In a perfect world, stock prices are rational reflections of market fundamentals. But the “efficient market hypothesis” is tempered by the fact that in reality, markets and securities are subject to waves of optimism and pessimism—bullish or bearish investor sentiment that influences the process of price discovery.
Behavioral finance exists as a discipline because people are not purely rational. We tend to trade on market “noise” and our emotions rather than facts. This sentiment phenomenon causes investors to have beliefs about the prospects for stocks, which are determined by factors that can outweigh traditional future cash-flow modeling and investment risk analysis.
Yet investor sentiment is complex and traditionally difficult to measure. As a result, for decades people have been using various proxies (and proxy indices) to quantify investors’ “feelings” about stocks. Popular sentiment proxies include:
- Investor surveys (such as the Bloomberg Consumer Comfort Index; the American Association of Individual Investors Sentiment Indices; Wall Street Strategist Asset Allocation)
- Market-based measures of “fear” (i.e. Credit Suisse Fear Barometer; CNN Money Fear & Greed Index)
- Retail investor trading flows and mutual fund flows
- Technical analysis of trading volume and breadth (i.e. advance/decline ratio; new high/new low ratio; up volume ratio; TRIN)
- Dividend premium; closed-end fund discounts; IPO returns and trading volume
- Option implied volatility; put-call ratios
- Insider trading
AI sentiment gathering beats proxy sentiment
What we’ve done at BUZZ is to demonstrate that AI and natural language processing technology makes sentiment proxies obsolete. We leverage artificial intelligence techniques to isolate the sentiment premia that have always existed within stock price returns.
Proxy sentiment measures are relics now that we have the ability to measure sentiment down to the individual stock, thanks to the digital trail left behind by millions of people sharing opinions and views across online platforms.
For example, since our launch last year BUZ has outperformed another sentiment themed ETF that holds large and mid-cap stocks of companies from the S&P 1500 Index which show positive insider sentiment from insider buying and favorable analyst ratings.
This other index is yet another proxy based approach to mining for sentiment. It is designed to reflect positive sentiment among corporate officers, directors, large institutional shareholders, and the Wall Street research analysts who follow the company—those “insiders” most knowledgeable about a company’s financial statements and business prospects.
Yet BUZ, which harnesses the collective wisdom of millions of investors on social media, in real time, has consistently better returns.
BUZ has outperformed other active large cap equity strategies and has performed well versus passive benchmarks. The index beat the S&P 500 by nearly 3% in the third quarter of 2017 alone.
The plethora of proxy based sentiment indicators is proof that investors recognize the importance of investor sentiment, and the potential value of the insights contained therein. BUZ is changing the game by using the latest technology to measure sentiment directly, in real-time and at scale—a far more accurate approach than the proxy methods of yesterday.