Monday, January 02, 2012

Using Reason+Emotion to Forecast 2012

2011:  An Emotional Year
Needless to say, it was an emotional year in the financial markets, and most investors underperformed substantially.  Since emotional markets are why we're here at MarketPsych, this month's letter will share our latest insights, lessons, and fascinating charts to help our readers going forward.

First of all, uncertainty is guaranteed to continue, with Europe's debt unresolved, U.S. political gridlock, and a potential hard-landing in China.  Fortunately, as you read below, you will see that uncertainty is predictive - in a counter-intuitive way.

Before we launch into today's discussion - while our schedules are filling, we still have availability for consultations and trainings in the first two weeks of February, so please reach out if you'd like some personal or organizational attention.

The MarketPsych "Awards:"  Not What We Thought
In December we launched the 2011 MarketPsych Awards:  identifying the Most Trusted Bank (WTFC)Most Innovative Tech Company (SNPS)Most Loved Consumer Products (STKL), and others based on social media chatter in 2011. 

But a word of warning for investors about these Awards.  We ran a predictive analysis of what happens to top ranked companies going forward, and the results are mixed.  In fact, past winners of the Most Loved Consumer Products award under-performed the market by -2% on average the following year, while least-loved consumer product companies outperformed by 10%.  This effect - where loved stocks fall and the unloved rise - reflects an old Wall Street aphorism - a "great company" rarely makes for a "great investment."

So if loved companies aren't the best investments for 2012, then what does predict stock prices going forward?

Answering this question became our mission in December.  

Our analytic team aggregated the past 13 years of social media chatter by year for 8,000 publicly traded U.S. companies.  Then we identified the topics and psychological factors expressed online about the company that were most predictive of company stock prices in the following year.  

Some Background:  Over the past 8 years the MarketPsych team developed the most sophisticated financial textual analysis software commercially available.  Every day we quantify hundreds of sentiments, topics, and tones in 1.8+ million social and news media articles.
For our study we examined a universe of about 2,000 highly discussed stocks, each year since 1998, with market capitalization over $1 billion.  We ranked each stock for each year relative to all other stocks using our Big 21 Macro Sentiment Indices.  We then grouped the stocks into deciles for each year based on their sentiment rankings and averaged the returns for each decile over the following year, from 1998 through 2010.  

We compared returns vertically (stepwise across deciles),horizontally (across sectors such as Materials, Tech, Consumer, Healthcare, etc...) and over time to ensure consistency.  Nonetheless, this isn't a rigorous study, and it is susceptible to over-fitting.  We did it for fun and not for trading (we use a different in-house testing technology for investment strategy research).

Investors are Smarter When They Think...

Our indicator FundamentalStrength, which measures the amount of positive versus negative accounting mentions in social media was most predictive of future returns.  Stocks in the highest decile of positive accounting discussions (FundamentalStrength) showed 14% out-performance, while the lowest decile showed -5% market under-performance. The results were even stronger for tech stocks, with the highest decile of tech stocks outperformingthe market by an average of 18% annually in the subsequent year, while the decile with the lowest FundamentalStrength under-performed by -7%.  

Another index with interesting results was EarningsExpectations, which measures online predictions of positive versus negative earnings reports.  The highest decile - with most positive earnings expectations - shows 10% out-performance versus the overall market the following year, while the lowest decile shows -3% under-performance.

...But When Excited, Their Predictive Power Lags

Our initial results indicate that investors analytically discussing company fundamentals in financial social media can be predictive of future returns, but does their expressed emotion also predict future performance?  First, an investigation into the seasonal dynamics of investor emotion.

The Happiest Time of the Year
December IS the happiest and most optimistic time of the year for investors, based on our data.  The red lines below are the average optimism for every year over the past 13 in social media (using the X-12-ARIMA method and courtesy Aleksander Fafula, Ph.D.).
 To be sure we weren't just capturing mutual well-wishers on stock message boards and Twitter (Happy New Year!, etc...), we looked at the data again with only future-oriented statements - what we call Outlook (Optimism - Pessimism).  December is at the top (although not as dramatically) in that graphic as well.

And what month contains the most distaste?
 In addition to being low in Joy, September sees the highest average of hateful, spiteful and angry language over the past 13 years.

Forecasting with Emotion+Reason
Now back to forecasting with our Emotion Macro Indices.  We have nine of them, including Anger Joy, Uncertainty, and others.  
The most striking predictive results across sectors and years occurred for Anger (sadly a common emotion among investors these days) and Uncertainty.  High Anger stocks (top decile) underperformed the market by -2%.  Anger is generally reactive, so perhaps bad news is priced-in quickly.  Stocks with little angry drama among their investors (or at least little anger expressed in social media) outperform the market by 10% in the following year.

Another "emotion" category - defined as such for the anxiety it provokes - is "Uncertainty." The high-decile Uncertainty stocks outperform by 6% annually - perhaps due to investor reluctance to invest until the uncertainty is resolved ("probability collapse" in the jargon).  Low-decile uncertainty stocks - perhaps swarmed by overconfident and too-certain investors - underperform the market by -7% annually as their cherished stability is revealed to be more fragile than expected.

When the best of our indices are combined into a single ranking index, their predictive power increases and becomes more robust,  The combination demonstrates 9% outperformance and -11% underperformance for the top and bottom deciles, respectively, over the following year.  We then used this single index to organize predictions for 2012.

