Aurora Borealis
1 April 2016 CFA Institute Journal Review

Measuring a Manager’s Trajectory—A (Very) Simple Approach (Digest Summary)

  1. Nicholas Tan, CFA

Traditional performance analysis of portfolios is typically done by examining static portfolios at particular points in time, which may overlook the role that the fund manager’s investment decisions played in a portfolio’s performance. The impact of a manager’s investment decisions can be studied in more detail by comparing the returns of a dynamically changing portfolio with theoretical static returns.

What’s Inside?

The author explores the impact of a fund manager’s actual investment decisions over a given period by proposing a metric called “trajectory,” which he defines as the portfolio’s calculated return minus the portfolio’s buy-and-hold return. This metric could better illustrate the effect that a manager’s trades have on a portfolio for a given period compared with traditional attribution models, which analyze only the composition of portfolios.

How Is This Research Useful to Practitioners?

Traditional performance attribution provides useful information to the portfolio manager but seldom offers actionable information to the asset owner. Rather, the traditional approach analyzes the composition of portfolios but does not isolate the impact of managers’ actions on that composition.

Trajectory, the author’s proposed metric, could be an additional metric for asset owners to use to assess their managers in addition to such metrics as excess returns and risk measures. The trajectory metric allows an asset owner to look in aggregate at the outcomes of the manager’s decisions and to see the difference in returns between the real and theoretical portfolio that can be attributed to the trading activity of the manager.

Together with the measure of a portfolio’s excess return over the benchmark, managers can be divided into quadrants by using the calculated trajectory. The four quadrants are underperforming benchmark but positive trajectory, overperforming benchmark and positive trajectory, underperforming benchmark and negative trajectory, and overperforming benchmark but negative trajectory. Trajectory versus excess return can be plotted on a coordinate plane, and asset owners can make retention and firing decisions on a pool of managers accordingly. The approach does not consider cash movements (inflows or outflows) outside the control of the portfolio manager.

How Did the Author Conduct This Research?

In the trajectory metric, portfolio calculated return is the portfolio return for the period that is computed by using either the internal rate of return or the time-weighted rate of return methodology. The portfolio buy-and-hold return is the theoretical internal rate of return based on the portfolio’s composition at the beginning of the period that is held constant through the period. The difference between the two returns is the impact of the portfolio manager’s trades for a given time period.

The author offers some caveats that may invalidate the use of the trajectory metric. For example, corporate actions or the transfer of securities between portfolios could be counted as active decisions by the manager. Another caveat is that the trajectory measure may not work in some asset classes, such as fixed income and private equity. Finally, making hedging/risk adjustments to the portfolio, having an extended time horizon, and linking returns to create extended period trajectories could all interfere with getting correct information from the trajectory metric.

Abstractor’s Viewpoint

The proposed trajectory metric, which uses available information and is simple to implement, could be a useful tool to help break down the attributes of alpha generated across funds. The metric could aid investors in quickly distinguishing bad managers from good managers and thus help less sophisticated investors who may lack the resources for in-depth performance attribution analyses.

We’re using cookies, but you can turn them off in Privacy Settings.  Otherwise, you are agreeing to our use of cookies.  Accepting cookies does not mean that we are collecting personal data. Learn more in our Privacy Policy.