What stocks would artificial intelligence select to produce outsized returns over the next 20 years?
That’s the question I wanted to ask ChatGPT, the breakthrough AI product from OpenAI, and the results were stunning.
- ChatGPT was asked to identify the characteristics of the top performing stocks since 2000 and then produce a list of 20 current stocks that possess those similar characteristics.
- The best performing firms had strong growth potential, brand recognition and customer loyalty, financial strength and stability, and focus on shareholder value.
How The Experiment Began
To begin, I asked the groundbreaking technology to evaluate what stocks had performed best since 2000. Now keep in mind ChatGPT data doesn’t have a data set current up to present day so I asked it to evaluate the data for the longest period to which it had access since the turn of the millennium.
This is the list of stocks it said had the best track record during that time frame.
- Amazon (AMZN)
- Apple (AAPL)
- Alphabet (GOOGL)
- Microsoft (MSFT)
- Facebook (FB)
- Visa (V)
- Mastercard (MA)
- Netflix (NFLX)
- NVIDIA (NVDA)
- Adobe (ADBE)
- Tesla (TSLA)
- Salesforce (CRM)
- PayPal (PYPL)
- Booking Holdings (BKNG)
- ASML Holding (ASML)
- PayPal Holdings (PYPL)
- Charter Communications (CHTR)
- Regeneron Pharmaceuticals (REGN)
- Biogen (BIIB)
- Illumina (ILMN)
Next, I asked ChatGPT to identify common themes among those companies.
What made them so successful over the past two decades?
What Do The Best Performing Firms Have In Common?
Here’s what AI said:
- Strong growth potential: Many of these companies are leaders in their respective industries and have demonstrated strong revenue and earnings growth over the years. Investors have been willing to pay a premium for their shares based on their growth potential.
- Innovation and disruption: Many of these companies have disrupted traditional business models and created new markets by leveraging innovative technologies and business strategies. This has allowed them to gain a competitive advantage and capture significant market share.
- Brand recognition and customer loyalty: These firms have established strong brands and have a loyal customer base that is willing to pay a premium for their products and services.
- Financial strength and stability: They have strong balance sheets, with significant cash reserves and low levels of debt. This has allowed them to weather economic downturns and invest in growth opportunities.
- Focus on shareholder value: They often have a track record of returning value to shareholders through dividends and share buybacks, which has helped to boost their share prices over time.
Now came the most intriguing part of the exercise. I asked the AI to identify which companies today possesses similar characteristics to those that outperformed over the past 20 years. And further, I asked it to limit the size of the company to a market capitalization of $10 billion or lower.
The reason for that final constraint was that some companies, like Apple, are undoubtedly extraordinary with wide moats that are unlikely to be disrupted for decades, if ever. But with a market cap close to $3 trillion already, Apple’s size currently eclipses the sum of the market caps of all companies in Germany and India combined. In short, the Law of Large Numbers will make it very difficult for Apple to produce massive percentage returns.
So what group of stocks did AI produce that had the characteristics of the best performing stocks since 2000?
Here’s the portfolio of stocks it identified:
- Fastly, Inc. (FSLY)
- Elastic N.V. (ESTC) – replaced with Alteryx, Inc. (AYX)
- Okta, Inc. (OKTA)
- Datadog, Inc. (DDOG)
- Zscaler, Inc. (ZS)
- Twilio Inc. (TWLO)
- Alteryx, Inc. (AYX)
- DocuSign, Inc. (DOCU)
- Zuora, Inc. (ZUO)
- Vroom, Inc. (VRM)
- Fiverr International Ltd. (FVRR)
- Upstart Holdings, Inc. (UPST)
- Sprout Social, Inc. (SPT)
- Bill.com Holdings, Inc. (BILL)
- Appian Corporation (APPN)
- PagerDuty, Inc. (PD)
- The Trade Desk, Inc. (TTD)
- Arlo Technologies, Inc. (ARLO)
- Unity Software Inc. (U)
- C3.ai, Inc. (AI)
Will AI prove accurate? We’ll track the list and see whether artificial intelligence can produce alpha.