AI Case Study
Wells Fargo demonstrates 88% short-term and 58% longer term recommendations accuracy using machine learning to predict stock price movements
Wells Fargo bank has begun to use an AI program to make buy/sell recommendations based on stock price predictions sourced from a variety of data including social media; however, actual decisions and implementations are still done by human analysts. Thus far its short-term call predictions have been more accurate than those over 5 days (88% vs 58%).
Wells Fargo has introduced AIERA (artificially intelligent equity research analyst) which issues buy and sell calls on over 500 stocks. However, while Wells Fargo reports the predictions AIERA makes, its human analysts ultimately issue their own recommendations that may or may not correspond with the AI's.
From Bloomberg Businessweek: "Aiera’s recommendations are for short-term periods. It scours the internet for stories, earnings reports, social media posts, and analyst research on more than 500 stocks and uses tools such as natural language understanding to turn the words into a measurement of sentiment. “Fear,” “anger,” “joy,” “sadness,” and “surprise” are combined into an overall sentiment score. Aiera then monitors the market to see if the emotions it identified move stock prices. If there’s a correlation, it stores that and uses it to make predictions and churns out one-paragraph summaries of the most relevant information for each stock."
Budgeting And Forecasting
Wells Fargo is purportedly the first major bank to make public the AI-powered recommendations from its equity research division.
"As of early November , calls it made for an eight-hour period had an overall accuracy rate of 88 percent, while 58 percent of its recommendations for eight days worked out. The eight-hour calls do best because there have been so many more of them to feed back into the learning algorithm. Most of those recommendations are holds, which work out well partly because nothing much happens to a stock in a few hours." (Bloomberg Businessweek)
Uses over 1,000 data sources, including articles from Barron's, Facebook and Twitter feeds which are sourced in real-time