Predictions for 2012  

Potential outperformers based on our basic analysis are the following:
1.  Royal Bank of Canada (RY).  Canada has strong banks, the U.S. does not, what can I say?  Opportunity here (but beware commodity bubble in Canada).
2.  Seabridge Gold, Inc. (SA).  Gold miner - underperforming industry this year.  Rebound candidate?
3.  Tessera Technologies Inc. (TSRA).  Semiconductor Equipment.
4.  MGIC Investment Corp. (MTG) - Mortgage insurance (really?).  Stock down, housing market is recovering.
5.  Natural Resource Partners LP (NRP) - high dividend (7.9%) coal mine leaser.

For 2012 we see a few surprises as potential underperformers.  
1.  Yum! Brands (YUM) is at the bottom of our list, perhaps due to China exposure, where a bubble is predicted to deflate over the next year.
2.  McDonald's Corp. (MCD) is the second from the bottom.   Chipotle and other upstarts are eating MCD's lunch.
3.  Delta Airlines (DAL)Procter and Gamble (PG), and one I struggle especially to write, CenterPoint Energy, Inc. (CNP) round out the bottom five.  Centerpoint has a 3.9% dividend, positive price momentum, and a P/E of 6.  It's REALLY difficult to consider selling CNP short as a good idea.

Many of these were totally unexpected (verging on psychologically disturbing) stock predictions.

After creating our ranking system, we decided to test FundamentalStrength tempered by Joyusing ETFs.  In the following study, you can see our thought process in deciding how to investigate the role of emotion in financial prediction.

ETF Performance:  Buy Strong Fundamentals With Low Joy Expressed by Investors
When we looked at the FundamentalStrength Macro Index (recall it is an aggregation of positive versus negative accounting-related conversation), we found that it had variable predictive capability- it was rendered less effective when positive emotion was associated with high values.  Just as with the "Most Loved Consumer Products," investor excitement destroyed the predictive power of positive accounting news.

It turns out that when people are excited that they are going to make money, they have activation in the brain's reward system.  After all, if we feel excited and happy, it must be coming from somewhere.

Importantly, that same excitement leads to a reciprocal deactivation in the "loss avoidance" areas of the brain.  That is, when we're excited that we've found a good opportunity to make money, we become less able to detect risk.  Just when we need it most, our ability to see and process potential dangers is turned down.  All systems GO! with no ability to turn back.  The following image demonstrates this effect in the lab.

We decided to explore this effect among investors, so we generated a hypothesis.  Perhaps investors become too excited about the FundamentalStrength (accounting fundamentals such as earnings, returns, margins, balance sheets) for some sectors and industries, or perhaps their excitement about the sector causes them to project over-optimistic accounting results into the future.

To explore whether this was occurring, we gathered a group of the 40 most liquid long-only ETFsincluding sectors such as utilities and consumer staples, industries such as gold miners and clean energy, and indices such as the Nasdaq 100 and the Russell 2000.

We then looked at the performance of the ETFs over the next quarter for two cases.  We looked at ETFs with an above average FundamentalStrength AND above average Joy (a.k.a. excitement, enthusiasm) VS. ETFs also with above FundamentalStrength BUT with below average Joy.  The idea is that ETFs with happy investors are likely to have good news already priced in, and to have investors blind to risk, while the others are under-appreciated and more realistic.

Sure enough, we see exactly that effect among all ETFs with high FundamentalStrength - less-loved ETFs outperform over time.
The Best Leading Indicator

As we watch financial news, it is common for our unconscious emotional state to shift from news story to news story - across a spectrum from joy to anger to fear - depending on the message of each story.  As these shifts occur, the brain shifts its decision-making from primarily neo-cortex (analytical) to primarily limbic (emotional).  And most importantly, this shift to emotional decision making is unconscious.

After such a shift, our rational brain (neo-cortex) believes that it is responding to financial circumstances and information as before, but in actuality it has been "emotionally primed" and is reacting to events from within an emotional frame.  

We can see the consequences of this effect in our own behavior.  After hearing frightening news, we react with more caution, after joyous news, we take more risk.  The effect is evident in today's MarketPsych Report research.  When investors are using their neo-cortices to formulate analytical decisions based on earnings and fundamental data, their forecasts are generally correct.  When they use emotion to guide their decisions - buying a stock because they feel good about it - they underperform.

MarketPsych coach Richard Friesen uses "Mind Muscles™" exercises to help clients build on their decision strengths.  The "Awareness Muscle" exercise will teach you to become aware of when the locus of your decision process is shifting from the neo-cortex to the limbic centers.  Email [email protected] with "Awareness" in the subject line and we will send you the exercise.

Richard Friesen still has sessions available on Tuesday, January 10th at 11am ET and 1pm ET and on Friday January 13th at 4pm ET.  

Next newsletter we will introduce "Extreme Event Mind Muscles™" which will teach you what to do once you've become aware of an emotional shift.  Given 2012's financial and political risks, it's important to plan ahead now - no one wants to be the deer in the headlights!

We also have training availability for your firm or organization in the first two weeks of February.  Please contact Dr. Peterson or Dr. Murtha at [email protected] for more information.

Best Wishes for 2012!
Richard L. Peterson, M.D.
and The MarketPsych Team


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.  THIS ADVICE IS WORTH WHAT YOU PAID FOR IT.  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.