Implementing a Dividend Yield Screen
The current market environment has been tough on many investors. With bond yields at historic lows and a volatile stock market, many investors are left wondering where to put their money. One investment that fell off the radar screen in recent years was the dividend-paying stock. However, dividends are in vogue once again, helped in large part by last year’s tax cut on dividend payments. The rate on qualified dividends was cut to 15% from marginal income tax rates that had been as high as 38.6% (dividends on most common stocks qualify for the tax break, while dividends on REITs and preferred stocks do not). Corporations also responded to the dividend tax cut: Since the beginning of 2003, according to a recent Money magazine, 23 of the S&P 500 companies paid dividends for the first time, while 233 upped their dividend payments at least once. All told, 372 of the S&P 500 companies now pay a dividend.
Dividends are attractive to investors because they contribute to returns in any market situation. In fact, since 1926 dividends have accounted for over 40% of stocks’ total returns, according to Thomson/Baseline. In addition, the income appeal of dividend-paying stocks helps limit steep losses during market downturns. However, just because a company pays a dividend does not mean it deserves your investment dollars. The question becomes: Which dividend-paying companies are good investment opportunities?
Employing a dividend-yield strategy can help you find potentially undervalued stocks with reduced downside risk, provided the dividend is secure. Because it is mainly mature firms that pay significant dividends, dividend analysis is geared toward established firms that are past their explosive growth and cash-consuming stage.
A dividend was paid in each of the last seven years and the annual payout increased over each of the last six fiscal years;
The seven-year growth rate in dividends per share is greater than 3%;
The current dividend yield is greater than the seven-year average dividend yield;
The payout ratio for the last 12 months is less than or equal to 85% for utilities and less than or equal to 50% for companies in other industries;
The total-liabilities-to-assets ratio must be below the norm for the industry;
The three-year earnings growth rate must be greater than or equal to the growth rate for the industry over the same period.
A stock’s dividend yield is computed by dividing the indicated dividend—the expected dividend over the next year—by the share price. For most stocks, the indicated dividend is the most recent quarterly dividend multiplied by four, although some firms have switched to a single annual dividend payment. If a stock’s price rises faster than its dividend, the dividend yield will fall, indicating that the price may have been bid up too far and may be ready for a decline. Conversely, if the dividend yield rises to a high level, the stock may be poised for an increase in price if the dividend can be sustained.
Like all basic value-oriented techniques, the dividend-yield strategy attempts to identify investments that are out of favor. Contrarian techniques such as this are based on the premise that markets tend to overreact to good and bad news and push the price of a security away from its intrinsic value. Value investors hope to identify these mispriced securities through the use of a consistent set of rules.
Screening is the first stage in this process and it involves scanning a group of securities to find those that merit further in-depth analysis. Stock Investor Pro—AAII’s fundamental stock screening program and research database is used to develop and test these screens. The software includes a built-in screen that seeks companies with a history of rising, sustainable dividend payments. The criteria for the screen are listed in Table 1. Table 2 shows the 39 companies that passed the dividend yield screen as of June 18, 2004, listed in descending order by current dividend yield.
AAII tracks over 50 different stock screening methodologies at the Stock Screens area of the AAII Web site. The companies passing the dividend yield screen have been reported and tracked on the AAII Web site for over seven years with results that exceed the broad market indexes.
Figure 1 shows that the stocks passing the dividend yield screen have surpassed the performance of the S&P 500 index since December of 1997, as well as the MidCap 400 and SmallCap 600 indexes.
Overall, the dividend yield screen has generated a cumulative return of 114.8% over the period from December 1997 to June 2004. Note that this performance does not include dividend payments or dividend reinvestments.
In this article, we apply the dividend yield screen using four on-line screening tools: CNBC at MSN Money Deluxe Screener, Wall Street City Power ProSearch, Morningstar Premium Stock Screener, and Reuters.com Investor PowerScreener.
The screen begins by requiring a company to have seven years of both price and dividend records. This immediately eliminates IPOs or companies lacking a proven record of performance. The choice of time period is a balance between using one that is too short and only captures a segment of the market cycle, and one that is too long and includes a time period that is no longer representative of the current company, industry, or market. Periods of between five and 10 years are most common for these types of comparisons. Seven years was selected because Stock Investor Pro has seven years of financial statement information, including annual dividends.
From a trading record standpoint, only Wall Street City’s Power ProSearch module tracks how many weeks a stock has been traded. The dividend yield screen requires at least seven years of trading history, so with Power ProSearch, we required a minimum of 364 weeks of trading (7 years × 52 weeks).
