Spotting trends with Big Data
More and more industries are discovering the predictive power of large-scale data analysis.
Another benefit of large-scale data analysis is trend prediction. The fashion industry for example, whose upcoming trends are usually forecast by a handful of insiders, is increasingly relying on insights from social media streams. According to an article in the Guardian, specialized consulting firms were able to extract trends from the social media chatter that followed highly publicized fashion events such as the London Fashion Week. Analysts found that street style and sportswear were the trends to watch and that the most talked about color of the season was red.
Another popular huge source of data is web search trends. With Google Trends, Google offers free daily and weekly reports on the queries users enter into Google. These search trends offer a good idea of public sentiment, and, as a recent study has shown, can also be used to measure investor sentiment. As a result, they might even be able to predict where markets are headed.
In a widely-quoted study published in Scientific Reports, researchers were able to find robust correlations between high volumes of finance-related search terms aggregated over seven years and dips in the Dow Jones Industrial Average, a price-weighted average of 30 significant stocks traded on the New York Stock Exchange and the Nasdaq. “When a search term was heavily searched for, the market would generally go down if that search term connoted anxiety”, says H. Eugene Stanley, one of the authors of the study, in an interview for Yahoo Finance. The best predictor of market dips were the search volumes for the term "debt". By adapting their investment pattern to these trends the researchers achieved investment portfolio returns of over 300%.