For Investors

BTD is straightforward: buy a temporarily undervalued stock and sell it after it rebounds.

So why doesn’t this strategy always work? Pundits claim dips can’t easily be distinguished from dives. And they’re right — for most of us, it’s nearly impossible to consistently find stocks that are briefly undervalued. Intuition, a gut feeling that a stock should bounce back, is just that — a feeling. But intuition can be wrong, and that “gut feeling” could result in a loss.

The AI behind DIP isn’t a single machine but a combination of machine learning algorithms — think of them as “brains” — working independently to track and qualify billions of data points. The collective AI then applies a proprietary set of filters to these qualifiers to identify and evaluate likely dips — the most promising of which are flagged for purchase. The AI’s prescribed transactions are executed throughout the market day, buying newly identified dips and selling rebounded stocks.

The result? A genuinely dynamic ETF. As the AI is optimized for finding and selling dips to maximize the delta — the difference in price between when a stock is bought and sold — you can expect 10 percent to 30 percent of the portfolio to turn over every day. The AI then directs the profit to be reinvested into more dips.

The system’s goal: Repeat this process and grow DIP’s value.

The AI powering DIP processes information on a level humans can’t comprehend and behaves in a way the human brain can’t easily or reliably replicate.

Consider DeepMind Technology’s AlphaGo, the first computer program to defeat a Go world champion. Given the complexity of the 3,000-year-old game — there are 10 to the power of 170 possible board configurations in Go — even advanced computer programs using standard AI methods could only play Go at the level of human amateurs.

But AlphaGo’s algorithm leveraged deep neural networks — like those powering DIP’s AI.

After AlphaGo was introduced to Go’s rules and typical gameplay, it began competing against some of the world’s most elite Go players. In one such match, AlphaGo made moves that were so unusual, so profoundly different from those that a human would play, that its programmers believed it to be malfunctioning. But by leveraging patterns unseeable to humans, AlphaGo won — upending hundreds of years of gameplay wisdom.

AlphaGo is now regarded as one of the greatest Go players of all time. Now imagine what neural net-based AI can do with the markets.