Sentiment Analysis: dealing with perception in finance – friend or foe?

In our daily lives we all handle a flood of information from various sources (environment, work related, social interactions etc.). Our understanding and follow-up reaction to such information is based upon how we identify, process, and understand it – namely based on our perception of it.

While this perception appears to come very naturally to us in daily situations, things are actually more complicated than that: all the information we receive is processed in our brain by various selective mechanisms such as learning, attention, pre-existing rules, intuition, and implicit expectations. As explained by Daniel Kahneman [1], our mind employs two thinking systems: Fast (intuitive) thinking and Slow (reasoning) thinking. As Kahneman points out, not surprisingly, humans like to identify themselves with the “conscious reasoning self that has beliefs, and make choices, and decides what to think about and what to do”[1]. However, as he further mentions, intuitive (fast) thinking is the main source for our explicit beliefs and choices subsequently used for our reasoning type (slow) thinking. Now that questions the human belief  that decision making processes are logic/reasoning driven!

When it comes to the financial trading area, the well-established link between perception and trading is highlighted by zentrader.ca [2] that states: ‘Perception is your reality in trading’. From Peracton’s experience, such statement is indeed very commonly accepted by traders and investors all over the world.

Figure 1 below is an example of financial information representation typically used by traders and investors. It is highly visual and mainly intuitive, where some basic reasoning is employed for decision making. Such reasoning is generally focused on what stocks to buy or sell and when.

While such charts look complex and very scientific, in fact they track just a few parameters that human brains can follow with relative ease. The below example has three financial parameters – Moving Average (50 days), Moving Average (200 days) and Volume’s variation over six months. These parameters are used in conjunction with two lines (red-resistance, blue support) that are usually drawn by hand in an intuitive manner, based upon where the trader ‘thinks’ there could be a ceiling or a bottom of price based on his/her perception of what is happening to that particular stock. The above mentioned moving averages and volumes are all inferred from price that is in itself a perception human traders and computers programmed accordingly have about the possible value of a share and implicitly, of a company’s value (market capitalisation).

Based upon some (quasi)simultaneous event occurrences, the trader/investor then makes a call and decides to buy or sell a particular stock. Such decision calls are also related to past memories of the trader where similar situations may have led to specific outcomes that influence today’s decisions.

Figure 1. LinkedIn Stock Chart

From the above discussion some conclusions can be drawn and open questions can be asked:

  1. We build our rational decisions on many factors of pure intuitive origin, which we are not aware of. How can we control the intuitive input and be aware of the biases we introduce in our decision processes?
  1. Perception (that is subjective) leads us to intuitive conclusions (also subjective and prone to error) that are hard to define as right or wrong. How can we figure out when perceptions serve us well and when they don’t?
  1. Humans employ formal ‘reasoning’ (slow thinking) that again appears to be used naturally. However, there is no formal reasoning over the reasoning itself. Why do we reason in manner A and not manner B or C? How can we step back and employ the most appropriate model?
  1. In trading, ‘perception’ equals reality. Such assumption, while comfortable to make, may serve well in some circumstances and not so well in others based upon many uncontrollable factors. Making it a generic rule can be risky. How can we differentiate between the right and wrong circumstances?

Going back to origins, trading, and investment have evolved enormously in complexity, size, and traders/investors’ behaviour since the 1920s when the first methods and principles of Technical Analysis were laid down. In 1948 Robert D. Edwards and John Magee published ‘Technical Analysis of Stock Trends’ that remains in use to date.

An update of tools, methods, and reasoning capacities is obviously required to catch up with the realities of the highly computerised present financial markets. In this respect, in SSIX project we are testing the first prototype version which will harness the market social sentiment. This will be combined with fundamental and technical factors in order to separate false and real signals for trading and investment by means of automated tools and reasoning algorithms.

We are just scratching the surface of a new way of looking at things and more questions will appear, questions that can be answered by going beyond classic fields and combining multi-disciplinary approaches such as IT-big data, finance, socio-economics and psychology.

 

[1]    Thinking, Fast and Slow, Daniel Kahneman, published by Farrar, Straus and Giroux 2011

[2]   Perception is your reality in trading – zentrader.ca http://www.zentrader.ca/blog/perception-reality-trading/ retrieved October 2015

For more information on SSIX, visit our website ssix-project.eu.

For the latest update, like us on Facebook, follow us on Twitter and join us onLinkedIn.