Your tweets may be moving markets.
An exchange-traded fund, launched earlier this year, uses online chatter about stocks—drawing from social media sites like Twitter and Facebook, but also blogs and comment boards—to compile its roster of portfolio holdings, betting that the aggregate views of thousands can point to outperformance.
The Sprott Buzz Social Media Insights ETF is premised on the idea that if huge masses of people share their trading activity—not only the names they’re buying and selling, but also the price levels they’re looking out for and other more technical details of their strategies—that data can be mined for measures of sentiment. The firm calls this metric “social momentum,” and it flies in the face of most sentiment measures by taking a non-contrarian view.
“When you look at views that people volunteer, as opposed to what they say when they’re being polled, the results are more reliable,” said Jamie Wise, founder of Buzz Indexes.
“The amount that people share on social media has grown astronomically—there’s a ton of commentary about what people are buying, what they make of certain stocks, what they like or dislike, what trades have worked or not worked for them,” he said. “And now machines have gotten smart enough that not only can they scan through huge amounts of data, but they’re able to detect sarcasm and understand financial terms.”
The ETF has struggled to gain traction. Flows into it have been slight—about $2.46 million in 2016, according to FactSet data; it has $5 million in assets—and over the past 30 days it has posted an average trading volume of only 157 shares. Wise admitted that flows “are lower than we expected.”
Adoption may also be stifled by the fund’s relatively high fee—an expense ratio of 0.75%, which Wise said was necessary given how expensive it was to run the strategy—and by the familiarity of its portfolio components.
Wise likened the fund’s thesis to the “ jelly bean experiment ” made famous in James Surowiecki’s 2004 book “The Wisdom of Crowds.” In that experiment, a group of people was asked to estimate the number of jelly beans in a jar. While single guesses varied widely, the group’s average was likely to be close to the actual number.
“You’re not picking individual pieces of data. ‘Joe from Iowa’ tweeting once about a stock isn’t worth listening to, but a million Joes tweeting a million times are,” said Wise. “This isn’t a high-frequency strategy or about what went viral yesterday; this is designed to measure sentiment over time. If thousands and thousands of people tweet or talk about liking a stock over a long period, we think that’s predictive.”
The ETF’s focus on longer-term sentiment allows it to avoid daily swings or the impact of institutions changing their views on a stock. “On a single day, no one can compete with the capital flow of a big manager, but over the longer term, it will be the fundamental factors that play out,” Wise said.
Because the fund launched in April, it is too early to tell how successful this strategy is, though Wise says that backtesting to 2013—the earliest date for which there is adequate social media data—has shown outperformance.
So far this month, it is down 2.7%, a decline that’s nearly twice as steep as the S&P 500’s /zigman2/quotes/210599714/realtime SPX +0.45% 1.4% drop. But over the past three months, it has eked out a gain of 0.2%, compared with a loss of 1.3% in the S&P over that same period.
Because the fund limits itself to stocks with a market cap above $5 billion, a measure to avoid the names most vulnerable to rumors or “pump and dump” schemes, most of its top holdings can also be found in large-cap funds that have much lower costs. The Vanguard Large Cap Index ETF /zigman2/quotes/209242880/composite VV +0.46% —which has a heavy overlap in components, if not allocations—has an expense ratio of 0.08%.
The fund’s top five components are Microsoft Corp. /zigman2/quotes/207732364/composite MSFT +0.76% , Google parent company Alphabet Inc. /zigman2/quotes/202490156/composite GOOGL +1.94% , Facebook Inc. /zigman2/quotes/205064656/composite FB -1.74% , Walt Disney Co. /zigman2/quotes/203410047/composite DIS -0.73% , and Apple Inc. /zigman2/quotes/202934861/composite AAPL +0.0000% . For the whole fund, the components have an average price-to-earnings ratio of 23.15 and an average dividend yield of 1.99%, along with average annual sales growth of 10.1%.
To compare, the average S&P component has a P/E of 28.06 and a dividend yield of 2.14%, along with much lower revenue growth, of 2.02%.