Video - Can Twitter Predict the Stock Market?

Can Twitter Be Used to Change Economic Direction?

Introduction  

• Analysis of social networks for emotional content suggests that mood is a predictive element for movements in the national economy ("Can Twitter predict the future?", 2011)
• Social networks are subject to being influenced by surprisingly small number of individual nodes within those networks (Eherenberg, 2011). 
• Those nodes can be identified and directly influenced themselves ("Untangling the social web", 2010).
• Consumer confidence is a key indicator in driving consumer demand, and consequently a key element in aggregate demand. 
• Governments could increase economic activity during a downturn by influencing national mood via social networks rather than by engaging in costly Keynesian stimulus spending.

Background

•  During an economic downturn, governments can help push an economy out of recession into recovery by increasing government spending.
• Keynes theorized if government spending increased, it could increase economic activity, setting the economy on an upward trajectory.
• Keynes believed "animal spirits" played a substantial factor in driving economic activity (Ruffin & Gregory, 1990, p. 175).
• "Animal spirits" are now generalized through the entire population via economic measures such as the consumer confidence index (The Conference Board, n.d.). When consumer confidence is high, one expects people will spend more increasing aggregate demand.
• Using social networks to obtain information about the members of those networks is not new (Crosman, 2011).
• Dr. Bolling's research takes the emotional content of messages sent using the Twitter system and uses this data to construct an overall national "mood" determining if people using Twitter are generally happy or sad ("Can Twitter predict the future?", 2011).
• When the national mood shifts, the Dow Jones Industrial Average does too ("Can Twitter predict the future?", 2011).
• A hedge fund is being launched using Dr. Bollen's Twitter algorithm to guide its trading ("Twitter research promises trading success", 2011).

Potential Benefits

• The connection between the national mood and the market offers societal benefits if the national mood could be affected.
• One must influence only a small portion of a network in order to influence the whole, which seems to be especially true of social networks. According to Ehrenberg (2011) 20 percent of a social network like Facebook or Twitter must be affected to affect the whole.
• Many companies use network analysis to identify "influencers" in to target their marketing efforts to them. ("Untangling the social web", 2010).
• This type of analysis could be used to determining which members of social networks are "influencers" with respect to mood.
• Keynesian style stimulus are expensive. The American Recovery and Reinvestment Act ARRA cost roughly $787 billion dollars.
• Identifying and altering the mood of "influencers" in social networks would change the mood of the entire network possibly changing economic activity for less cost than a Keynsian stimulus.

Legal and Ethical Issues

• Use of social networks to influence national mood delivers a substantial tool of public policy to the government that would effectively be invisible.
• If the government could drive the economic direction of the national economy via influencing social networks then it would be able to engage in actions that affect the economy more subtly, and almost clandestinely, which raises concerns that this ability could be overly used, abused, or misused.

Security Concerns

• Danger that any organization with sufficient resources and determination could attempt to shoift national mood.
• An organization hostile to a particular country could attempt to influence its national mood towards increased anxiety and depress economic activity.
• A political organization might attempt to depress economic activity prior to an election if they believed that an economic down turn would be to the benefit of their preferred candidates for office.

Social Problems

• Many individuals react quite negatively when their social media data is used for commercial purposes, even when it is used to provide additional products that they enjoy more efficiently (Guyunn, 2011).
• Social media data can be used to violate the privacy of the individual.
• Possible resentment against using social media data to control national mood by influencing social networks, could impair, negate, or even result in the antithesis of the intended effect.

Further Research

• Whether these various moving parts would work together has yet to be researched.
• Entirely speculative whether intentionally attempting to affect national mood via influencing social networks is possible
• Open question whether intentionally affecting national mood will result in movements in the direction of the economy. An artificially induced shift in national mood may not have the same impact on economic activity as a spontaneously occurring shift.
•  Could an artificially induced shift in national mood could be successfully accomplished without provoking a negative reaction that would counteract the intended effect.

Conclusion

• Use of twitter to assess national mood is potentially a powerful tool to predict movements in the stock market.
• Connection between national mood and economic activity could potentially be used as a powerful policy tool.
• This tool could be put to malicious uses.
     

References

Boyde, E. (2011, May 9). Twitter research promises trading success. Financial Times, p. 16.

Traders plan to use Johan Bolling's method of assessing national mood by analyzing tweets to make trades in the stock market.

Can twitter predict the future? (2011, June 4). The Economist, 399(8736), Technology Quarterly p. 12.

Indiana University professor Johan Bolling has discovered that tweets can be used to determine national mood which can be used to predict the movement of the Dow Jones Industrial Average.

The Conference Board | Trusted Insights for Business Worldwide. (n.d.). Retrieved June 19, 2011, from http://www.conference-board.org/

The non-profit Conference Board gathers information to produce the consumer confidence index and provides analysis of the resulting information.

Crosman, P. (2011, May). Mining the social network. Bank Systems & Technology, 48(3), 16. Retrieved June 19, 2011, from ABI/INFORM Global. (Document ID: 2334689731).

Banks use social networking to obtain information about their customers in order to more effectively target marketing towards them.

Eherenberg, R. (2011, June 4). A few master switches can rule a network. Science News, 179(12), pp. 5-6.

The actions of a small subset of a network can influence the entire network, and social networks are especially easy to influence in this way.

Guyunn, J.  (2011, April 22). Mining of Facebook data a two-edged sword; Marketers like ability to target ads; privacy watchdogs sound alarm. Edmonton Journal, E.2.  Retrieved June 22, 2011, from Banking Information Source. (Document ID: 2328211221).

Corporations mine social networking data to target marketing towards individual consumers, but consumers react negatively to these efforts.

The Recovery Act. (2009, February 17). Recovery.gov. Retrieved June 19, 2011, from http://www.recovery.gov/About/Pages/The_Act.aspx

The official website for the American Recovery and Reinvestment Act of 2009 including detailed information about the purposes of the Act.

Ruffin, R., & Gregory, P. (1990). Principles of macroeconomics, fourth edition. Glenview, Ill: Scott, Foresman and Company.

A basic text on macroeconomics, with relevant sections related to Keynesian economics including stimulus spending and "animal spirits".

Untangling the social web. (2010, September 4). The Economist, 396(8698), Technology Quarterly pp. 16-17.

Corporations, including cell phone companies, analyze social networks to identify individuals who are especially influential and target marketing towards them.