A large body of literature examines whether managers of actively managed funds add value to their investors by generating abnormal returns. Unfortunately, not only do the vast majority fail to do so, but the evidence, as presented in my book, “The Incredible Shrinking Alpha,” demonstrates that the already-small percentage of managers able to beat their benchmarks has been diminishing at a rapid pace.
Indeed, 20 years ago, approximately 20 percent of active managers were adding statistically significant alpha on a pretax basis. Today that figure is down to about 2 percent.
Given that a large majority of investor dollars are still invested actively, either investors are unaware of the evidence, or choose—irrationally—to ignore it. Perhaps some investors ignore the evidence as a result of an all-too-human trait: overconfidence.
I suspect investors who suffer from overconfidence believe that (even though they acknowledge how difficult it is to outperform, and know even Warren Buffett warns that the vast majority of professionals fail at it) somehow they will succeed.
Another well-documented trait some investors possess is an irrational preference for dividends.
In their 1961 paper, “Dividend Policy, Growth, and the Valuation of Shares,” Merton Miller and Franco Modigliani famously established that dividend policy should be irrelevant to stock returns. This theorem has not been challenged since. Moreover, the evidence supports this theory, which is why to my knowledge there are no asset pricing factor models that include a dividend factor.
Unfortunately, it appears that the managers of mutual funds are well aware of the irrational preference some investors have for dividends, and they exploit that preference to the detriment of those same investors.
Lawrence Harris, Samuel Hartzmark and David Solomon—authors of the paper “Juicing the Dividend Yield: Mutual Funds and the Demand for Dividends,” which was published in the June 2015 issue of The Journal of Financial Economics—found that some mutual funds purchase stocks before dividend payments as a way to artificially increase their dividends.
In fact, more than 7 percent of the authors’ fund-year observations have dividend payments that are more than twice as large as their holdings imply. The authors called this behavior “juicing.”
No Tax Or Income Justification
While fund manager behavior is consistent with an underlying investor demand for dividends, it cannot be explained by either taxes or the need for income. Funds can generate equivalent tax-free distributions by returning capital instead of acquiring and then distributing dividends.
Similarly, investors in individual stocks can generate self-dividends by selling shares. If investors sell shares, generating the same proceeds that a dividend would have paid, they will end up with an identical amount of dollars invested in the company’s stock as they would have had if they’d kept all of their shares and the company instead had paid a dividend. This occurs because the company’s share price would drop, reflecting the dividend that has reduced the company’s assets.
However, while distributing cash to fund shareholders is easy, such distributions can only be labeled as “dividends” if they correspond to dividends received by the fund on its underlying securities. As a result, mutual funds must hold dividend-paying securities in order to pay out dividends themselves.
There are two ways a mutual fund can meet investors’ desire for large dividend payments. Either it can buy high-dividend-yield securities, or it can artificially increase their dividend yields by “buying the dividends” (or “juicing” them). The process involves purchasing stocks before the day on which the dividend will accrue to investors (known as the “ex-dividend day”), collecting the dividend and then selling the stock afterward.
Uncovering The Juicing
The authors noted that a few funds actually advertise their juicing behavior. In 2010, Morningstar identified an illustrative list of seven funds that explicitly describe a juicing strategy in their prospectuses. The properties of these funds are sometimes quite astounding. The authors noted that in 2009, the First Trust Dividend and Income Fund (FAV) listed a ratio of income to assets of 19.3 percent and an annual turnover rate of more than 2,000 percent.
Since mutual funds are not required to report every trade, the existence of a dividend-generating trading strategy must be deduced through indirect means.
The authors were able to infer the total dividends that a fund received based on the dividends distributed to shareholders, and then comparing this total to the dividends the fund would have received based on reported quarterly stock positions. If funds trade without regard to dividend ex-dates, the total implied dividends from quarterly reports should be an unbiased estimate of total actual dividends received.
