I’m having trouble understanding the logic behind the within sector allocation return. I know it calculates the impact of performance attributed only to individual security selection decisions (basically did you pick better stocks than the benchmark.). And the formula is: WS return= Sector benchmark weight*(sector portfolio return-sector benchmark return) I don’t get how this formula calculates the return. If you over-weighted an outperforming sector, wouldn’t this also result in a positive number for the within sector allocation even if you picked underperforming stocks using this formula? I don’t get how this formula accounts for how the portfolio sector weighting affects the portfolio return. It seems like you should account for how much you over or underweighted the sector and subtract that effect out first, and then you could find out how well you picked stocks in the sector. Can anyone explain what I’m missing? Thanks.
Read my comments (second from bottom) from this thread: http://www.analystforum.com/phorums/read.php?13,684898 Essentially, the within sector return is just one of three components of a managers return differential attribution (microattribution). You’re right, the WS affect alone is not sufficient. Hopefully reviewing the other factors will help fill in the gaps.
What about my effect? 
Thanks Patacon. I’ve reviewed the material and everything and I have the formulas memorized, its just the formula for this doesn’t make intuitive sense the way the others do. It seems that to find out the attribution of stock picking within a sector it should be a formula like this: [(Portfolio sector weight/benchmark sector weight)*benchmark sector performance] - portfolio sector performance For example, say that you weighted the Tech sector 10% and the benchmark had a 5% weight, and the benchmark had a 2% return and the portfolio had a 4% return. Then basically you within sector return would be zero because you had double the return because you doubled the benchmark allocation. However, after typing this I just realized that the returns aren’t weighted as a percent of the whole, so my above paragraph doesn’t really work. I’ll just memorize the formula…