you’re an investor. You’re just gambling.

Myself, I’m going to go back to what I’m good at, which is making early investments in tech startups where I have inside information on their performance and co-investors I trust to do good due diligence. It’s where I made my fortune, and I’m sorry I got entranced by the constant stream of public market pro-investing garbage published in newsletters and on TV every day.

Figure out what you’re good at. Go do that. And when you’re in the mood to relax and spend some money at the casino, your portfolio will be there.

1. My original article (with links to updates) on AI investing.

2. How to make money investing in startups.