New VanEck ETF Taps Into ‘BUZZ’

Today, VanEck Global launched an ETF that trawls through social media and other internet postings to glean nuggets about the stocks investors are trading, looking to find those that are the subject of sustained positive investor interest.

The VanEck Vectors Social Sentiment ETF (BUZZ) tracks an index that covers the 75 stocks that have the highest “positive insight scores” while meeting size and liquidity requirements. BUZZ comes with an expense ratio of 0.75% and lists on the NYSE Arca.

If the concept sounds familiar, you probably remember the 2016 launch of the Sprott Buzz Social Media Insights ETF (BUZ). It was among the first ETFs to rely on artificial intelligence in its underlying index. While it’s not uncommon to see an ETF incorporating artificial intelligence into its methodology today, BUZ was probably a bit ahead of its time five years ago, and it eventually closed.

New Kind Of BUZZ

However, VanEck’s BUZZ uses the same index, which has been maturing and adapting to changes in the way we use language throughout its five years of existence. The index pulls data from a wide range of sources, including social media, news articles and blog posts, among other venues, and uses natural language processing to determine the nature of the sentiment. From there, stocks are scored and ranked based on how positive the sentiment around them is and how much they appear in the surveyed sources, according to a fact sheet.

“For a long time, we’ve believed in the power of the insights from the collective convictions of all the millions of people that are engaging in stock-related discussions across those platforms. We were strongly convicted about that community back then and we remain so today,” said Jamie Wise, founder of BUZZ Indexes.

With five years of live calculations backing the BUZZ NextGen AI US Sentiment Leaders Index, Wise notes that the increase in activity to analyze has been quite high. He says the index originally analyzed roughly 2 milllion posts every month five years ago, whereas now it is analyzing more than 15 million posts.

And all along, the model has been adapting to accommodate the changes in how investors communicate with each other. One such change was a dramatic increase in the use of emojis in stock-related conversations, Wise says. He points out that the model had to be retrained to account for that shift.  

Index Design

Companies must have a market capitalization of at least $5 billion and have a three-month average daily trading volume of at least $1 million, while meeting a minimum threshold for online mentions. The top 75 stocks meeting those parameters are selected for inclusion in the index, with updates occurring on a monthly basis, the document says.

The index’s components are weighted by their sentiment rank, and individual securities are capped at 3% of the index, according to the fact sheet.

Importantly, the methodology is designed to respond to sustained trends around large cap stocks, which means that manipulating the index would be hard to do, and that short-term market disruptions won’t result in changes to the index.

“We think about sentiment through a longer-term lens that’s consistent with monthly rebalancing of the index, so those short-term events won’t really impact the index at all, and the index won’t change immediately to reflect those short-term sentiments,” Wise explained.

The end result is an index that encompasses a diversified mix of large cap stocks that represent different factors and styles, although perhaps with a more significant tilt to momentum. The fact sheet notes that sentiment can be driven by any number of factors, and while price momentum can generate chatter, so can stocks that may be undervalued by the market.

Among the index’s top components as of mid-February were Twitter, Amazon.com, Ford Motor Company, Virgin Galactic and DraftKings.

(Use our stock finder tool to find an ETF’s allocation to a certain stock.)

Contact Heather Bell at [email protected]