How to Use Machines to Beat the Market

My first boss had a mantra that still influences me to do this day.

“Never trust anyone!” she’d declare, while reviewing the same spreadsheet over and over again. “Never trust anyone!”

That may sound a little crazy. But we worked in accounting.

And it was up to us to question every last number. And all the assumptions behind them.

She called this approach “audit skepticism.” Of course, that’s just a fancy way of saying, “Never trust anyone!”

Indeed, my boss was so mindful of the potential for human error that she was obsessed with automating everything as much as possible. The more automation, the less room for stupid mistakes—at least in theory.

But the reality is that automation can introduce new mistakes or even inspire faulty assumptions.

Recently, I’ve noticed that a lot of investors put too much faith in their brokers’ automated stock-rating systems.

Charles Schwab has its own automated stock-rating system, while Fidelity uses one provided by a third party.

These systems pull together numerous financial and market data to gauge whether a stock is a good buy.

The problem is that no amount of financial data can ever tell a stock’s full story.

I should know.

In my former life, I used to track stock-rating systems. And I found that such systems rarely beat the market.

Never Enough Data

These days, I have my own proprietary rating systems that I use to pick stocks for Utility Forecaster. But when you wade through reams of data, you quickly learn its shortcomings.

I’ve certainly tried to come up with the perfect stock screener. And I’ll never stop trying.

However, even with layers and layers of data, you still need the human element.

That means things like listening in on earnings calls, talking to management, and seeing what customers have to say.

It’s impossible to quantify all those things. And even if it were, how would you put it all together into a system that tells you whether to buy, hold, or fold?

Don’t get me wrong. These systems can be incredibly helpful.

When you’re looking for new investments, they can help quickly winnow down thousands of possible choices to a dozen.

And when you think you’ve found the perfect stock, these systems might reveal a fatal flaw that you hadn’t considered.

Sometimes I use Utility Forecaster’s Safety Rating System this way.

After I’ve looked at a company from the top-down and bottom-up, I then run it through the Safety Rating System.

With its eight fundamental criteria, this system functions almost like an instant credit check. And while no system is perfect, I have to admit that sometimes it’s a lot smarter than I am.

But I would never use the Safety Rating System as the sole measure of whether a stock is worth buying. And I wouldn’t rely on any other system for that purpose either.

Man vs. Machine

Ultimately, I believe the best way to use these systems is to combine them with good, old-fashioned human intuition.

If you think about it, that’s the best of both worlds. It’s essentially a combination of human and artificial intelligence.

In fact, nearly all of Investing Daily’s services use a similar combination of proprietary systems coupled with an expert’s knowledge, research, and intuition.

At Personal Finance, Jim Pearce uses his IDEAL Stock Rating System to track every stock in the S&P 500.

At Profit Catalyst Alert, Linda McDonough created a Launchpad System based on her years as an analyst for hedge funds.

Our resident options guru, Jim Fink, even has his own Seasonality Timing app that both he and subscribers can use to come up with new trades.

And each one of these experts knows exactly how to use these systems to tease out insights that others have missed.

We may live in age of automated rating systems, robo-advisors, and indexing. But the human element can still give investors the edge they need to beat the market.