One of the most stringent requirements of the dividend yield screen is that a company must have paid a dividend for each of the last seven years and increased it every year. Dividend levels are set by the board of directors based on consideration of the current company, industry, and economic conditions. Recently, tax law changes that lowered the tax rate on qualified dividends to 15% have led numerous companies to institute dividend payments for the first time in their history. Because dividend cuts are tantamount to an announcement that the firm is financially distressed, dividends are typically set at levels low enough that the company should be able to sustain payment throughout the economic cycle.
A lack of dividend growth or a decline in the dividend growth rate can also be troubling, especially after a period of regular dividend increases. Roughly 8% of the companies paying dividends over the last seven years have not increased their dividend payment over this period. Investors such as Benjamin Graham required that stock dividends at least keep pace with inflation. The dividend yield screen is more aggressive and demands an annual increase in per share dividend payout for each of the last six fiscal years.
Only Morningstar.com’s Premium Stock Screener allows year-to-year dividend comparisons, although it is for the last four fiscal years, not six.
Beyond requiring annual increases in dividends, the dividend yield screen also imposes a 3% annualized, compounded growth rate in dividends per share over the last seven years.
The best these on-line screening services could do was offer a five-year dividend growth rate—MSN, Morningstar, and Wall Street City all offer five-year dividend growth rates while Reuters offers a three-year.
Underlying some of the growth rates offered by various screening services are the requirements that earnings, dividends, etc., do not decline year-to-year. MSN’s Deluxe Screener, with data from Media General Financial Services (which is now offered by CoreData) has growth rates calculated using least-squares regression. Underlying the calculation is the assumption that dividends have been paid for each year of the calculation period. Therefore, the five-year dividend growth rate requires a company to have paid dividends over the last five years and at least maintained the annual dividend level over the period.
Wall Street City’s Power ProSearch also offers a unique variable that examines the consistency of dividend growth over the last five years. The field compares a company’s dividends against its history of dividends and shows whether or not they have increased each quarter. Using ProSearch’s relative screening capabilities, we required those passing companies to be among the most persistent dividend increasers. In fact, of the 25 companies that passed the dividend screen using Power ProSearch, 10 had increased their quarterly dividend 19 times over the last 20 quarters, including such companies as Abbott Labs, Allstate Corp., and Coca Cola.
The next filter requires that the company’s current dividend yield be higher than its seven-year average dividend yield. This filter seeks out companies whose dividends have increased at a faster pace than their share price, or whose current share price has declined recently relative to the dividend payment.
Absolute or relative levels may be used in screening for high-yield stocks. A screen requiring an absolute level might look, for example, for a minimum dividend yield of 3% before an investment is considered. Absolute screens can lead to passive market timing—cash levels tend to build up when investors cannot find suitable investments that meet the minimum requirements during times of market extremes. Also, screens that only look at absolute levels can be weak because they may turn up securities from a single industry that traditionally has higher dividend yields—such as utilities or REITs.
Screens based on relative levels compare the yield against a benchmark that may fluctuate, such as the current yield for the S&P 500 index. In this case, the investor does not require that the yield meet some minimum level, but instead that it maintains its historical relationship with the benchmark figure. Common screens examining relative yields include comparisons against some overall market level, industry level, historical average, or even some interest rate benchmark. This dividend yield screen is performed using a historical average as the benchmark.
Again, we were forced to alter our screen to fit the offerings of each of these on-line screening services. Two of the services—Morningstar Premium Stock Screener and MSN’s Deluxe Screener offer the ability to screen for yields relative to some benchmark. Using MSN’s screener, we screened for current dividend yields equal to or above the company’s five-year average yield. With Morningstar, we required the current dividend yield to be greater than or equal to the sector average.
With Reuters.com and Wall Street City screeners, relative comparisons are not available so we set “absolute levels” for the current dividend yield. In setting up the dividend yield screen with both of these services, we required a minimum dividend yield of 1.9%. We chose this level because it was the yield of the S&P 500 on May 31, 2004, as calculated by Barra (Figure 4). Barra presents a number of useful benchmark statistics at its Web site (www.barra.com).
Having identified companies with some history of dividend growth and/or stability in dividend payments and some relative or absolute dividend yield, you might think that the screening process is finished. However, before investing in a dividend-paying company, it is important to assess the security of the dividend as well. A high dividend yield may be a signal that the market expects the dividend to be cut shortly and has pushed the price down accordingly. A high relative dividend yield is a buy signal only if the dividend level is expected to be sustained and increased over time.
