Friday, July 14, 2017

Current Thought on Momentum

Momentum and MFI

Sure, I studied physics in college.  Struggled at first, but as my math improved, so did my Physics.  Plus I got a tutor.  Any way, here is a definition of momentum:

momentum definition. In physics, the property or tendency of a moving object to continue moving. For an object moving in a line, the momentum is the mass of the object multiplied by its velocity (linear momentum); thus, a slowly moving, very massive body and a rapidly moving, light body can have the same momentum.

So in physics, momentum is a function of velocity and mass.  In investing, we are thinking more about the first part, the property or tendency of a moving object to continue moving.

I have been studying whether magic formula stocks can have momentum.  The question I am trying to answer is whether the movement of the stock price in the past 6 months have any predictability on where it will go over the next year?

I built a database going back to 2011. Being just a guy with a spreadsheet - so using some tools available on the web, I did the best I could. But always possible on individual stocks I made errors as sometimes hard to adjust for splits, special dividends, name changes, and stocks that have disappeared or spun off other stocks.

Here are my findings by purchase year:

Year       Overall   Momentum  Dividend
2011 -0.8% -2.1% 9.0%
2012 27.2% 25.5% 41.0%
2013 24.5% 24.9% 26.9%
2014 3.8% 11.2% 8.9%
2015 -3.8% -1.5% 0.9%
2016 16.1% 21.2% 17.9%
2017 1.0% 1.0% 3.4%
Grand Total 10.2% 12.1% 17.2%
Cash  $183,720   $205,765   $261,315 

So momentum does seem to have some merit, over 6 years you'd have over 10% more.  But it does not seem as good as dividends.  You will see that relative to overall, the worst momentum year was 2011.  That was the year the US Govt was being downgraded in August and a lot of stocks dropped substantially.  But then in ensuing year, some bounced back a lot.  It seemed to me, that in this type of falling market that the momentum approach did not work particularly well.  A lot of the stocks had dropped as Mr Market went nuts and they were the best values.

All being said, I think there is enough here to be interesting. So I think I will split my "Formula" approach into Dividend and Momentum (and increase my money in them overall).  In September/October, I will randomly pick 5 momentum stocks.  

My filter for momentum took out both tails.  So I screen for 50 stocks over $100m.  Then I throw out the 25 with the worst/lowest momentum. But I also throw out the top 5 (I found in backtesting that was an improvement).  So I will have 20 to choose from.

Happy investing!

No comments: