Issue link: https://resources.envestnet.com/i/1528379
3 Active vs. Passive Asset Management: An Update First, we measure the statistical significance (positive and negative) for each percentile of the cross- sectional alpha distribution for all mutual funds (both dead and alive) in the Morningstar database (see Table 1) from January 1980 to April 2022. We use regression analysis to calculate manager alphas, employing all the available data for that manager and using the benchmark for the manager's Morningstar category as the independent variable, to determine the "manager success rate" and the "manager failure rate." We then classify a category as "active" if the lowest statistically significant positive percentile was at 67 or below. In other words, an active manager's success rate is at least one-third, meaning at least one-third of a distribution of alphas in a given category are statistically positive. We classify a category as "passive" if the manager failure rate is at least two-thirds, with the highest statistically negative percentile for a passive category at 67 or above, meaning that at least two-thirds of a distribution of alphas in a given category are statistically negative. We classify a category as "neutral" if it is neither active nor passive. Further, managers with an alpha above or equal to a statistically significant positive percentile are denoted as "skilled," whereas managers are considered to be "unskilled" if they have an alpha below or equal to a statistically significant negative percentile. Thus, each Morningstar category is divided into a group of skilled, unskilled, and indeterminate managers. Note that we grouped all the managers from the same category in the same peer group, regardless of the time that they were active, and the results show the average proportion of active managers for a particular category across time. 2.2 | Time Dimension Of Active or Passive Classification During our more than four-decade long study period, managers had vastly different market environments, asset management approaches, and technologies available to them. Changes in these factors can produce markedly different proportions of managers generating positive alpha over time. In fact, as we will see later when discussing the time trends of active management, these results can, and most likely should, be used in making the active or passive investment decision. Our trend analysis measures performance over rolling periods at monthly intervals, where at any given month we analyze only those managers who are alive at that particular time period. Also, we use 36 months of data for the alpha regression analysis. To be consistent with whole-sample analysis, we measure all managers against their Morningstar category benchmarks. This gives us a time series at monthly frequency of the proportions of positive alpha managers in a particular category, and we repeat this analysis across all of the categories. We apply the same estimation methodology as we used in the whole- sample analysis to determine whether a particular positive or negative percentile is statistically significant. Regression analysis: A statistical method to obtain a linear explanatory relationship be- tween a particular variable ("depen- dent variable") and a set of potential explanatory variables ("independent variables"). Whole Sample Analysis: Refers to using all the available historical data, when running a regression analysis for a particular manager. Since this approach uses the maximum available data, it may provide analysis that is less relevant for more recent periods, since the results are influenced by the whole available, and potentially distant, history. Manager Failure Rate: The proportion of managers whose alpha is below or equal to a statisti- cally significant negative percentile, calculated from ranking all the managers in the peergroup. Manager Success Rate: The proportion of managers whose alpha is above or equal to a statisti- cally significant positive percentile, calculated from ranking all the managers in the peergroup.