Broker Check

Can Returns Help Predict Future Results? (Some rough math)

February 21, 2020
Share |

If you have ever tried to select your investments in a retirement plan, you probably have seen a list of available funds. It's a universal list, showing each fund, its category, and then performance results for the past 1 year, 3 years, 5 years, and 10 years. Then near the bottom of the page is a disclaimer saying something to the effect of "The performance data featured represents past performance, which is no guarantee of future results." 

Although there is no guarantee of future results, many investors still make their investment decisions based on the return information in front of them. They may think to themselves something like, "Here are the returns. They provided them to me. They must be important and relevant. Right? Hmmm, that fund seems to be "doing" pretty well. Let's use that one." As it turns out those returns are pretty useless if you are trying to forecast the future results. In fact, there is so little relationship that you may as well be rolling dice if you choose funds based on returns. 

I had always heard this over the past 25 years (Past performance is no indication of future results), but I wanted to see for my self. So, yesterday, I looked at 120 years (1900-2019) of the Dow Jones Industrial Average (DJIA for short). I wanted to do the math. What is the relationship between past changes and future changes in something like the DJIA? (FYI, I ignored dividends to keep the math simpler, and can only assume that dividends wouldn't change the outcome enough to change the conclusions.)

What I found follows. See if you can find any relationships. For orientation, you can see a statistic called R-squared on each chart. This measures the exact relationship between the two sets of numbers. A perfect correlation between the two numbers would have an R-squared of 1.00 (or 100%). 

Does last year's price change have anything to do with next year's price change of the DJIA? Answer: NO. Only a 0.67% correlation!

Does the last 3-year price change have anything to do with next 3-year price change of the DJIA? Answer: NO. Only a 3.2% correlation!

Does the last 5-year price change have anything to do with next 5-year price change of the DJIA? Answer: NO. Only a 6.55% correlation!

10 years is a pretty long time. Surely that one will work.
Answer: NO. Only a 0.3% correlation!

What about really long periods, like 30 years?
Answer: No again. Only a 1.7% correlation!

Past performance is useless as a tool for predicting future results. Great. So now what do you do?

Funny you should ask! There is still hope for us all. The math just isn't as simple as we'd all like. The global economy has a lot of complex variables.
But stay tuned. We'll be following up with more on that soon. The coming topics include:

  1. Understanding Valuation for Normal People 
  2. Risk, Emotions, and your Portfolio Design
  3. Building Investing Strategies Around Goals

As always, if you any specific questions about your situation, just contact us. We're happy to help.