Facebook, Social Unrest, and the Predictive Power of Big Data Feb 2012, v2, n1
"The Extraordinary Popular Delusion of Bubble Spotting." November 5, 2011. Jason Zweig. Wall Street Journal.
November 28, 2011. Ted Schwartz. ABCNews.com.
November 26, 2011. Ben Levisohn. WSJ.com.
"Bull Market Bear Market Bring It On!" WSJ.com. October 15, 2011. Ben Levisohn.
List of Past Press.
Big Data and Social Prediction
At MarketPsych we are at the vanguard of social prediction in the financial industry. We developed predictive models that not only are monitoring the buzz about Facebook, but that use the characteristics of that buzz to predict (and trade) Facebook’s shares. We have spent decades modeling social data and trading on it - our MarketPsy Long-Short Fund LP was successful in using this data to outperform the S&P 500 by 27% from launch on September 2, 2008 to the end of 2010 (when we closed it). It’s not only our firm that does this - there is an entire industry of social prediction based on mining behavioral and psychological data.
But we and many of our customers have problems with Big Data – it’s unwieldy and noisy. Consider that our Macro data feed distributes 30 sentiment and topic data points every minute for major economic sectors and industries (40), Commodities (60), Countries (100), and Currencies (50) in two feeds (one feed derived from social media and one derived from news media). In case you didn’t do the math yourself (and I hope you didn’t!), we’re releasing 15,000 data points minutely. This data deluge is actually a summary of our core data. It is derived from 2 million daily articles, analyzed for the presence of 40,000 entities, and scored into 1600 sentiment and topic combinations – all condensed into the 30 Macro indices to render it “usable” by humans. Here’s the general idea:
Predictive analytics is occurring in every industry. One company - from whom I recently received a “we will hire anyone with a pulse who has statistical skills”-type email - Accretive Health (NYSE: AH) reports mining 1 billion health insurance claims in order to identify trends that can be arbitraged to reduce health care costs while increasing health care delivery. But this newsletter isn’t about Big Data companies per se, it’s about social prediction and how we can do better in our own decision making using the insights we are gleaning from social data. More about Big Data generally can be read in this McKinsey study and here is a great summary of companies in this space on Quora.
While it appears cutting edge, it turns out that social prediction using big data is old news.
Isaac Asimov and Psychohistory
The other day, after the fifth person in a week told me, “sounds like your data is recreating Isaac Asimov’s Psychohistory,” I decided it was time to investigate.
Now I’m not exactly a science fiction fan, but I do enjoy a good story regardless of whether it occurs in our solar system or Qo'noS. In high school I occasionally wrote book reports on science fiction novels (lenient teachers!). But it was largely a matter of finding Jane Austen a bit too stuffy , not a personal fascination with the other-wordly (Prime Radiant? Psionic Suppressions?), that drove such reading.
So today in Wikipedia I read, “Psychohistory is a fictional science in Isaac Asimov's Foundation universe which combines history, sociology, and mathematical statistics to make general predictions about the future behavior of very large groups of people, such as the Galactic Empire.” Cut off the Galactic Empire bit, and that is EXACTLY what we’re doing at MarketPsych. Go figure. I guess we’ll have to withdraw our “Universal People Prediction Device” patent application – Asimov beat us to it.
(I actually started reading the third book of the Foundation Trilogy in eighth grade. But as it’s best not to start a trilogy on the third book, and the concepts were obtuse to me, I put it aside. But I wonder if it didn’t plant a seed…)
Macro Indices: Country-level Social Unrest
In addition to economic sectors, stocks, and commodities, few people know that we also monitor social and psychological phenomena by location - cities and countries. For example we have indices of “Social Unrest” related to negative chatter about national governments, authorities, and business leaders in social media. It’s likely this data will prove to be predictive of social events, although we may not have enough unrest incidents (thankfully) to test it thoroughly.
Per an informed source, China experienced 160,000 significant social protest actions last year (from tens to hundreds of thousands of citizens participating). Many of these actions could be (briefly) detected in social media, but they were not revealed in news media due to Chinese government concern about social stability. As a result we are not only tracking news media unrest mentions (since these are censored), but perhaps more importantly, we pick up on psychological predictors of unrest in social media. Such predictors include expressions of anger and frustration towards authorities and mentions of personal hopelessness (a positive predictor of impulsive violence).
