Dividends aren't the only way that firms return capital to shareholders.
There are about 30 exchange-traded funds focused on dividend-paying stocks, but only two focused on stocks that buy back their own shares. While academics might claim that investors should be indifferent between stocks that pay a dividend and those that buy back shares, fund providers know that investors like the allure of dividend funds. But there is good reason to pay attention to firms that reduce their shares outstanding.
Firms have three choices to deploy excess cash: invest back into the business, distribute the cash as a dividend, or buy back shares. Expected returns among the three options should be the same, at least according to the of Modigliani-Miller dividend irrelevance theorem. But that theory holds only under some strict assumptions, chiefly that firm managers will be good stewards of shareholder capital and will invest only in projects that earn rates of return at least as high as the firm's cost of capital.
In practice, we know that managers are often poor stewards of shareholder capital, more concerned about empire building than achieving a high and sustainable return on invested capital, so dividends are an important way to keep management on a tight leash. If management is sitting on a hoard of cash with a bloated balance sheet, it does not need to seek the approval of the capital markets to fund new ventures; it can just raid its internal corporate piggybank.
Over the last few decades, there has been a large shift in corporate finance away from dividend payments. According to Standard & Poor's, about 75% of stocks in the S&P 500 pay dividends, down from 94% in 1980. At the same time, share buybacks have become more popular. Between 1999 and 2004, companies in the S&P 500 spent about the same amount of money on dividends as buybacks. But between 2004 and 2010, they spent about 60% more on buybacks than they did on dividends.
In an efficient market, a share repurchase should have the same effect as a dividend. When a firm pays a dividend, its stock price should go down approximately by the amount of the dividend. So, the higher dividend yield is offset by a lower capital gain. If a firm does not pay a dividend but instead buys back stock, the dividend yield will be lower, but the capital gain will be higher. Investors in nondividend-paying stocks who prefer the periodic income that dividends provide can sell a portion of their holdings to replace the missing dividend.
Under current law, the top tax rate on long-term capital gains of 15% is the same as the top tax rate applied to qualified dividends. But in 2013, those in higher income tax brackets will likely face a higher tax rate on dividends than capital gains. If the tax rate on dividends is higher than the tax rate on capital gains, management will prefer to buyback shares. A more dubious reason for management to utilize buybacks rather than dividends is that a buyback increases the value of management stock options while dividends dilute it. This is why certain industries tend to make liberal use of management stock options, such as tech, to see more buybacks than dividends. So when screening for firms that utilize share buybacks, it is important to make sure that the net amount of shares outstanding actually decrease, rather than screening for cash flow spent buying back stock. Otherwise, some of that cash flow used to buy back stock may not be enough to offset the dilution from new issuance as part of employee-stock ownership plans.
According to data from Kenneth French, the dividend-paying stocks in the top third of all payers returned 11% per year since 1927 compared with just 8% per year for nonpayers, resulting in an ending wealth that was 8 times greater for the high dividend-paying group. So we know that dividend-paying stocks have vastly outperformed nonpayers over the long term. What about companies that buyback shares?
Jason Zweig recently warned about buybacks in The Wall Street Journal: "At their worst, buybacks can be a form of corporate cannibalism. Often the unspoken motive is to use extra cash to boost earnings per share by reducing the number of shares among which the company's profits are divided.” But that is exactly the point--better to boost earnings per share through a lower number of shares outstanding than waste it on ill-advised acquisitions.
According to a study by Bali, Demirtas and Hovakimian, between 1972 and 2002, firms that repurchased their shares outperformed firms that issued new shares by 6% per year. A share repurchase is an important signal of management confidence. If managers were uncertain about business conditions, they would prefer to issue equity to increase their capital cushion. If they are confident, they will issue debt and buy back stock, which will boost their leverage and crank up their return on equity. Among Dow 30 stocks over the past year, the 15 firms that have repurchased the most shares have returned about 9%, compared with about 3% for the 15 firms in the bottom half. International Business Machines IBM, Intel INTC, Pfizer PFE, Wal-Mart Stores WMT, and Verizon Communications VZ were among the top repurchases, but so was Hewlett-Packard HPQ, which dramatically reduced its share count only to see its share price fall further. Bank of America BAC and Alcoa AA were among the largest issuers of new shares.
The best ETF option dedicated to firms that buy back their shares is PowerShares Buyback Achievers PKW. The fund has achieved a 5-star rating over the past three years by seeking stocks that have bought back 5% or more of its shares outstanding in the previous year. It follows a modified market-cap-weighted approach, which caps stocks to no more than 5% of the portfolio at each rebalance date. At 0.60%, the fund is expensive compared with dividend-themed ETFs, which typically charge between 0.20% and 0.50%. It currently holds 142 stocks, including WMT and IBM. With only $59 million in assets under management and 14,000 shares trading daily, it is somewhat illiquid, so bid-asks spreads can be wide. Unlike most dividend funds, which typically have a value tilt, PKW is decidedly core.