More than half of equity mutual funds are misclassified by investment style, much of it done intentionally to gain a marketing edge, a new study suggests.
Mutual fund managers "have an incentive to misclassify their funds on purpose to obtain higher rankings," concluded Erik Witkowski, who is a research assistant at Northfield Information Services Inc., Boston.
"There's a lot of incentive to be misclassified, to rank well within a group to gain a marketing advantage," said Mr. Witkowski in an interview. If the misclassified funds had been placed in their proper groups, their performance would not have stood out among their peers.
Dan DiBartolomeo, Northfield president, added, "While there is tremendous anecdotal evidence about funds not being clearly delineated in their true style, this is the first robust study showing misclassification not only occurs, but that it is purposeful.
"It is not random, and the study shows one can identify where it is likely to occur."
In Mr. Witkowski's study of 709 equity mutual funds, he found 394, or 56%, were misclassified, leading him to reclassify them into categories other than their own.
Of the 394 misclassified funds, the study found 75 were what he calls "seriously" misclassified.
The largest in terms of assets of the seriously misclassified funds are:
Vanguard Windsor Fund, which has $19.9 billion in assets. It originally was classified as a growth and income fund, but reclassified by Mr. Witkowski as an aggressive growth fund;
Fidelity Growth & Income Portfolio, $8.4 billion; originally growth and income, reclassified as aggressive growth;
INVESCO Industrial Income Fund, $3.9 billion; originally income, reclassified as growth;
Mutual Series Shares Fund, $3.5 billion; originally growth, reclassified as income;
Phoenix Series Growth Fund, $2.1 billion; originally growth, reclassified as income;
Fidelity Capital Appreciation Fund, $1.7 billion; originally aggressive growth, reclassified as income;
Fidelity Fund, $1.6 billion; originally growth and income, reclassified as aggressive growth;
Sequoia Fund, $1.5 billion; originally growth, reclassified as income;
Lindner Fund, $1.5 billion; originally growth, reclassified as income; and
Mutual Series Beacon Fund, $1.5 billion; originally growth, reclassified as income.
Under his definition, he considers a fund misclassified if it is one category or more away from its proper peer group. But he considers a fund seriously misclassified if it is two categories or more away from its proper peer group.
Mr. Witkowski completed the study for a thesis in the department of economics at Harvard University, Boston, with assistance from Northfield Information Services. He used returns-based analysis - a methodology developed by William F. Sharpe, a professor in finance at Stanford University and a Nobel laureate in economics - to classify the mutual funds.
Based on this method, he used only the returns of the mutual funds (rather than a fundamental analysis of the securities in a mutual fund portfolio) to determine their investment style and classification of the funds, each with performance histories of at least five years.
Northfield markets to pension funds and others in portfolio management a software program it developed using the Sharpe returns-based style analysis methodology.
A separate study by Jon A. Christopherson, analyst, Frank Russell Co., Tacoma, Wash., challenges the validity of using such returns-based analysis developed by Mr. Sharpe and others to determine investment style.
Without question, Mr. Christopherson acknowledges the importance of studying investment style. "(I)nvestment style involves implicit links between equity characteristics and subsequent returns," Mr. Christopherson notes in his study. Style is important, he added, because it helps investors anticipate the pattern of future returns. Thus, discerning changes in style are important for investors to manage risk.
But Mr. Christopherson in his study questions relying on the relatively new returns-based methodology.
It "fails to fully appreciate the role of noise in the data," he says in his study. "By underestimating the effects of noise, the method can lead an analyst to conclude that a portfolio is exposed to factors of return when, in fact, no exposure exists."
In short, Mr. Christopherson warns such a method "is no better off at forecasting manager returns than simple style indexes assigned on the basis of universe membership."
To determine a portfolio's style characteristics, one has to look at the specific securities in the portfolio, he said.
In addition, he noted the returns-based analysis is highly sensitive to different assumptions, especially the timeframe used. Often, analysts using returns-based methodology use five-year periods. Mr. Christopherson suggested results would be different if different time periods were used.
In an interview, Professor Sharpe disagreed with Mr. Christopher's criticism.
On one of Mr. Christopherson's points, Mr. Sharpe said returns-based style analysis shows what he called the "effective asset mix," of a portfolio, which he said is more important to know in determining investment style than the actual securities in a portfolio.
A returns-based analysis may, indeed, show a portfolio exhibits characteristics of having, say, some international securities, when in fact the portfolio has only domestic securities, Mr. Sharpe said. But he said some domestic securities can be sensitive to international effects, such as international interest rate movements or currencies.
"Just because it's a domestic security doesn't mean it's not sensitive to interest rates abroad," or other international effects, Mr. Sharpe said.
"The (returns-based) analysis is telling you something that is not self-evident if you look only at the securities," Mr. Sharpe said.
In his study, Mr. Witkowski used the returns from Des Moines, Iowa-based Investment Company Data Inc.'s database and compared his resulting categories with its mutual fund classifications, which he said generally were selected by the mutual fund company.
The study didn't specifically compare data from such sources as Morningstar Inc., Chicago, or Lipper Analytical Services Inc., Summit, N.J., so he couldn't draw conclusions about the accuracy of their classifications. But he said all of the mutual fund data sources use classifications that rely to a large part on the manager's own category and are similar to those classifications ICDI uses.
Mr. Witkowski grouped the mutual funds into six conventional categories - aggressive growth, growth, growth and income, income, international, and small capitalization.
He concedes a reason for the misclassification for some funds lies in the ambiguity of the current groups. Funds typically are grouped by the shared fundamental characteristics of their investment style, especially their level of risk.
In his study, he found in general the worst misclassification occurred for those mutual funds that are part of small mutual fund complexes.
Mutual funds of large mutual fund complexes tend more to be classified in their proper investment style group. But such tendencies weren't always consistent, he conceded.
"When you have a big mutual fund complex group, the chances are some of their funds might show up on the list" of being misclassified, he said.
In general, Mr. DiBartolomeo said, "The big guys seem to be more honest. It is the small guys that seem to be mucking around."
The reason, he added, is "the big guys tend to have a fund for every category, so they don't need to fill marketing niches. Little companies don't have the advertising budgets big companies do. So the only way they can get the attention of the press is to get high rankings. The easiest way to stand out in a category is to be in the wrong category."
In pointing out the ramifications of the findings, Mr. Witkowski said misclassification misleads investors to allocate their money to mutual funds whose investment characteristics and risks are different from what they expected.
To put it in academic terms, he said the misclassification prevents investors from maximizing their utility, or in other words, their investment objective.