Tapping into Twitter for more than a hot IPO

As today's wildly popular Twitter Inc. (TWTR) stock offering clearly proves, social media is hot. Shares of the 7-year-old microblogging site spiked more than 90% above the $26 offering price Thursday, with trading volume exceeding 55 million shares in the first 90 minutes after the initial public offering. This is all music to the ears of James Ross, not because he is an investor but because his company, HedgeChatter Inc., is in the business of tracking, filtering and analyzing online financial chatter for the purpose of generating quantitative investment projections. “Twitter is not just a platform for 14-year-old girls to talk about their favorite ice cream,” he said. “Twitter makes its money by selling data and selling ads, so we're monitoring how people are going to be interacting with Twitter as a stock, and we're utilizing Twitter's own information to determine where their stock price is going to go.” Hedgechatter.com, which deliberately scheduled its launch to coincide with the Twitter IPO, is the latest iteration of a fast-evolving industry of online financial sentiment analysis. Platforms such as stocktwits.com categorize and run live streams of tweets that include ticker symbols posted with a dollar sign in front of them to distinguish them as financial or investment-related posts. Another platform, sntmnt.com, provides a stream of data and sentiment scores for financial chatter posted in Twitter. Also, platforms such as dataminr.com allow subscribers to set up templates to mine specific online financial chatter. All these platforms, including HedgeChatter, target the full range of investors and traders, from the casual amateur to the most sophisticated institution. But where HedgeChatter hopes to stand out is by both qualifying and quantifying the online chatter, including weighing and ranking the source of the information. “Financial chatter is all over the place, but the knock against it has been its actual usefulness, because most of the unstructured data is incredibly noisy and unfiltered,” said Chris Camillo, a HedgeChatter board member. “What HedgeChatter has done is taken all of this financial chatter and created rankings and is sorting the data based on its usefulness and influence of various stocks," Mr. Camillo said. "That provides you with sentiment analysis and scores based on the historical influences of those chatters.” One example of the power of financial chatter is seen in the market's reaction to two Aug. 13 tweets by billionaire investor Carl Icahn. In the two microblog entries, he cited his large position in Apple Inc., called the stock extremely undervalued, referenced a conversation with Apple's chief executive, and said he hoped the company would increase its stock buy-back plan. From the time of the first tweet, less than two hours before the end of the trading day, Apple shares gained nearly 3%, increasing the company's market capitalization by $12.5 billion. “That's an extreme example, and obviously some people have more influence than others,” Mr. Camillo said. “You might have an individual who chats about 100 securities but is only influential on 20 of them, and that's what this analysis is about.” Another example of the power and influence of online chatter is the April 23 tweet that came from a hacked Associated Press Twitter account that claimed that two bombs had exploded at the White House and that President Barack Obama had been injured. Before any news reports could correct the false tweet and report the hacking, the Dow Jones Industrial Average lost 140 points, which were quickly recovered once it was clear that the tweet was a hoax. “There are 550 million tweets a day, and one rogue tweet moves the market,” Mr. Ross said. “The fact is the markets require the emotional insights of humans in order to make the financial arbitrage, and that's the kind of stuff we're mining because we're looking for the micro-Jim Cramers out there who get it but they have a TV show to talk about it.” The data mining by HedgeChatter includes monitoring the tweets from nearly 80,000 Twitter accounts, in addition to monitoring all the news, blog posts and financial chatter related to every U.S. stock every 30 seconds. The platform gives weight to comments and posts based on the calculated influence of each source. Initially at least, subscribers to the platform won't be able to focus on the sources, only on the resulting analysis that shows up on the online dashboard in the form of bar charts illustrating trends. On Wednesday, the day before the much-anticipated Twitter IPO, for example, the HedgeChatter main dashboard showed that in several categories Twitter was actually ranked second and third in terms of market interest behind electric car maker Tesla Motors (TSLA) and Apple. The information is broken down in myriad ways, including top stocks by news volume, stock price activity related to news coverage, top stocks just from tweet activity, tweet volume and increases in tweets and in addition, each stock is linked to rolling tweets, chatter and news reports on the company. The real value, for which subscribers will be charged of between $20 and $500 monthly, is that the platform offers projected performance of each stock if it held over various intervals. “Every day, 315,000 financial professionals turn on their Bloomberg terminals, and they're all looking at the same thing,” Mr. Ross said. “Our job is to filter out the noise, spam and duplicates.” Mr. Camillo compared online financial sentiment analysis today to online advertising in the late 1990s. “Back then, people were sure about the effectiveness of online advertising, but it took years for industries to figure out how to use it,” said Mr. Camillo, who is already planning the next evolution in the form of a hedge fund that invests based on non-financial Twitter chatter. “Historically, financial analysis has been sorted into fundamental and technical, and I believe sentiment analysis is really the third leg to that stool,” he said. “This is the start of a new era of financial analysis.”