Imagine two groups of students post-graduation. The first group is informed that one of their teachers will go to Europe for a year for a sabbatical period. In this first group of students is asked to write what they imagine that the teacher will do during the year to come. A second group of students receives the same information that the teacher has done a gap year in Europe and these students are asked to describe what they imagine that the teacher did during that period.
At first glance, do not expect that there is a big difference between the reports of the groups, because it is representative and practically doing the same task. However, the differences in reports of the two groups were relevant. The reports of the past were much more detailed imagery and rich when compared to attempts to describe the future. The latter, always vague and generic. This study, conducted by Canadian psychologist Janet Bavel, and other experiments demonstrate that we have easier to imagine what someone did as opposed to someone who will. That is, there is a clear imbalance between our ability to explain the past and our ability to predict the future.
Once an event has occurred, even if it is something extremely rare, we humans have easy to find a plausible explanation. Terrorist attacks such as September 11, 2001, and major air disasters are typical examples, whose sequences of events become well known to just a point of time to be easily assimilated and even considered predictable in hindsight. However, even the day before, nobody is able to down them or avoid them.
The financial theory of behavioral styles that easily explain the history of "retrospective bias" (in English, hindsight bias). The financial media is full of specialists always available to explain the movements of markets in the year, month, or last week. Ask then to predict the future and the explanation becomes vague and generic, as in the study. Anyone can fit the past into a sequence of events perfectly rational and intuitive. Now, predict the future is another story.
How accurate is the Brazilian market in an attempt to predict the path that the main economic and financial variables will go in the future? Every week, the Central Bank publishes such projections in the Focus newsletter. Does this group of analysts can anticipate with some accuracy and 12 months ahead of the GDP, inflation, nominal and real interest and exchange? A quick look at the difference between the forecasts made 12 months ago and the values actually observed for these variables showed significant errors, between 40% and 12%. In addition to these two variables, nominal and real interest rates, the average expectations of experts was unable even to predict the direction in which they would go. A year ago, predicting an increase of these variables, when in fact we have seen significant decline of both. The same trend was observed for the "Top 5" in principle the most accurate in their forecasts, even with a minor deviation.
Many will say that in that period the world was affected by financial crisis and that makes the forecasts less accurate. However, for other periods in which predominated more normal (as between 2003 and 2007), or without a crisis so serious, the margin of error of the forecasts proved to be too large. In fry eggs, which move with the market is exactly the unexpected! All that is predictable and does not change the expected price. Is already largely built on expectations of economic variables and prices of assets. The opportunity to profit is anticipated that the other not yet see, to make it faster and more efficiently than the other and translate that into positions in certain assets, the "tip" right (bought or sold) before the others do. In this respect, the newsletter of the Central Bank Focus acts more like a picture of the feeling prevailing in the economy at any given time than as a source of reliable estimates of these indicators or even the tendency to assume that in the future.
During the planning of flight of the Apollo 11 mission to the moon, NASA set the acceptable margin of error in initial estimates and calculations in 1961 the level of 1%, which was then reduced to 0.5%, 0.2% and finally 0.05% in 1969, the year of the famous first flight to the moon. When NASA needed to rely on the accuracy of the forecasts of the newsletter Focus, the flight of Apollo 11 probably would not have reached the height of a Saint-John's balloon ...
Aquiles Mosca, strategist for personal investments and business executive superintendent of Santander Asset Management. He is the author of the book "Investment under measure" and "Behavioral Finance."
E-mail: aquiles.mosca @ bancoreal.com.br
Nenhum comentário:
Postar um comentário