Here is our top ten list of Countries with the most buzz about Social Unrest in the English-language media recently:
While social unrest is related to social forces often out of our immediate control, we can gain control how we prepare for and manage ourselves when experiencing setbacks if we understand our propensities to reaction. Genetic and hardwired cognitive biases play a role in our responses to loss, as you can see in today’s Researcher’s Corner, but such factors are by no means deterministic. We can intelligently use our minds to make better decisions.
Researcher’s Corner: Trader Genetics
Our good friends Steve Sapra, Ph.D. and Paul Zak, Ph.D. Director for the Center for Neuroeconomics Studies and author of the forthcoming “The Moral Molecule” this week published a new study of trader genetics: “A Combination of Dopamine Genes Predicts Success by Professional Wall Street Traders.“
The authors took genetic samples from 60 Wall Street traders in 2008 before the market meltdown. They found that traders with genes conferring moderate dopamine tone were more likely to have longer careers as traders. The longetivity of traders (the study’s dependent variable) was correlated with fewer D4 receptors and also less catabolic (enzymatic breakdown) of dopamine, theoretically leading to higher tonic levels of dopamine and less variability per receptor over time. This dopamine environment might lead to more stability during the ups and downs of trading and should be correlated with behaviors such as less trading in volatile markets (which the authors found) and perhaps more emotional and analytical equilibrium.
The authors conclude: “Combining the personality analyses and genetic findings from the present study, reveals that our sample of traders are analytical, integrative, and can delay gratification. They have a genetic profile associated with balanced levels of dopamine.” As we gather more data we may see additional interesting results.
Researcher’s Corner: Blind Spots in Financial Ethics
Last Friday I attended a talk by one of my academic heroes – Harvard GSB professor Max Bazerman – on the subtle cognitive biases that underlie unethical behavior. He has a new book out called Blind Spots which is a fascinating account of his research. Bazerman introduced the topic by explaining that MBA students are taught ethics through case studies that profile high-impact criminal behavior, yet the vast majority of unethical behavior is not obvious or dramatic at first. It is subtle, gradual (slippery slope), and permitted to some degree by observing others. As he points out, the standard MBA ethics education – which focuses on high-profile criminal cases, but not common unethical behavior - is inadequate. In some ways Bazerman’s research gets at the underlying causes of the financial crisis and contributes to our understanding of why no one has been criminally convicted.
Bazerman provided behavioral evidence of such socially important biases as 1) The unethical practices of auditors - by which independent auditors are blinded by their desire for ongoing income into gradually “cooking the books” of their clients, 2) The slippery slope by which most unethical behavior progresses, as our minds gradually rationalize misbehavior, 3) The process by which observers will permit minor unethical behavior with no compensation or kickbacks at first, ultimately putting themselves in a bind as serious violations mount and they feel unable to speak up after a history of silence, 4) How people judge ethical violations as less “bad” than ethically proper behavior that ends with worse result, 5) People judge anything that personally benefits them as less unethical, 6) Losses accelerate unethical behavior – similar to the disposition effect – in which people are also more likely to break the rules to get back to even, 7) And amazingly to me, people judge others hiring a proxy to do their ”dirty work” as significantly less unethical than doing the dirty work oneself.
Some of these biases explain, in part, why rogue traders begin and can escalate their misdeeds – collaborators looking away with good results (at first) lubricating further off-the-books trades, a slippery slope of small accumulating losses accrues, and losses accelerate unethical rule-breaking (disposition effect), and finally the climactic blow-up occurs. The same process explains the social, financial, and government biases that gradually allowed the housing bubble and bank risk-taking to grow to huge proportions.
As Bazerman points out, simple self-awareness only slightly ameliorates most cognitive biases. Per Bazerman the best (and only proven) solution to such biases is to reform institutions to help us be less harmed by our biases – the “Nudge” example of Richard Thaler. I would add that 1) learning about the nature of biases and 2) gaining specific self-awareness of biased decisions - by working through bias examples in one’s own decision making including practicing alternative decision strategies – also appears to reduce biasing.
Escaping the Pull of the Herd
Given what we’ve learned above – that IPOs lure naïve investors in due to hype and fear, that big data and social prediction is altering our future trajectory, that researchers have identified that moderate tonic dopamine levels can prevent us for falling for such hype and volatility as that around IPOs (and thus be better investors), and that our biases and Blind Spots are prevalent and explain some social trends and events like the financial crisis. The result of all this data and new understanding is a science of prediction (once called PsychoHistory by Isaac Asimov). Given all this, what are we as investors to do?
