I decided to revisit momentum investing using data from Sharelock Holmes on companies over £20m market cap. Using 6 month and 1 year relative strength, I was able to define 2 statistics for a company: rs6a (the relative strength for the first 6 months) and rs6b (the relative strength for the last 6 months. The data was split into quintiles for rs6a, and the medium of rs6b was calculated for each quintile. Here are the results:
Q1: -12.11
Q2: -4.13
Q3: 1.51
Q4: 3.61
Q5: 6.89
As my previous study had shown, there is a very nice relationship here, which basically confirms the momentum effect, at least over the time period studied. The conclusion here is pretty straightforward: buy the top quintile of strongest-performing stocks over 6 months. Buying the bottom quintile would have been, in aggregate, disastrous. Note that the outperformance of 6.89% would be for a period of 6 months, not 1 year.
Re-running the stats for companies over £50m market cap, I obtained the following results:
Q1: -8.26
Q2: -3.52
Q3: 3.03
Q4: 5.07
Q5: 10.79
So the effect had proved better for larger-cap companies in the top quintile. They would have also been easier to invest in.
Mark Carter's blog 