How to Get Out of Sync
Correlation is the degree to which two investments
move together. When you build your portfolio, you
want to invest in asset classes with low or "negative"
correlation, because these move independently and
allow you to hedge (protect) against fluctuations in
return. Though it may sound odd, in this case
"negative" is good!
If two investments have perfect positive correlation, their performance will be identical. You could
track their ups and downs point-for-point, and they would always match. Financial economists call
this a correlation of +1. On the other hand, if they always move in opposite directions -- if they have
perfect negative correlation, their correlation is -1. If two investments move independently, their
correlation is 0.
Any variation between +1 and -1 is possible, and indeed, this is where all correlations occur. No
two assets have a perfect negative or positive correlation. Finding even a partially reliable
relationship between two securities is difficult.
Economists use some rules of thumb to judge how different industries and different types of
securities move, but determining economic cycles is itself an inexact science. So when you think
about correlation, remember that there really are no perfect examples of it in the real world.
Correlation can be measured at the level of asset class. The oldest and most often repeated
example of negative correlation is the relationship between stocks and bonds. For years, it has
been believed that stocks generally do well when bonds do poorly, and vice versa. This belief took a
real pounding in the mid-1990s, however. Just to prove that we are living in the real world, stocks
and bonds both fell dramatically in 1994 and shot up in 1995.
It is also important to look at the level of correlation among stocks in different companies. The
performance of stocks in commodities (such as chemicals or metals) often lags behind the rest of
the economy by a few months, so they may continue to do well when retailers have begun to slide,
and then will probably remain low for a while once the economy rebounds.
Some economists see a negative correlation between restaurants and supermarkets. When people
have a lot of money to burn, the theory goes, they can afford to eat out and restaurant stocks do
well. When money is tight, consumers economize by buying food from the store and eating in, so
supermarket stocks do well.
If these examples seem tenuous, it's because they are. When you're dealing with even one asset,
it's difficult to guess future performance with any reliability. Guessing how two assets will perform in
relation to each other is even harder.
Here are three hypothetical companies whose respective stocks might have different degrees of
- Big Bang Corp., The largest manufacturer of fireworks in North America.
- Appleland Unlimited, a startup company gambling that consumers will flip for the great
- taste and safety of shrink-wrapped, factory inspected candy apples.
- Great Pumpkin Ventures, a nationwide distributor of farm-fresh pumpkins.
All three companies' products can be said to have seasonal demand. For Appleland and Great
Pumpkin, demand picks up through the fall and peaks around October 31. Big Bang has its best
sales in the summer, with peak business coming on or about July 4.
Over the course of the year, stocks of the three companies might look like this:
Appleland and Great Pumpkin have a very high positive correlation. Big Bang has a very strong
negative correlation to the other two. Its stock does well when the others do poorly, and vice
Even in this simple example, there is no perfectly reliable correlation. Big Bang does most of its
selling in July, but sees upticks in demand at New Year's Day, Labor Day and other national
holidays. Appleland (which probably has the shakiest business plan of all three companies), finds
that after Halloween consumers lose all interest in candy apples. Great Pumpkin's business also
peaks around Halloween, but its sales remain brisk through Thanksgiving.
In real life, calculating a pattern of correlation requires sophisticated math techniques and is never
simple. But knowing that you should try to invest in assets with low correlation already puts you
ahead in the game. With a bit of research and thought, you should be able to find some.