Funds paying substantially higher dividends than indicated by their holdings are likely the juicers. Note that funds must hold stocks for at least 60 days for dividends to be considered qualified, and thus receive preferential tax treatment (qualified dividends are taxed at a lower rate than the ordinary income tax rate).
The authors found that juicing is a persistent, and thus predictive, behavior. Funds that juice in one year (i.e., have an excess dividend ratio outside what chance alone would predict) are much more likely to juice in other years. This is consistent with juicing being a deliberate behavior. When funds have a high excess dividend ratio, the average time to the next dividend for reported holdings is lower than for similar funds, and their turnover is higher.
Not unexpectedly, the authors also found that juicing is costly to investors through higher trading costs (commissions, bid/offer spreads and market impact costs). They found that funds with an excess dividend ratio of above 1.38 have 11 percent higher turnover (with a t-stat of 4.2). Funds with an excess dividend ratio above 2 have 17 percent higher turnover (with a t-stat of 4.0).
In addition, juicers incur increased taxes, ranging from 0.6 to 1.5 percent of fund assets per year. And this assumes that all dividends are qualified.
The implication is striking: “Investors who seek an income stream are better off creating it by selling fund shares than by investing in a fund that juices.” Note that there may be no cost to selling mutual fund shares as they trade at their NAV. If capital gains result from divesting shares, only the gain is taxed, not the full proceeds (as is the case with dividends).
High Turnover’s Impact
As one example of the impact of high turnover, we can examine the performance of the aforementioned First Trust Dividend and Income Fund (FAV). Morningstar reports that for the five-year period ending Oct. 8, 2015, the fund provided a return of 6.8 percent per year.
During the same period, Vanguard’s 500 Index Fund (VFINX) returned 13.8 percent. Since Morningstar classifies FAV as a large value fund, we will compare it with other passively managed large value funds. The Vanguard Value Index Fund (VIVAX) returned 12.8 percent per year and the Dimensional Fund Advisors (DFA) Large Cap Value Fund (DFLVX) returned 14.5 percent per year. (In the interest of full disclosure, my firm, Buckingham, uses DFA funds in constructing client portfolios.)
Using the regression tool available at Portfolio Visualizer, we can analyze the impact of the high-dividend strategy. For the period October 2007 through August 2015—the longest period for which the data is available—FAV produced an annual alpha of -5.6 percent. The focus on high dividends was an expensive proposition.
Despite the many issues with juicing, the authors found that funds with an excess dividend ratio greater than 1.38 received, on average, an additional 6.8 percent of inflows a year compared with other funds with similar observable characteristics. An excess dividend ratio greater than 2 is associated with an additional 12.2 percent of inflows a year.
Identifying The Buyers
Furthermore, the authors sought to identify the buyers of funds that engaged in this bad behavior. The hypothesis would be that juicing must appeal to less sophisticated, uninformed investors. And consistent with this expectation, juicing is significantly less likely for funds with an institutional share class (institutions are considered to be more informed).
In addition, dividends should be more valuable to investors with lower income tax rates, or to those who pay no income tax at all. Yet juicing is more likely to occur among retail funds, whose investors have a greater likelihood of paying income tax on dividends when compared with institutional funds (which are more likely to have tax-free investors such as retirement accounts or charitable institutions).
Not surprisingly, given the less sophisticated nature of their buyers, the authors also found that juicing is more common for funds with higher expenses (which further serve as a drag on returns).
Finally, the authors noted their results are consistent with investors who psychologically distinguish between consuming income produced by their assets and consuming the capital value of their assets. This is simply a labeling error (or framing problem) and thus leads to irrational behavior. Unscrupulous mutual funds, however, cater—or pander—to the unsophisticated investor, charging higher fees and delivering lower returns with less tax efficiency.
To prevent such behavior, or at least to expose it and thus perhaps minimize it, the authors recommended that the SEC require funds to report their excess dividend ratios. I couldn’t agree more.
Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.