State of Social Media in the Financial Sector

In the two years that Gnip has been working with the financial industry, the state of social data and social media in the financial sector has changed dramatically. Here’s a look at where the industry stands and where we think it’ll go.

Social Media Moves Markets

Perhaps the highest profile event to involve social media in the financial sector this year was the Hash Crash.

On April 23, the AP Twitter account was hacked and tweeted that two explosions in the White House had injured Obama. The result? The Dow dropped more than 140 points within two minutes. It was an eye opener for many, on the power of social media to move markets. But what got lost in the coverage of this incident was that Twitter was part of the reason that the market rebounded so quickly. What Gnip has seen time and again is that rumors on Twitter are defeated just as quick as a rumor is started. Immediately after the initial Tweet, many others started to debunk the rumor giving live reports from the White House and setting the record straight.

The Hash Crash provided another, audible and clear reason for using social data if you are a participant in the financial markets. Our hedge fund and asset management customers have known this for sometime. If you weren’t following and analyzing social, you were most likely slower than others to understand what was happening in the market – in the dark.

SEC and Reporting on Social Media

Another big change to shape the industry was an official clarification in SEC policy on social media from the Securities and Exchange Commission allowing companies to announce key information on social media as long as investors knew that such channels would be used. As of today, more than 150 companies are using social media to report financial results or performance. Real-Estate Tech company Zillow, Nasdaq:Z ($Z) took this concept even further, opening their earnings Q&A up to questions from Twitter. Earnings calls have always been intended to provide color and transparency for all investors and potential investors of a publicly traded company, but the reality has been that they have been events attended and monitored almost exclusively by investment professionals. Opening up the announcement and especially the Q&A portion to Twitter isn’t as much a radical new move as it is a use of new technology to help re-align these events with their initial intent to give everyone access to information on the company to make investment decisions.

Social Data in the Markets

When Gnip first started looking at the ways the financial markets could use social data, we never would have guessed how fast the market would grow and how hungry people would be for data. In two years, we’ve seen large-scale growth of large hedge funds using Twitter social data as part of their trading strategies. Twitter provides a broad based stream that can answer questions about sentiment about companies, brands, ideas and rumors.  Investors are finding value both through intelligent aggregation and data mining. When a merger rumor is breaking, you can find speculative deal values on Twitter before official numbers have been released. In addition to Twitter, financial institutions have found value in similar content from Stocktwits as well. Stocktwits has a curated community of financial investors who buy into sharing their thoughts online. Stocktwits has been especially valuable for traders and hedge funds who don’t want to sift through the noise on Twitter.  If you search for Justin Bieber on StockTwits, you won’t find anything.


And earlier this year, Gnip signed a partnership with Estimize, a crowdsourced earnings estimates platform that provides open sourced financial estimates with incredible transparency, making it a valuable and unique set of social data. Estimize has a platform to capture and provide structure around the long explored concept of a whisper number. They’ve recently added Vinish Jha, a former Starmine Quant, to help add a layer of intelligent analytics on top of the open community, and to really work towards an open estimate that includes only the most accurate prognosticators.

The Adoption of Social Data in Trading Terminals

One of the oft passed around anecdotes at Gnip is how financial institutions talk about traders and analysts using their iPhones under the desk so they can keep an eye on Twitter. Due to regulation, most banks or brokerages don’t allow traders to post or use social media. To enable traders and analysts to access social media (but not to post) a number of banks and terminal providers have been adding social data to terminals  – thus enabling users to at least look up conversations and research online.  In the case of Bloomberg, for now they provide a curated feed, so it isn’t always the complete and full conversations.

New Uses – Risk Management

Over the next two years the acceptance of correlations between stock prices and social data will allow for deeper insights. The area I see making the most progress is in risk management. A good portion of making money in investing is figuring out how not to lose money.   With the S&P on a 5 year growth run, it’s no secret that there is a risk of a pullback, the big question is when?

Social data allows for risk modeling that removes one of the inherent biases of price/volume based modeling. Price and volumes of a security or asset only move when investors are ready to take action. Social media volumes and sentiment move around thought and discussion. Given the hope that thought and discussion still generally precedes action in the strategy of most investors, there exists a huge opportunity to pick up on early, previously undetectable correlations between companies and concepts.  A quick teaser example below shows normalized rolling 24 hour Twitter volumes for 2 related securities (LNKD and FB) and two unrelated securities (LNKD and IBM).  In the next year I expect more companies to start looking at these types of correlations for risk management, both between securities and concepts like “government shutdown”.


So Where Are We Headed?

Many of the initial uses cases have been reading social media for actionable trade ideas. The growing number of firms trying to offer social media based signals shows the success in this area. The next 1-2 years will be about expansion in two directions:  improvements in implementation/standardization and expansion of insights. Now that social data has made it through the sandbox phase for certain applications, the focus turns to integrating with existing processes and data sets. The most successful aggregators and indicators will partner with exchanges and traditional financial data vendors to help their data flow through to existing trading and research systems making the information more broadly accessible and cheaper to implement. On the raw data side, more tools will emerge to standardize linking data back to existing security/company identifiers and accepted industry and index classifications.

Using social data in the financial sector is fast becoming a must have, not a nice to have.


When an Earnings Release Isn’t an Earnings Release

Going into an earnings release, it’s important to know what recent news may be already baked into the stock price and what recent events are being discussed. Blackberry provided a great example on how you can easily look at social media to get context around an earnings release.

On Friday, Sept. 26th, Blackberry released their official results which included a loss of almost $1bln and a decline in revenue of 45%. Despite those dismal results, shares of Blackberry were trading slightly up during the day. How? Those following the stock knew this was because of a partial result pre-release a week earlier and an offer for privatization mid week. This post uses an internal interface we built using Gnip’s new Search API for Twitter to show how social data tells a full story.

