Disputed loud and long
Each in his own opinion
Exceeding stiff and strong
Though each was partly in the right
And all were in the wrong!”
The poem Blind Men and the Elephant by John Godfrey Saxe used to be one of my favourites during school days. It’s an old parable from the Indian subcontinent about six blind men, who have never come across an elephant before. When they actually do, they try to learn about it by touching it. Each blind man is feeling only one part of the elephant’s body and inferring what it is like. For example, the blind man who touches the leg of the elephant proclaims that an elephant is like a tree. They describe the elephant based on their limited experiences and their descriptions are different from each other.
The men end up in a loud argument; each of them having partial information; each one of them overconfident of their experience and knowledge, and each believing the others to be wrong.
Even though the poem is meant for a child in school, it has important observations for many disciplines. More so, in investment research!
Rush to Collect Information
The initial days of this calendar (so much has happened, they seem like far behind!) were characterised by uncertainty about the spread of the virus, multiple theories about how it can impact businesses and consumers, fear among investors and panic in markets. The only parameter that had some predictive value for the market during those days was the number of Google searches of coronavirus. There was extremely high negative correlation between the number of Google searches and the equity markets. In the last month, however, things changed.
As lockdowns eased, analysts and investors seem to have entered a race to collect more and more real-time information about what is happening ‘on the ground’ in various businesses. I haven’t seen a day when some research house hasn’t sent me takeaways from channel checks, excerpts from company interactions, management commentaries, dealer surveys or consumer surveys.
Add to that numerous anecdotes that start with “I know someone who…” It feels good to have more and more information, but how good are we at making good decisions based on this information? Turns out not much!
We make many mistakes in judgement like the six blind men.
The Errors
In the ancient times, heuristics or ‘rule of thumb’ helped humans make quick decisions and survive under adverse conditions. However, using more and more of these to make judgements leads to biases creeping into the decision-making process. I have highlighted below some of the important biases that can affect our decisions while collecting and analysing data during the current crisis.
Strength & weight of information
While determining a trend or testing a hypothesis based on new incoming data, one has to check its strength or extremity of the evidence and its weight or the predictive validity. Suppose we wish to evaluate the evidence for a hypothesis that a coin is biased in favour of heads rather than tails. In this case, the proportion of heads in a sample reflects the strength of evidence for the hypothesis in question, and the size of the sample reflects the credence of these data.
In India, let us say we want to test the hypothesis, that consumer durables’ sales were sharply lower due to Covid-19 related issues. We check with three large consumer durables dealers across India, and two of them confirm our predictions. Here the strength of the evidence is high, but the weight is low. But if we can survey 33 large dealers across the country, and 19 of them confirm our hypothesis; here the strength is comparatively low but the weight and its predictive validity is higher.
Studies conducted by psychologists Amos Tversky and Dale Griffin showed that we tend to be influenced more by strength of the evidence, than its weight. This leads to a couple of other problems:
- When the strength of the information is higher than its weight, we are likely to be overconfident about our conclusions
- We tend to be happy using small data sets. We may draw wrong inferences if the sample set does not represent that population correctly.
- Ignoring base rates
We often ignore the background data or trend in the presence of new evidence. Let us take the example of the residential real estate. Suppose a set of developers, even pre-Covid-19, were experiencing falling demand and footfalls. To access the exact impact of the new crisis, one has to compare the incoming new data with the sluggish background trend (base rates). Even without the crisis, one would not have come across any good news.
Similarly, in the case of sectors that display sharp seasonality, one has to account for background data during these months. For example, cement sales are typically low during monsoon. Ignoring these background numbers or base rates can lead to wrong conclusions.
The Personal Touch
Most of us give importance to the information that we perceive directly (that we hear or see with our eyes) than information received second hand, even though the latter may have greater evidential value.
Events that people experience personally are more memorable than those they only read about. When we personally visit a place or conduct a survey, we form certain impressions, make certain observations.
We hold these observations as treasures to be protected than hypothesis to be tested. Cognitive dissonance sets in and we rationalise and disregard any contradictory information. If the dataset we have referred to is small or lacks diversity, we may end up making wrong conclusions (same as the blind men).
Anecdotes and Case Histories
In a research paper, social psychologists Richard Nisbett and Lee Ross showed that we put high weightage to anecdotes and case histories than statistical data. Anecdotes are vivid, more emotionally interesting, easier to remember and easier to deal with for the human mind.
Let us take the example the researchers used: Suppose you wish to buy a new car and have decided to purchase one of the two brands -- A and B. As a sensible buyer, you go to consumer reports, which inform you that the consensus of their experts is that Brand A is mechanically superior, and the consensus of the readership is that the Brand A has a better repair record. Armed with this information, you decide to go for Brand A. In the interim, however, you go to a party where you announce this intention to an acquaintance. He reacts with disbelief and alarm: ‘Brand A! You’ve got to be kidding. My brother in-law had a Brand A. First, that fancy fuel injection computer thing went out. Cost a lot of money. Next he started having trouble with another part. Had to replace it. Then the transmission and the clutch. Finally, sold it in three years for junk.”
Although a logical, dispassionate response to such an encounter ought to be a very slight revision of the statistical data contained in consumer reports, few people will so act.
Availability Bias
The availability bias is a mental shortcut that relies on immediate examples that come to a given person’s mind when evaluating a specific topic, concept, method or decision. Suppose you conduct a survey of businesses in Mumbai, which have the highest number of Covid-19 cases in the country (about 16% of the India total), the answers you get are likely to be gloomier than say Chandigarh, which has very few active cases. People in Mumbai would be hearing about the problems of the pandemic with more frequency than people in Chandigarh, hence these issues are at the top of the mind recall for them.
Using Information the Right Way
A lot of time, effort and resources are invested in collecting information to make decisions. It is important that this information is understood and processed in the right manner. Using the right sample set (in terms of size and diversity), determining and anchoring on base rates and conducting unbiased (and statistical) analysis can lead to better decisions.
"bad" - Google News
July 08, 2020 at 04:09PM
https://ift.tt/2O7QVEK
Often wrong, but rarely in doubt! What really leads to bad investment decisions - Economic Times
"bad" - Google News
https://ift.tt/2SpwJRn
https://ift.tt/2z7gkKJ
No comments:
Post a Comment