Where You Get Your Data Matters

When Chaikin Analytics founder Marc Chaikin recently called me about a trading setup he spotted, I knew he’d found something good…

Vic, the technical setup here is near perfect. I want to get this one out as fast as possible.

Marc had identified a trade opportunity for Chaikin PowerTrader newsletter subscribers. And “as fast as possible” is as tight of a deadline as you can get in our business.

On top of that, Marc made the task even more challenging…

He had picked a company that’s so boring, most investors never think about it.

I pondered how to tell this story in an exciting way. And I knew that I needed to start right away.

My first step?

I bought a book.

Now, I realize that doesn’t sound exciting at all. And on its face, it’s not…

You’d expect financial analysts to do a lot of reading. I’m no exception to the rule.

But the book I bought had a major problem. And that’s the big point of today’s essay…

Put simply, the author filled the book with pages and pages of AI-generated slop.

The cover promised an in-depth look at the company Marc had spotted…

I hoped to find exciting stories – and more importantly, interesting data – to use as a springboard for my research. Instead, I found nothing more than ChatGPT-style summaries.

I won’t name the book. After all, I can’t prove that the author “wrote” it with evil intent.

But it’s one of the worst examples of bad data I’ve come across in my years in this business.

And it proves the value of trustworthy information in our increasingly digital world.

Just look at what happened at Sports Illustrated last November…

The magazine first published 70 years ago. It was long known as one of the most reliable names in sports coverage. But even it allegedly fell victim to the lure of “easy” content…

Tech website Futurism reported that Sports Illustrated used AI-designed headshots and bios for several authors. Sources told Futurism that Sports Illustrated published AI-generated articles as well. (Sports Illustrated‘s parent company blamed a third-party contractor.)

This example should make you wonder…

How much bad AI data is out there these days?

This problem doesn’t end with the book I bought and the Sports Illustrated scandal, either.

People have caught ChatGPT and other AI models regurgitating all kinds of bad data…

Alphabet’s new AI search feature told users that cheese stays on pizza better with “nontoxic glue.” And it said that geologists recommend “eating at least one small rock per day.”

That may sound ridiculous to you. But this nonsense represents a huge data problem.

You see, Google trained its latest AI model on everything it could scrape off the Internet. That included online message board Reddit and satire website The Onion.

Bad data got in. And it messed up the output.

Sure, I’m sharing silly examples of data gone wrong. But in finance, data quality is as serious as it gets. It drives hundreds of millions of dollars’ worth of decisions every day.

Marc knows that…

It’s why he built his decadeslong career in this business around learning to retrieve data from the most dependable sources. Along the way, he also learned to interpret it correctly.

Some of Wall Street’s most powerful names trusted Marc to provide everything they needed when it came to data and analysis. And nowadays, data makes the Power Gauge such an incredible tool.

Folks, the world is changing quickly. To be sure, AI is a powerful tool. But as I’ve discussed today, it still has its limitations when it comes to data.

Put simply, where you get your data matters – especially when it comes to your investing decisions.

Good investing,

Vic Lederman