A quick general search for “Blackberry” can help easily show whether or not there has been any major news in the week leading up to the call. On the chart you can see multiple spikes in the week prior to the official release.

Blackberry Mentions on Twitter

By zooming in on the spikes and looking through some of the details you can find three major events:

  • An earnings/revenue pre-release
  • Early talk of a privatization buyout around $9/share
  • The full official release of results

Here’s an example of digging into the one of the spikes. I added a second term “earnings” to get more relevant results on the earnings pre-release.

Blackberry Earnings on Twitter

Looking at the price chart, you can see all three events reflected on the price of the stock during the course of the week. A huge hit on the initial earnings release, a recovery rally on the news of the buyout and little movement on the final official earnings release.

Aside from being an interesting example of monitoring news through social media velocity, this example provides a great case study of where understanding investor reactions and volume could be useful but need to be paired with additional information. In this case, stock price movements accompanied the first two high Twitter volume events, but the third event had very little price movement despite high volume and bad news. A strong example of where more analysis is needed to determine if trending news is already reflected in the stock price!

Quantifying Tweets: Trading on 140 Characters

Social media analysis… for traders? That statement 5 years ago elicited a much different response than it does now. The market now recognizes the importance – and the impact – of social media channels, and as such, has recognized the need to monitor and trade off the research created from that data.

One of the earliest and most important voices in this conversation was that of Joe Gits and the Social Market Analytics team. Which is why we are incredibly excited to announce that they’ve joined the Plugged In to Gnip partner program.

Social Market Analytics quantifies social data for traders, portfolio managers, hedge funds and risk managers. Their technology extracts, evaluates and calculates social data to generate directional sentiment and volatility indications on stocks, ETFs, sectors and indices – providing predictive indicators for clients. They have succeeded in turning qualitative text into quantitative indicators that can easily be incorporated into trading strategies – broadening the types of traders and firms who can now access and incorporate social signal into their decision making.

As shown in their recently announced agreement with New York Stock Exchange (allowing NYSE reseller and distribution rights to their sentiment feed), SMA is helping bring social analytics to a wider group of financial firms than has ever been possible. That client base requires the highest-level of enterprise reliability in the products they buy –  which means SMA’s product requires the strongest data foundation possible. And Gnip is honored to be the company providing the reliable and complete access to the social data that fuels this solution.

To see what their technology looks like, check out the webinar we recently held with them.

Tweets, Texts and Tickers

A look at social data in the financial markets with Tom Watson, Vice President of Global Market Data at NYSE Euronext; Brian Hyndman, Senior Vice President of Global Information Services at NASDAQ OMX Group, Inc., Rich Brown, Global Head of Elektron Analytics at Thomson Reuters and Heidi Johnson, Global Product Lead for Hub and Collaboration Services at Markit. 

Financial Social Data Panel

The use of social media data in the finance industry presents some inherent and unique challenges. This panel explored how social data could and should be used on Wall Street.

Leadership Vacuum
There is a great opportunity for leadership in this space, as no one entity is currently driving the charge on how to structure the use, the infrastructure or the verification of social data in financial markets. Firms need to focus on how to get this volume of data from the social channels and to the customers in order to trade and make investment decisions based on that data. Firms are receptive to using social data, but regulatory and compliance oversight make this tricky. Who is to say when information becomes public? Firms are looking for real-time data solutions but must examine this information within the context of historical models. Historical trends put real time data in perspective; both are critical.

How do firms vett social media accounts as the professional, official individuals and groups? A verification process or Klout-esque score or index is needed in order to confirm that the data from social channels or social sources are reliable, trusted, consistent: Is that the company or individual you think it is? Currently we passively consume this social data, but how can financial firms weed out false positives and anomalies? The nature of social data is that it moves so fast, people and companies react before verification can be made of a trusted source. Fake accounts can and have tanked stocks. Identity verification of expert, trusted sources is crucial. It’s not the first tweet, it’s the conglomerate of the tweets, blog posts, etc., but people often react to the first data they see instead of looking at trends and patterns, as well as the original source. Think snowball, not snowflake. Social amplifies data. Everyone has the ability to reach millions of people now on social media. Data needs to be corroborated with other sources of information.

“Hash Crash”
Earlier this year, the Associated Press Twitter account was hacked and sent out a falsified tweet about a bombing at the White House. The fake news event caused a real market event, dubbed the Hash Crash- a dip (and reversion) in the stock markets. But even from the start, a large proportion of overall Twitter conversation doubted the veracity of that tweet, and traditional news sources showed a different story. This disproved Tweet led to the V-shaped dip and recovery of the markets.

Social Data Examples in Financial Markets

  • Spikes in weather reports and crop prices
  • Violence in Iran and oil prices
  • Sentiment and Psychological index – fear, greed, optimism
  • Geospatial – Florida orange grove region and supply chain data

There is a great need for historical models, real time data analytics, data mining, and verification processes in the financial realm, and firms are receptive to finding these solutions.

Big Boulder is the world’s first social data conference. Follow along at #BigBoulder, on the blog under Big BoulderBig Boulder on Storify and on Gnip’s Facebook page.

Big Boulder Speakers Using Social Data in Innovative Ways

Big Boulder is next week and we’re excited to add four new speakers who are using social data in amazing ways, from disaster response and epidemic tracking to predicting the stock market and monitoring political developments.

If you want to follow the conversation about Big Boulder, be sure to follow the hashtag #BigBoulder , the Gnip blog for live blogging and pictures from the conference on our Facebook page.