Measures exist that may help to confirm the safety of the dividend. The payout ratio is perhaps the most common of these and is calculated by dividing the dividend per share by earnings per share—it is the percentage of earnings that are paid out as dividends. Generally, the lower the number, the more secure the dividend. Any ratio above 50% is considered a warning flag. However, for some industries, such as utilities, ratios of 80% or higher are common. A 100% payout ratio indicates that a company is paying out all of its earnings in the form of dividends. A negative payout ratio indicates that a company is paying a dividend even though earnings are negative. Firms cannot afford to sustain such behavior over the long term. The dividend yield screen requires a payout of between 0% and 85% for utilities and between 0% and 50% for firms in other sectors.
Only Reuters.com allowed us to impose varying payout ratios depending on the company’s sector classification. With Morningstar, we were able to screen for companies whose payout ratio was less than or equal to its sector average, while with MSN’s Deluxe Screener we screened for payout ratios that were less than or equal to the average industry payout.
Dividends are paid out in cash, so it is also important to examine the liquidity of a company. Financial strength helps to indicate liquidity and to provide a measure of safety for the dividend payout in the event a company experiences rough times.
One must consider both the short-term obligations along with long-term liabilities when testing for financial strength. Common measures of the longer-term obligations of the company include the debt-to-equity ratio, which compares the level of long-term debt to owner’s equity (debt as a percent of capital structure); long-term debt dividend by capital, which includes long-term sources of financing such as bonds, capitalized leases, and equity; and total liabilities to total assets.
The dividend yield screen uses the ratio of total liabilities to assets because it considers both short-term and long-term liabilities. Acceptable levels of debt vary from industry to industry, so the screen looks for companies with total liabilities to assets below the norm for their industry. The higher the ratio, the greater the financial leverage and the higher the risk. The financials and utilities passing will tend to have much higher values than stocks in the consumer sectors.
None of the services had a total-assets-to-total-liabilities ratio, so we used the total-debt-to-equity ratio in all four instances. Only Morningstar allowed us to screen for debt-to-equity ratios on a relative basis—we isolated those companies whose ratio was less than or equal to the sector average debt-to-equity ratio. For the other three services, we imposed a 50% maximum value.
CNBC at MSN Money Deluxe Screener
Morningstar.com Premium Stock Screener
Reuters.com Investor PowerScreener
Wall Street City Power ProSearch
It is also important to examine the historical record of earnings. Dividend growth cannot deviate for very long from the level of earnings growth, so the pattern of earnings growth will help to confirm the stability and strength of the dividend. Ideally, earnings should move up consistently. The final filter requires that growth in earnings over the last three years exceeds the norm for the industry.
Three of the four services allow you to screen for the three-year growth rate in earnings—Reuters.com, Wall Street City, and Morningstar. However, of this group only Morningstar allows for screening on a relative basis, offering a three-year earnings per share growth rate greater than or equal to the sector average. For the others, we required a minimum growth rate of 5.1%.
Why 5.1%? Again, whenever using some absolute level—in this case a growth rate—you have a lot of discretion, so it is important to come up with a logical, and reasonable, value to use in your screen. It’s not a good idea to merely screen for those companies with the highest growth rates. Companies with abnormally high growth rates cannot sustain these rates in the long run and their stock prices tend to suffer as growth falls to the earth. Conversely, it’s unwise to merely require positive growth. With this in mind, we chose to screen for companies whose earnings are growing at a “premium” to inflation. In other words, we required a growth rate that exceeded the trailing 12-month inflation rate (3.1% as of the end of June) plus a two percentage point premium—5.1%. (The two percentage point premium is purely arbitrary.) This is the rate we used in setting up our dividend yield screen with MSN, Reuters.com, and Wall Street City. Morningstar’s Premium Stock Screener allowed us to screen for companies with a three-year earnings growth rate greater than or equal to its sectors’ three-year average growth rate in earnings.
The figures shown here display the screening criteria and screening results using the four on-line screening services. While there is some overlap of passing companies, different companies passed each screening system due to unique data, calculation methodology, and screening capabilities. Nevertheless, each screener captured the basic flavor of the dividend yield screen—identifying undervalued, dividend-paying stocks with a degree of financial strength.
However, using this screening methodology does not mean that you should buy every stock that passes the filters. Screening is only the first step when attempting to find a stock worthy of your investment dollars. The results of any screen—irrespective of the means of screening—are merely a stepping-off point for performing additional due diligence. Any of the top screeners will give you a reasonable starting point for your analysis, but it is important to understand the quantitative and qualitative shortcomings of the screening system you are using.