So, we can conclude that a manager’s value-added return is statistically different from zeo if value-added return falls outside the confidence intervall ?
Is the “value-added return” the cumulative value-added return or the daily value-added return? Same thing goes with the 2sd range for the value-added return; do we use the deviation of the cumulative alpha or the daily alpha. When I look at the chart in the Schweser notes, it appears to be using a cumulative return as I would expect the line to dart around zero quite a bit if daily returns were used – but I can’t get my head around how the standard deviation of the cumulative returns would work – wouldn’t it be kind of skewed?
发表于 2013-4-17 18:37
| 