S&P 500 index, outperformed by 121 basis points. Given our definition of active return, its exceptional active return was 111 basis points. Of that, 174 basis points came from stock selection and 54 basis points from factors, while market timing actually detracted 117 basis points from the account's performance. If you look at factor contributions, both the industry and style exposures added value, 30 basis points for industries and 24 basis points for styles. Currency and country contributions were nil since this is a single country portfolio. The report also provides a more detailed breakdown of attribution at the stock (specific), sector, style, and industry levels. Contributors to specific return are computed by taking each stock's active weight and multiplying it by the difference between the stock's total return and the return attributed to factors (excluding market timing). This difference is what forms specific return. Taking a look at the "Contributors to Specific Return" section of the table, we find that the majority of the top and bottom 10 contributors over this period are made up of positive active weights (i.e., higher weight in the portfolio than in the benchmark). If we consider positive active weights as representing stocks that the portfolio manager prefers, then we can see that many of his or her preferred stocks are some of the biggest contributors and detractors of specific return over this period. Next, we explain an alternative return attribution methodology-asset grouping-that forms the basis of variance analysis. Asset Grouping Methodology Portfolio managers want to view their portfolios' sources of return in a simple and relatively straightforward manner. Some prefer not to use a factor model at all, as they do not view their portfolio construction process as being driven by some predefined, quantifiable set of factors. These managers usually rely on commercially available systems that employ an asset grouping methodology to generate so-called variance analysis reports. This methodology consists of three steps: 1.    Group assets. For each time period (e.g., a day) we group assets according to the value of some factor. For example, we may group stocks by their industry classification or by their exposure to a particular investment style. In the case where we group assets by their style exposure, we may first generate deciles of the distribution of all exposures4 to a particular style and then group assets into deciles based on their particular exposures. 2.    Compute the return of each group. Once assets have been grouped, we compute their one-period returns. The return for the group is computed by taking a weighted average of all returns in the group where the weights are based on the group's total market value. 3.    Compute the contribution of each group to the total return. The contribution of each group is computed by taking a weighted average of all returns in the 4A popular way to define all exposures is to use the exposures corresponding to the assets in the benchmark portfolio.