22.214.171.124 The "New Common Sense"
Having built the case for regarding manager structure and selection as an optimization problem, reflecting the reality of pure active return and risk as the real dimensions of active management, we can state what a portfolio reflecting these principles might look like. Moreover, a portfolio reflecting the principles we've advanced should have pretty much the same characteristics whether one formally optimizes or not. Not everyone has the time, focus, or patience to run an optimizer,
26 As stated above, for a given level of skill, the information ratio must be lower for a concentrated high-risk active manager than for a lower-risk manager. And, of course, if the concentrated manager really is unskillful, its high risk level will mean that downside realization will be especially painful. On average, the highly concentrated manager performs at the average (before fees and costs - just like any other active manager).
and we're all equipped with pretty good "fuzzy optimizers" above our shoulders. For investors who use an actual optimizer and for those who don't, the following can inform one's intuition, providing common sense that is directed by the nature of the underlying optimization problem:
1. Be disciplined in forming expected alphas, and in giving the greater weights to managers with higher ratios of expected alpha over active risk squared. Be rigorous when examining historic alphas, looking for clues to the future. If a manager's alpha is not why are you looking at it, your final estimate of expected alpha should be strongly supported by fundamental analysis of the manager and its process.
2. The best portfolio is one that balances the two key dimensions of active management - maximizing pure active return and controlling the total active risk, summed across managers.
(a) The portion of the portfolio that moderates its overall risk will consist of some combination of "good" risk-controlled active funds (sometimes called enhanced index funds) and traditional index funds, which together will likely comprise roughly one-third to two-thirds of the total fund. If an investor has little tolerance for active risk, this portion will lean more toward index funds; if there is more tolerance for active risk, it will lean toward risk-controlled active funds.
(b) Risk-controlled active funds will be more heavily weighted than equally skillful traditional active funds at most risk levels chosen by institutional investors.
(c) Among traditional active managers, prefer skillful lower-active-risk managers over higher-risk, concentrated managers. Bias toward diversified portfolios, away from concentration (unless completely carried away by the concentrated manager's extraordinary forecasting skill).
(d) "Good" market-neutral long-short funds will receive a substantial weight for investors not limited to long-only managers. 27
3. Set the overall active risk at a comfortable level, your "risk budget." For US equities, a typical investor seems to be most comfortable (in the experience of the authors) at an overall active risk level of 1.5-2%, with the very largest investors preferring even less active risk (between 0.75% and 1.25%).
4. Keep a careful eye on misfit risk, trying to minimize it while still maximizing expected alpha. If you don't use an optimizer designed especially for this purpose, you'll have to use a style map or "effective asset mix" table as a supporting tool.
27 Alphas delivered by high information-ratio managers sourced in one asset class can, at least theoretically, be ported to another asset class (where, perhaps, high information-ratio managers are scarce) by the use of futures or other derivatives. Such a "portable alpha" strategy is most frequently used to add alpha, generated in hedge-fund programs, to an equity or fixed-income account.
24 Understanding Active Portfolio Management 561
Was this article helpful?