asset class as exhibited by the higher intra-asset class correlation. It should be clear that the ability of investment managers to understand and manage the risk in their portfolios is of direct benefit to the client. Arguably managers' ability to quantify portfolio risks is a strong indication of skill and should positively correlate with their ability to consistently outperform the market. The foundation of successful portfolio construction is predicated on a manager's ability to understand and quantify sources of risk in a portfolio, to size intended exposures appropriately, and to avoid unintended exposures. Risk managers can implement a simple approach to measuring the success or failure of their investment manager's ability to size their risk appropriately. We call this approach the "green zone." The idea is to define three levels of outcomes for tracking error. The first range of outcomes represents those that are close enough to a manager's targeted realized tracking error to be considered a successful event. This is the green zone. The second range of outcomes, the yellow zone, represents outcomes that are not successful, but that are close enough to target to be expected to happen on occasion. While the yellow zone is deemed to be unsuccessful, we should nonetheless expect even the most skilled investment managers to operate occasionally in the yellow zone simply because realized tracking error isn't fully controllable. Yellow zone outcomes should be viewed as warning signals to the risk manager. However, there may be a reasonable explanation for the event. Finally, we will define bad tracking error outcomes as the red zone. Events in this zone should occur rarely, if at all, for an investment manager who understands the sources of risk in their portfolio. Red zone events should be thought of not only as warnings, but as likely indications of a lack of control in the portfolio construction process. This green zone discussion brings us back to our earlier example of the domestic equity manager who was able to beat the Russell 3000 index. Clearly, we are delighted that one of our managers is able to generate performance in excess of the benchmark. However, the manager was using higher levels of risk than we expected in order to generate positive performance. These unsuccessful tracking error outcomes not only are warnings for the risk manager but are likely indications of a lack of control in the manager's portfolio construction process. A thorough review of this manager should be conducted to ensure that inclusion in the total plan is wise. The manager analysis should also take into account not only the amount of risk being taken by the manager, but the manager's impact at the asset class and plan level as well. We say this because if our manager is taking on larger unintended exposures in the portfolio, it will typically mean that our domestic equity asset class will have a higher tracking error and contribution to total plan risk than budgeted. It's worthwhile to spend time describing the process of setting manager level tracking error targets. The process entails the use of a manager's performance history or track record and the benchmark to which our manager's portfolio is compared. It is not uncommon for managers to have daily track records in the case of mutual funds; however, some institutional managers produce composite performance only on a monthly basis. In either case, data frequency should not present a major hurdle as long as the managers with monthly performance data have long enough track records. The objective is to compute rolling tracking error over various periods of time (i.e., rolling 20- and 60-day with daily performance data and