Today we turn the concept of innovative disruption to our home field of asset management and consider how technology is changing the capital allocation process. In focus: two novel approaches to stock selection, both expressed through US listed exchange traded funds. We highlight them not as investment recommendations, but as examples of novel applications of advanced technology.
All three case studies here use artificial intelligence or natural language processing, and there’s a voyeuristic element for those who pick stocks for a living. What do the machines like right now?
Example #1: AI Powered ETF (symbol AIEQ). The security selection process here runs on the IBM Watson platform, screening for 30-70 companies with the best potential for 12-month price appreciation. The program then constructs a portfolio whose risk profile matches the broader market.
The top 5 holdings/weights are:
- NVIDIA (NVDA, 3.5% weighting)
- Alphabet (GOOGL, 3.3%)
- Nasdaq (NDAQ, 3.0%)
- S&P Global (SPGI, 3.0%)
- Salesforce.com (CRM, 3.0%)
- Also worth noting: Financials are a large overweight at 24% and Energy is a modest underweight at 4%.
Our take: we’ll hand it to the folks that run this product – with just 87 positions in the portfolio they clearly aren’t running an “AI tilt” to the S&P 500. We will be curious to see how the process works in a down market and if the machine ever insists on a weighting over 7-8% for a single stock (indicating an outsized conviction in the name).
Performance YTD: +5.4%
More details here: http://www.equbotetf.com/about-aieq/#investor-materials
Examples #2 & #3: Innovation Shares ETFs (symbols KOIN and EKAR). There are very few public companies entirely dedicated to the blockchain or electric/autonomous vehicles, but plenty of businesses working on these two technologies. That creates a potential investment opportunity to find these plays before their efforts (if successful) are more completely incorporated into stock prices.
The ETF/index designers here use Natural Language Processing (NLP) to scan reams of publicly available information looking for word patterns related to these technologies. Based on the strength of those signals, they create a portfolio around companies relevant to each theme. We met one of the designers last year through a client – the technology is well beyond our knowledge base but we thought it was an interesting way to construct an index.
Here are the top 5 positions for KOIN, the blockchain ETF:
- Visa (V, 7.3%)
- Amazon (AMZN, 6.5%)
- Microsoft (MSFT, 6.5%)
- Intel (INTC, 6.3%)
- Tencent (HK 700, 5.7%)
And EKAR, dedicated to autonomous/electric vehicles:
- Apple (AAPL, 7.3%)
- Intel (INTC, 7.3%)
- Toyota (TSE 7203, 7.3%)
- Alphabet (GOOGL, 6.4%)
- NVIDIA (NVDA, 5.7%)
Our take: time will tell how well the NLP approach to algorithmic stock picking works out, but it strikes us as a thoughtful approach to the difficult problem of finding and leveraging emerging investment themes.
Performance last 3 months (funds only opened this year):
More details here: https://innovationshares.com