Swedroe: Madness And Dividends

Careful thinking about dividends would lead investors to think of them quite differently.

Today’s post will continue the discussion on dividends and the complications and confusion that often surround them. Hopefully, the detailed examples will shed light on some of the denser components.

In The Real World There Are Frictions

Unfortunately, we don’t live in a world without frictions. And where frictions exist, they can lead us to prefer capital gains over dividends, or vice versa. For example, if dividends are taxed at higher rates than long-term capital gains (as they once were) taxable investors would have a clear preference for capital gains.

But even with the current tax regime, taxable investors should prefer capital gains, because when creating a self-made dividend, you could sell only the shares that would receive long-term capital gains treatment, and taxes would be due only on the portion that is a gain, not the total amount. With dividends, you’re taxed on the full amount.

In addition, investors can manage taxes by choosing the highest cost basis lot to sell, minimizing taxes. And if there are losses on the sale, the investor gains the benefit of a tax deduction. Even in tax-advantaged accounts, investors who diversify globally—which is the prudent strategy—should prefer capital gains, because in tax-advantaged accounts, the foreign tax credits associated with dividends have no value.

Finally, if dividends were throwing off more cash than needed to meet spending requirements, the total return approach would benefit from not only the time value of not having to pay taxes on the “excess” amount of dividends, but dividends could also push investors into a higher tax bracket.

It’s also possible that investors could prefer dividends. The reason would be transactions costs. Let’s assume that all assets are in tax-advantaged accounts and all assets are domestic. In this situation, taxes don’t create any frictions. If the creation of self-made dividends does incur transaction fees that could have been avoided by a cash-flow approach—as in dividends providing the cash flow—then those fees could lead to a preference for cash dividends.

In today’s world, with very low commission rates, and with many mutual funds trading at no costs, this is not likely to be much of an issue.

There’s one other point to consider. There often isn’t much difference in the yield of dividend strategies versus the yield on either more market-like strategies or the yield on value-oriented strategies. For example, the current yield on the SPDR Dividend ETF (SDY | A-72) is about 2.3 percent. The current yield on Vanguard’s Total Stock Market Fund is about 1.7 percent, just 60 basis points percent lower than SDY’s.

For a typical 60 percent stock/40 percent bond portfolio, to make up for the lower dividends, an investor would have to sell just 0.36 percent of the stock portfolio to create that self-made dividend. And if the yield on Vanguard’s Value Index Fund is about 2.2 percent, just 10 basis points lower than SDY’s yield, the investor would have to sell just 0.06 percent of his stock portfolio.

For those interested in learning more about the differences between a cash flow and a total return approach using home-made dividends when needed to supplement the interest and dividends generated by the investment portfolio, Vanguard’s research team has a paper on this subject. The writers recommend using a total-return approach with the investor first deciding on an asset allocation that’s based on his or her unique goals and objectives, as well as their ability, willingness and need to take risk.

“This decision should be the investor’s highest priority,” the paper says.

They should then stick to their plan, rebalancing along the way as needed.

The Math Of Cash Dividends Vs. Homemade Dividends

To demonstrate the point that the two are equivalent, we’ll consider two companies that are identical in all respects but one: Company A pays a dividend and Company B does not.

To simplify the math, we’ll assume that the stocks of both companies trade at their book values (while stocks don’t always do that, the findings would be the same regardless).

The two companies have a beginning book value of $10. They both earn $2 a share. Company A pays a $1 dividend, while Company B pays none. An investor in A owns 10,000 shares and takes the $10,000 dividend to meet his spending requirements. At the end of year one, the book value of Company A will be $11 (beginning value of $10 + $2 earnings - $1 dividend). The investor will have an asset allocation of $110,000 in stock ($11 x 10,000 shares) and $10,000 in cash for a total of $120,000.

Now let’s look at the investor in Company B. Since the book value of B is now $12 ($10 beginning book value + $2 earnings), his asset allocation is $120,000 in stock and $0 in cash. He must sell shares to generate the $10,000 he needs to meet his spending needs. So he sells 833 shares and generates $9,996. With the sale, he now has just 9,167 shares. However, those shares are $12, so his asset allocation is $110,004 in stock and $9,996 in cash, virtually identical to that of the investor in Company A.

Another way to show the two are equivalent is to consider the investor in A, who, instead of spending the dividend, he reinvests it. With the stock now at $11, his $10,000 dividend allows him to purchase 909.09 shares. Thus, he now has 10,909.09 shares. With the stock at $11, his asset allocation is the same as the asset allocation of the investor in B; namely, $120,000 in stock.

It’s important to understand that Company B now has a somewhat higher expected growth in earnings because it has more capital to invest. The higher expected earnings offset the lesser number of shares owned, with the assumption being that the company will earn its cost of capital.

There’s one more issue that should help to clarify why dividend-based strategies are not optimal.

The Explanatory Power Of Dividends

For the past 20 years, the workhorse model in finance has been what is generally referred to as the Fama-French four-factor model—the four factors being beta, size, value and momentum. The model explains the vast majority—more than 90 percent—of the differences in returns of diversified portfolios.

If dividends played an important role in determining returns, then the four-factor model wouldn’t work as well as it does, since dividends are not one of the factors. If, in fact, dividends added explanatory power beyond those of these factors, we would have a factor model that included dividends as one of the factors. But we don’t.

The reason is that stocks with the same “loading,” or exposure, to the four factors have the same expected return regardless of their dividend policy. This has important implications because about 60 percent of U.S. stocks and about 40 percent of international stocks don’t pay dividends.

Thus, any screen that includes dividends results in portfolios that are far less diversified than they could be if dividends were not included in the portfolio design. Less diversified portfolios are less efficient because they have a higher potential dispersion of returns without any compensation in the form of higher expected returns, assuming, of course that the exposures to the factors are the same.

Larry Swedroe is director of Research for the BAM Alliance, which is part of St. Louis-based Buckingham Asset Management.