Before (and after) hiring an investment manager whose approach is to actively manage a stock (or bond) portfolio, a thorough examination should be performed to determine if the manager has added value compared to a market-based (passively managed) benchmark. The reason these tests need to be performed is that overwhelming evidence shows the vast majority (80% - 99% depending on method of measurement) of active money managers have not added value. Since there is such an enormous risk to investors of underperforming a passive approach, active money managers should prove how they have added value through skillful management.
The manager or mutual fund track record should be analyzed from it's inception. While a minimum of five years of data should be reviewed (any less is probably meaningless) ten years or more is preferable. A major problem with most track records is that statisticians suggest between 30 and 140 years of data may be required to sufficiently document manager 'skill.? This is obviously a huge challenge. Few managers have the minimum statistically useful 30 years. Just 20 of 4,874 (or .41%) of equity funds have a manager with greater than 30 years of documented track record and those with 30 years of experience are probably nearing retirement.
The following is a series of analytic tests that an active money manager or mutual fund should be able to pass before engaging their services. These tests focus on the equity part of a portfolio. However, similar assessments apply to bond management as well.
The failure of any of these steps below should disqualify a manager or fund from further consideration. These constraints not withstanding, let's review the steps necessary to document manager skill.
What is the investment return compared to the most appropriate passive benchmark, after expenses? How reliable and representative is the data? Is the data A.I.M.R. compliant (an industry standard of performance measurement)? This first step will usually eliminate 75%-80% of managers or funds.
What are the returns compared to the financial factors-based (equity style) regression analysis (Fama French data series) ? Was there added value? This step adjusts for known factors of additional return associated with small capitalization and value stocks. This portion of the test can be done on international funds or managers as well. However, the complexity of the analysis increases dramatically due to the variability of associated with country allocation.
Next - evaluate risk-adjusted return. Typical statistical measurements are: Standard deviation, beta, Sharpe Ratio, Sortino Ratio, alpha, down cycle frequency, and magnitude. If a fund or manager took greater risk in achieving returns, this isn't necessarily value added.
What is the statistical significance of factor-based, risk-adjusted return (steps 1-4): Is the T statistic greater than 2.0? This step compares investment returns to a normal distribution of expected returns (bell shape curve). A T-statistic greater than 2.0 suggests that the excess positive return (if it was present) is greater than you?d expect from a normal distribution of returns. In other words, manager skill might be present.
What is the consistency of steps 1-5 over independent time periods? If there isn't persistency of risk-adjusted return from one time period to the next, one might conclude luck was the source or excess return ? not skill.
After steps 1-5, an examination of individual securities contribution to return should be conducted (assuming the data is available). If the bulk of excess returns are attributed to just a few stocks, it begs the question ? Can the selection process can be repeated?
What is the termination criterion for underperformance? What are the acceptable underperformance parameters and time horizons? There isn't a ?right? or perfect answer to this question. A general rule of thumb in the institutional world is that money managers are given at least three years to prove themselves. Would the manager's track record have triggered the termination criterion any time during the evaluation period? If so, it begs another question ? Why would one consider hiring them or buying the fund?
What has been the tax impact for taxable accounts? What is the tax loss harvesting philosophy/methodology? Is there coordination between taxable and tax deferred accounts to minimize tax? The negative effect of realized gains can dramatically reduce returns. Steps 1-7 should be evaluated on an after ? tax basis for taxable accounts.
Steps 1-5 generally eliminate about 99% of money managers or mutual funds from consideration because there wasn't added value. Steps 6, 7, 8 usually eliminate the majority of the remaining 1%. Since so few active managers or funds have added value in a statistically meaningful way, a casual observer may conclude that active managers are unintelligent, greedy, lazy, or incompetent. Occasionally, these criticisms are valid. However, the real issue is that active money managers are competing aggressively over the same sets of information. Therefore, it's extremely difficult for them to gain a consistent competitive edge over their peer group and even more difficult against a passive benchmark (or fund) where costs are lower.
Consider this. What is the additional return (over a passively-managed market portfolio) one might reasonably expect from an active manager or mutual fund if the entire evaluation criterion described above were met?
Most objective professionals might suggest .10% - .50% per year.
So here's the sixty four million dollar question (drum roll please) ? If only 1 in 500 active money managers could be expected to add .10% - .50% per year additional return, is this an endeavor that's worthwhile pursuing, considering the costs of being wrong? The cost of being wrong isn't just the poor odds of selecting a value-added active manager or fund. The magnitude of the average ?winning? manager is substantially less than the magnitude of average underperformance of the ?losing? group
The graph below shows the 20 year surviving U.S. Large Cap Funds compared to the S&P 500 Index. Due to the limitation of commercial databases as Morningstar, funds that did not survive the full twenty years are deleted from the database. This has the effect of overstating the percentage of winning funds. Adjusting for this survivor bias would reduce the winning funds to about 10% instead of 18%.
Below are a couple of examples of the difficulty in beating a passively managed benchmark or fund:
The DFA Large Value Fund (passively managed) ranked #12 of 349* US large value funds for the ten years ending March 2006, produced 11.74% annualized return, 3.40% annualized better return than the group average. The #1 ranked fund, the FAM Value Fund produced just .74% better annualized return. However, the return for the full common time period of the two funds (April 1993-March 2006) shows the FAM Value Fund underperformed the DFA Large Value fund by .53% annualized, hardly inspiring the idea of value added associated with skillful management.
The DFA Small Value Fund (passively managed) ranked #2 of 89*. US small value funds for the ten years ending March 2006, produced 17.02% annualized return, 3.50% annualized better return than the group average. The #1 ranked fund in this group, the Hotchkis and Wiley Small Value Fund, managed to beat the DFA Small Value Fund, but only by .39% annualized. However, the Hotchkis and Wiley Fund managed to underperform the DFA fund by nearly 10% annualized for a prior seven year period. Comparing the returns of these two funds for their common period (April 1993-March 2006) also shows that the Hotchkis and Wiley Fund underperformed the DFA Small Value Fund by 1.30% annualized from inception. The inconsistent performance of the Hotchkis and Wiley Small Value Fund would not be acceptable to a discerning investor.
These examples also dispel the erroneous, yet popular, notion that passively managed funds produce average or mediocre returns. The DFA US Large Value Fund produced 66% more gain than its peer group average. The DFA Small Value Fund produced 49% more gain than its peer group average, over the last ten years ending March 2006.
After all of the analysis one might attempt, it does not mean there won't be winning managers or funds relative to their appropriate index. There probably will be. The more important issue is that there is no systematic way to identify the winners in advance.
The Leap of Faith
Suppose that a fund or manager satisfied the tests above. Additional challenges still persist:
- Market conditions that enabled the manager to beat the market may no longer exist.
- The strategies that were used to extract inefficiencies from the market to earn superior returns may be duplicated by other skillful managers eliminating the competitive edge.
- The strategies employed may no longer be effective if the managers? assets get too big. Trading volume and liquidity will eventually become a constraint.
- The manager may raise their fees (as justified by the superior record) potentially eliminating the anomalous extra return. Hedge fund managers are notorious for this.
While everyone would like to know the next market-beating funds or managers, the pursuit of superior managers will more than likely produce disappointment, not superior returns.
* Source: Morningstar March 2006