The decision is difficult in the face of an abundance of divergent views. In his paper “Do Dividends Really Matter?,” the renowned Nobel Laureate Merton Miller argued that “the seeming evidence that dividends do matter. . . is not to be trusted. It’s an optical illusion.” He claimed the apparent effect of dividends on share prices is not caused by the dividends per se, but is merely the market’s recognition of what the dividends communicate about investment policy and future earnings trends. Given this, it would seem an open-and-shut case that dividends don’t matter.
Then why the fuss? One main reason is that many investors pester management by demanding new or increasing dividends. Those investors cite studies showing that dividend payers are better than nondividend payers at delivering total shareholder return (TSR), which reflects a company’s stock-price change plus the dividends paid during a period. A flurry of such recent studies has attracted attention by the press and in the boardroom. What should a CFO do?
It depends. Our capital-market research on the dividend policies of the 1,000 largest nonfinancial companies, excluding those that were not public for the full decade of the 2000s, indicates there is no one-size-fits-all answer.
Our research shows that dividends have a positive effect on valuation during a financial crisis and a negative effect when markets are booming. When economic confidence is low, investors fear the worst and prefer the cash in hand. In boom times, however, there is more confidence that companies will invest wisely.
To determine this, we measured corporations in terms of enterprise value to gross operating assets and compared that to the cash-on-cash return on capital. There is a strong correlation between those measures of valuation and return, as documented in “Postmodern Corporate Finance” in the spring 2010 Journal of Applied Corporate Finance.
Some companies are valued higher or lower than their returns alone would imply. We tested to find if dividends influenced those premiums or discounts. From 2004 through 2007, a majority of companies trading at a premium were nondividend payers. This flipped in the first quarter of 2008, and from then until today a majority of companies trading at a premium are dividend payers. The gap is closing and seems to be reversing.
This cyclical dividend influence partly explains why recent share-price performance has been better for dividend payers. Dividends went from reducing valuation to increasing valuation, benefiting the capital gains of dividend payers. Depending on the forward trend in the economy and stock market, this may flip again — meaning that this may not be the best time to initiate a dividend.
Do some company share prices benefit more from dividends? If so, when do dividends really matter? In principle, investors should prefer higher dividends from companies with few desirable investment opportunities so that the investors themselves can redeploy the capital elsewhere. And companies with an abundance of desirable growth opportunities should maintain low or zero dividends. From the outside, we cannot assess the quality of investment opportunities. We must thus characterize companies by the returns and revenue growth they have generated historically to serve as a guide as executives look forward.
We evaluated average cash-on-cash returns over the past decade to separate our data base into high- , medium- , and low-return groups. We then divided each of those groups based on whether a company experienced revenue growth above or below the 8.1% median annualized revenue growth for the sample. Each company was classified as a nondividend payer, a low payer, or high payer based on whether it was below or above the median dividend as a percent of after-tax operating cash flow.
Regardless of their cash-on-cash returns, the low-revenue-growth companies delivered a higher TSR if they paid a dividend. The size of the dividend did not matter in terms of the TSR. This was most significant for the low-return group, where both low-dividend and high-dividend payers delivered a median TSR of 2.5% per year while the TSR for nondividend-payers was -11.2% per year, a gap of 13.7%. For the medium-return group, the high dividend payers delivered 6.9% more TSRs per year than nondividend payers, and for those with high returns this was 3.5%. Dividends seem to help the share prices of all low-growth companies, but the benefit was larger for low-return companies than for those with high returns.
This pattern reversed for high-revenue-growth companies. Comparing the high-payers to the nondividend payers, the TSR gaps for the high- , medium- , and low-return groups were -0.7%, -3.6%, and -0.5%, respectively. For high-growth companies, there seemed to be a drag on TSRs from paying dividends, though the data is more scattered as company-specific circumstances vary more.
What are the implications for CFOs contemplating dividend policy? First, you should recognize that changes to dividend policy will be heavily scrutinized by investors, analysts, and the press. But that’s no reason to avoid desirable changes. Also, remember that your reinvestment rate and future earnings stream will influence your share price more over time than your dividends, so keep your capital-deployment priorities straight. Given the research, here are some guidelines:
1. If your company doesn’t earn the required return and you expect growth to be lower than 8.1% per year, then you should absolutely pay a dividend.
2. Even if your company does earn a high return, if you are not rapidly growing, your TSR would probably benefit from paying a dividend.
3. If you expect to grow rapidly, dividends may constrain your TSR, particularly if they consume cash you could have devoted to funding profitable growth. This is even true if your returns are currently below the required return.
Consider your prospects for deploying capital in high-return investments over the next few years. If the amount of expected investments is a large percentage of the cash you generate, you should be less inclined to pay dividends. But recognize that if you have had returns persistently below the required return, your share price may be penalized if you do not pay a large enough dividend.
For companies with fewer desirable future investments, dividends should be more attractive. But if you have been aggressively investing in growth and have delivered returns well in excess of the required return, a dividend increase may hurt the share price. That’s because investors may conclude you have run out of desirable investments.
If you are on the fence — and if you believe, as I do, that the economy and financial markets are likely to improve over the next few years — then more modest dividend increases may be warranted as dividends may become less important for a while. This will leave more cash available for reinvestment in future growth.
