|
Submitted by: Howard B. Aschwald, CFA
Howard B. Aschwald, CFA is Executive Vice President and Chief Investment Officer of Quantum Capital Management. Howard is a current member of the Security Analysts Society of San Francisco and the Marin County Estate Planning Council. He has over 25 years in the financial services industry and holds MBAs in Finance and Financial Planning. Quantum Capital Management is a full service, SEC-registered investment counselor and advisor.
Investment advisors and their clients looking for index-beating returns have not had much support from academic research. Most of the published articles conclude that the average mutual fund provides less risk adjusted return than a corresponding passive index. Fortunately, some very good research published by Yale University professors Martijn Cremers and An Petajisto in early 2007 demonstrated a convincing case for active mutual funds under the right circumstances. On average, the most active funds by portfolio composition outperformed their respective index benchmarks by 1.5% a year after fees and expenses.
What worked
The right circumstances included smaller relative total asset size, a lower number of total stocks in the portfolio and an emphasis on stock selection as opposed to market timing or sector rotation. The largest 40% of funds in their categories added negative value after fees and transaction costs. Furthermore, funds with a large number of holdings essentially became (closet) indexers and after expenses, significantly under performed their indexes. Fortunately, many of the largest funds by total assets were also the same funds with hundreds of holdings. (This happens because many funds are constrained to hold less than 5% of the total shares of a company. Above that threshold more onerous reporting requirements are triggered by the SEC.) The funds that were smaller in total asset size and that were the least index like in their composition outperformed their behemoth closet indexing category funds by an average of 5 to 6% a year.
What doesn’t work
So, what are some good ways of identifying funds that will provide superior performance (and at least avoid the ones that will be performance laggards)? Let’s look first at what doesn’t work. While lower fees are generally better, many closet indexing funds charge the same as active funds. Selecting for low versus high turnover also doesn’t work as a large percentage of closet indexers exhibit high turnover. Finally, while the beta of a closet indexing fund will be at or very close to 1, funds that excel in stock picking also tend to have betas that are similar to their benchmark, even though the portfolio’s composition is substantially different.
What will work
A set of reasonable screening parameters will incorporate the following: No-load and no transaction fees (for RIA’s and their clients); an expense ratio less than 1.5% for small mid cap and emerging markets and less than 1.25% for large cap and international; less than 100 holdings for US stock funds, less than 200 holdings for international; total assets less than $1Billion for small mid cap and less than $2 Billion for large cap US and International; the top 10 holdings should be 18% or greater of the total holdings.
In terms of quantitative filtering, we have found that funds with an R-squared that is at least .85 or greater relative to their index are more likely to adhere to that index in the future. This also demonstrates more stock selection skill (or lack thereof) relative to a validated peer group. When R-squared falls below .85, we have found it necessary to look at three year standard deviation or the Sharpe ratio to determine if there has been any real alpha generated. Finally, we have found success choosing among funds with a positive 3 year alpha relative to the best fit index as our own research has shown more performance persistence over the subsequent three year period. Three year records seem to work better for identifying future success than five year records, although both periods can be used for a more comprehensive evaluation.
Qualitative screening
While using quantitative screening to create a filtered universe of mutual fund candidates is a great time saver, it is still necessary to do qualitative research. The overall objective of this research would be to ascertain how the past three year performance history came about and how likely past processes will be sustained going forward. As an example, for a number of recent years, value strategies have performed better than growth strategies. A fund’s out performing a blended index benchmark might be more attributable to a value tilt rather than stock picking skill. In this situation, it would be better to look at how the fund did relative to other value funds in the same category.
More qualitative criteria to look at would include a review of the fund’s written description of their investment process and comparing what the fund currently owns and what they have owned (via past 13F filings) to what they say they do. If there is a reasonably close match, then it would be fair to assume that they will continue in the same process going forward. As a shortcut to reconstructing portfolio changes, funds that have had key manager turnover in the past three years should be avoided.
While reviewing a fund’s written material (or, better yet, interviewing a member of the fund’s decision team), it is also a good final step to determine a fund’s unique research process. What do they do that adds value in the stock selection and allocation process? How do they use publicly available information differently from the way other managers use the same information and why does that add value? Hopefully, the answers should have enough compelling logic to inspire confidence that the fund’s superior performance will persist long into the future.
Separately Managed Accounts
The same methodology for fund selection can be used to select superior separate account managers. In addition to better potential tax management, many separate account managers have a smaller number of holdings relative to their mutual fund counterparts. This generally requires more conviction in their stock selections as opposed to the comfort of numbers that is the norm in most large mutual funds. The only additional screening criteria to incorporate into the search would be to select for managers where performance track records were verified as GIPS compliant and audited by an independent third party. There are more than enough audited separate account managers to choose from without the risk of relying upon numbers of doubtful quality. In addition, as all mutual funds are independently audited, this puts separate account managers under the same microscope for comparison purposes.
Not to index
According to research from Vanguard Mutual Funds, the average index investor stays with an index fund for three years. While Vanguard did not release any conclusions, it is fairly safe to project from other studies that the undisciplined switching of funds, even with index funds, will lead to relatively poor performance over time. (The Dalbar study showed average long term underperformance of 6.5% a year.) Apparently, most investors would rather try for a superior outcome rather than passively accept whatever an index will give them. Even in down markets, if a superior active fund can beat its index, the relative out performance may stop an investor from making a poor decision.
Based on recent academic and our own internal research, mutual funds (and separate account managers) that are truly actively managed have a much better chance of out performing their respective indexes on a risk adjusted basis. By incorporating a disciplined quantitative screening and filtering process, past performing superior mutual funds can be identified. With the further application of good qualitative judgment, it is highly probable that an investor will achieve better long term performance than an indexing strategy.
> Return to Financial Investments |
Authored by the
founder of the
PaladinRegistry