The chart below shows the holding period and return on capital (inclusive of dividends) of the 17 stocks that I have bought and sold, excluding short term trades. The actual returns can be found here. Using Excel, I ran a regression of the two variables and found that there is almost no correlation between the two of them (low R-square of 0.001).
From an academic standpoint, my results do not contradict the conventional wisdom that the you can get good returns when you hold stocks over a long period. This is because the defining study, Stocks for the Long Run, uses a 30 year period and a very big basket of shares. This is in contrast to the fact that share that I have held longest is at most 6 years and that most of them have small market capitalisation.