We tend to look for patterns and relationships between variables, whether is it in investing or in other aspects of our lives. Perhaps, it is simply humans’ primal need to understand complexity, or just the way we were taught at schools. More often than not, we only analyse the relationship between two factors and this can lead to some misleading conclusions.
I can think of three reasons for this tendency. One; a 2 factor chart is can be easily understood by almost everyone. Two; for practical reasons, an X-Y graph is easy to present on print. Can you imagine showing a 3 dimensional graph on paper? Lastly, not many people will get exposure to the concept of multi-factor regression, much less understand it. Naturally, because of these factors, I believe we have a tendency to look at relationships with a 2-dimensional view.
I recall coming across an amusing, but apt, quote about statistics.
Continued from part one... Q1 hedge fund letters, conference, scoops etc Abrams and his team want to understand the fundamental economics of every opportunity because, "It is easy to tell what has been, and it is easy to tell what is today, but the biggest deal for the investor is to . . . SORRY! Read More
“Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital”
Sometimes, the third factor (which is what is concealed), can dramatically change the dynamics of the entire situation. In statistical parlance, the addition of a new variable may change the coefficient of an existing independent variable. But enough of geek talk, here are 2 examples which I came across recently.
Roy Ngerng on Wages: How PAP has betrayed Singaporeans
Politics aside, Roy asserts in the article that Singaporeans are disadvantaged because we have one of the lowest wages in the world. One argument he presented was that Singapore has the lowest minimum wage among advanced nations.
The graph seems convincing, but the key is what new information is able to change the context of the argument? One statistic I can think of would be the proportion of workers earning minimum wage. If the entire country has only 1 person earning minimum wage, then the minimum wage level simply becomes irrelevant. I do not have the relevant statistic, because my only purpose is to provide a conceptual argument (and not to be dragged into some political sling fest).
RHB Securities REITs report
Unfortunately, I can no longer find a copy of the report. However, I recall that the analyst was saying S-REITs should do fine even in the event of a U.S. interest rate hike. His argument was that S-REITs did well the last time when U.S interest rates rose during the 2004-2007 period. This is peculiar as most people would know that REITs, as leveraged financial instruments, are highly susceptible to interest rate changes. Well, the third factor here would be that 2004-2007 was also the period when Singapore property prices through the roof.
The bottom line – knowing what to conceal will make you a very good sell-side analyst. I kid.