Based on emerging evidence of the role of genetics and hardwired biases influencing decision making and performance, in combination with awareness (always a first step), external behavioral “nudges” are the best tools we have for setting up an optimal bias-lite decision environment. Secondly is practicing self-awareness and working through biases in one’s own decision making (at MarketPsych we have developed workbooks for this practice).
Let’s consider an example of how we can use our knowledge of biases to improve our decision making. We know from Bazerman’s research that financial losses predispose us to violate rules (e.g., not honoring stop-loss rules, anyone?). One solution to this misbehavior is to plan ahead – one can enter automated stop-losses at the time a buy order is entered, before they are put in the position of biased thinking such as, “Will this falling investment rebound? Of course it will. Clearly there’s no need for that silly little stop-loss at this level.”
For financial advisors one application relates to their increasingly common role as an emotional coach for their clients. Advisors often need a nudge to engage with clients when they themselves are feeling beleaguered by falling markets. Advisors can set up their daily routine to incorporate simple tricks like seeing clients in the mornings or have interactions scheduled after refreshing activities like exercise or enjoyable activities. (We have many more ideas for advisors – contact us for more information).
Talks in February
We will be speaking in the San Francisco Bay Area next week (two classes at Stanford and two progressive corporations) as well as in New York at a Risk Management conference and in Chicago in February 2012. Contact us if you’d like to attend or you are in one of those cities and would like to meet.
We also have speaking and training availability for your firm or organization in late April. Please contact Dr. Peterson or Dr. Murtha for more information.
Best Wishes for 2012!
Richard L. Peterson, M.D. and The MarketPsych Team
Both books named "Top Financial Books of the Year" by Kiplingers.
Both books named "Top Financial Books of the Year" by Kiplingers.
Who We Are
2400 BROADWAY, SUITE 220 - SANTA MONICA, CA 90404
+1 (310) 573-8523
This material is not intended as and does not constitute an offer to sell any securities or a solicitation of any offer to purchase any securities.
The information of the MarketPsych Report is presented free of charge. It is no substitute for the services of a professional investment advisor. Investments recommended may not be appropriate for all investors. Recommendations are made without consideration of your financial sophistication, financial situation, investing time horizon, or risk tolerance. Readers are urged to consult with their own independent financial advisers with respect to any investment.
Past performance is no guarantee of future results. Screen and model signals and related analysis are for informational purposes only and should not be construed as an offer to sell or the solicitation of an offer to buy securities. Most financial instruments (stocks, bonds, funds) carry risk to principal and are not insured by the government. Anyone using this newsletter for investment purposes does so at his or her own risk.
Data accuracy cannot be guaranteed. Opinions and analyses included herein are based on sources believed to be reliable and written in good faith, but no representation or warranty, expressed or implied, is made as to their accuracy, completeness, timeliness, or correctness. We are not liable for any errors or inaccuracies, regardless of cause, or for the lack of timeliness of, or for any delay or interruptions in, the transmission thereof to the users.
As a matter of policy, we may act upon the investment information that this newsletter provides prior to making it available to the public. We do not accept compensation of any kind from any companies mentioned herein.
MarketPsych is not responsible for any special, indirect, incidental, or consequential damages that may result from the use of, or the inability to use, the Information contained on this newsletter whether the Information is provided or otherwise supplied by MarketPsych or anyone else. Notwithstanding the foregoing, in no event shall MarketPsych total liability to you for any and all claims, damages, losses, and causes of action (whether in contract or tort or otherwise) exceed the amount paid by you, if any, for accessing this newsletter.
MarketPsych expressly disclaims all warranties and conditions with regard to the Web sites, their Content, and the Information, including, without limitation, all implied warranties and conditions of merchantability, fitness for a particular purpose, title, and non-infringement. By using the Web site, Content, and Information, I assume all of the risks associated with their use, and I release and agree to indemnify and hold harmless MarketPsych from any and all liability, claims for damages, and losses arising from or connected with such risks.
IF YOU DO NOT AGREE WITH ANY OF THESE TERMS AND CONDITIONS OR FIND ANY OF THEM TO BE UNACCEPTABLE, SIMPLY UNSUBSCRIBE FROM THE EMAIL LIST. If you understand and accept these caveats, feel free to read the newsletter.