“EMT is logically impossible and practically incorrect. Share prices NEVER incorporate OR reflect all relevant information.
First, all information is never available to everyone. IBM has more information available about it than could be read in a single human lifetime. Hence by simple logic nobody is taking it all into account; every analyst is considering differing parts and with differing motives and differing degrees of effort. So nobody has the same information and nobody has all of it.
Second, much information about a company is false. In enough cases, staff deceives CEO's, CEO's spin data to auditors and analysts, who then deceive brokers, who in turn deceive investors, whose actions are further distorted by the press. The resulting available information set has almost no logical chance of being entirely correct, aside from the fact it is too voluminous to be read.
Third, nobody looks at the data exactly the same way. Once investors receive their part of incomplete and sketchy data set, they do not assign the same importance or weight to each element, again making the resulting investment decisions vary widely from actual value.
So basically, share prices incorporate various degrees of ignorance, deception, fad, laziness, selectivity of review and other imperfections, the result is that they NEVER reflect all relevant information. They are always 'off'. Classical scholars will recognize EMT as Parmenides' theorem in disguise...and it’s just as stupid as his was 2500 years ago.”
McWillia’s first two points are spot-on, and obviously relevant in noting the incompatibility between theory and reality. But what I find even more instructive (laughable?) is that you can give those first two assumptions a pass – that ALL investors have ALL relevant information (obviously this idea is quite abstract, but the assumption would be that MORE information is a step in the direction of increased relevancy) – and you can STILL show that the idea of market efficiency does not hold:
In James Montier’s book “Value Investing” (as I’ve noted before, a truly fantastic collection of insights on investing), he discusses a study that examined the concept of information overload:
“Tsai, et al. (2008) show once again that, beyond a remarkably low level, more information translates into excessive confidence and static accuracy. They tested American football fans’ ability to predict the outcome and point spread in 15 NCAA games. The information (selected by surveying non-participating football fans) was presented in a random order over five rounds. Each round revealed six items of information (called cues)… A reward of $50 was promised to the participant with the best performance. In order to take part in the study, participants had to pass a test demonstrating that they were highly knowledgeable about college football.
With just the first set of information (six cues) the [computer] model was around 56% accurate. As information was gradually added, the predictive accuracy rose up to 71% by the time all available information was presented. So, from a statistical modeling point of view, more information was indeed better information. However, when dealing with humans rather than computers a very different result was uncovered.
Accuracy pretty much flatlines at around 62% regardless of the amount of information that was being provided. This performance is higher than the model on the earlier rounds, although not statistically so, but lower in the later rounds.”
With more information, results are essentially unchanged; beyond a few key sets of information, continued "data" reaches a point of irrelevance, and causes performance to flat line.
But what do the participants believe?
“However, confidence tends to soar as more information is added. [Mr. Montier discussed overconfidence in a 2002 article, calling it one of the two most common biases; the illusion of knowledge, which stems from overconfidence, is the tendency for people to believe that the accuracy of their forecasts increases with more information.] So, it starts off at 69% with 6 cues, and rises to nearly 80% by the time participants have 30 cues… This finding reflects the cognitive constraints suffered by the human mind…
The simple truth is that we aren’t supercomputers with unlimited power. Rather than trying to push the cognitive bounds of our brains, we should seek to best exploit our natural endowment. So, rather than collecting all the available information, we should spend more time working out what is actually important and focus on that.”
Mr. Montier’s opinion is that the investor should do two things: Find what is actually important, and focus time/analysis on that information.
So, what is actually important? This, in my opinion, is where a sustainable advantage is developed for the intelligent investor – finding what is pertinent and then following a disciplined approach to implement it; for the large majority of individual investors (of which I believe I am one), they will either not be able to discern what is truly pertinent across a vast supply of raw information, or they will not find the time/dedication to devote themselves to extracting useful information from this data and implementing it in an investment strategy.
I'm challenging myself and the readers of my articles to truly immerse themselves in a mission to define what is important; this MUST be a deep dive: for example, being a “dividend aristocrat” is not enough - we need to determine WHY a company is able to attain such a position in an intensively competitive marketplace and examine any potential events that may threaten their ability to continue to do so in the future. When people like Charlie Munger, Mohnish Pabrai and Guy Spier talk about checklists, this is what I imagine the core of their list seeks to answer.
A good place to start building an answer/checklist of sorts is by examining what has worked historically; conveniently, this information has been compiled by Tweedy Brown in a helpful document entitled “What Has Worked in Investing” (link here).
In the future, I hope to expand on this topic further and present my findings in an article; comments are helpful and encouraged, so please put in your two cents if you’ve leaned on any tools/ideas in investing that you’ve found particularly helpful!