As readers of this blog are aware, MKTSTK likes to monitor non-traditional sources of information about financial markets. One particularly rich data source has been search volume, and in the past we have shown several of the ways Google search volume can be used to analyze the stock market including:
- Timing Tesla with Google Trends
- Trading Google Trends with Logistic Regression
- This Google Search Term Just Set a Record and Investors are Noticing
This is why we found it particularly interesting that search volume for the term “stock market” has reached levels not seen since October 2008: i.e. the period immediately following Lehman Brothers collapse.
As this chart makes clear, it is extremely dangerous when your average investor starts paying attention to their portfolio. Instead of seeking out advice from their advisers, investors respond to unexpected declines by searching the internet for explanations. This allows us to measure investor fear levels using social data sources derived from search volume and social media messages. These data sources can provide us with novel ways to predict the stock market.
The evidence is mounting that is current spike in volatility will persist for some time. Last week we pointed out that volatility tends to remain elevated after the stock market breaks a significant nonlosing streak. All this suggests taking a very cautious view towards shorting volatility.
One of the best ways to stay alive in the trading world is to identify when things are different than normal. When you train your quant model on past data you are implicitly hoping the future will be like the present. Oftentimes, however, we only need simple rules to differentiate normal from abnormal markets.
Want to learn how to mine social data sources like Google Trends, StockTwits, Twitter, and Estimize? Make sure to download our book Intro to Social Data for Traders
Interested in resources for Stock Market Visualization?
Categories: Quantitative Trading