While perusing the contributor boards, I came across a discussion centered around the notion of high dividend yields. I focused on three very basic, but also critical, points raised in this discussion thread. The first is how exactly should you measure dividend yield. Different data sources have different approaches that result in different yields. In many cases, you can make an argument that both or more yields are correct. The second notion is the issue of the impact of pull forward and special dividends that were rushed in at the end of 2012 to benefit from the expiring favorable tax rates. The third notion is what types of dividend yields should be deemed to be high dividend yields.
Dividend Yields Depend on Definitions
The first point is simply a matter of definition if you want to stick to historical data. Yahoo!Finance uses a current approach which looks at the most recent dividend and then, when quarterly, multiplies by four and divides by the closing stock price. This approach is fine as long as dividends are consistent. However, many stocks, including the Blackstone Group LP (BX), do not pay consistent dividends. Its last nine quarterly dividends have been $0.42, $0.10, $0.10, $0.10, $0.22, $0.10, $0.10, $0.20, and $0.32. So Yahoo, dutifully reports an annual dividend of $1.68, giving a yield of 8.5% based on the $19.78 closing price. Dividend.com sometimes takes a perhaps more accurate approach and uses the last four dividends to give $0.72 or a yield of 3.6%. One at least sees some consistency in these dividends. Other companies, often ADRs, can have even more bewildering patterns. For example, Southern Copper Corporation (SCCO) pays very irregular dividends. The last six quarterly dividends have been $0.24, $2.75, $0.24, $0.53, $0.19, $0.70 - go figure. Most likely the core driver is financial performance with a significant boost to take advantage of the tax benefits. In this case, Yahoo suggests a yield of 2.6%; however, Dividend.com suggests that same result. While I wouldn't use the TTM, I would probably use something higher; however, it would require a more detailed look at SCCO's financials. The last stock I will look at is AT&T (T). A long time favorite of many dividend investors, T has increased their dividend like clockwork each year since 2009. The first quarter of the year, everyone gets a penny more per share. The most recent quarterly payment was $0.45 so Yahoo!Finance figures on a $1.80 for the annualized dividend. In this case, I would probably look at the forward yield and give T credit for the expected one cent increase for Q1 2014, this gives it an expected $1.81 for the next twelve months. The larger point here is that in assessing a stock, one should make a best efforts attempt to think about the forward performance and consider an estimated forward dividend. Historical dividends are not always the best guide to future dividends as most Frontier Communications Inc. (FTR) investors can tell you.
Tax Optimization has Boosted TTM Yields
The S&P 500, as measured by the SPDR S&P 500 Trust ETF (SPY), as a dividend yield of 2.0% based on a recent closing price of $156.67 and a trailing twelve month ("TTM") dividend of $3.18. However, it should be noted that there was a flurry of special and pull forward dividends at the end of 2012 to take advantage of more favorable tax treatment. Without question, this has had an impact on dividends and thus most likely future projections of dividends, since the pull forward dividends have now been paid out. The following table shows the current TTM dividend for SPY in comparison to prior years.
|TTM Ending*||TTM Dividend($)||Percent Change|
Source: Yahoo!Finance, Author Calculations *Note that this means that the dividend in row Q1 2013 is the sum of dividends for Q1 2013, Q4 2012, Q3 2012, and Q2 2012 for SPY.
This table shows a clear impact on the overall pull forward and special dividend effect. These items might have created around $0.20-$0.30 of increase for the current TTM period. This is more clearly seen in comparing the Q4 2012 dividend of $1.022 to the Q4 2011 dividend of $0.77, a 33% increase. The corresponding Q3 dividends also showed a 25% increase. So, on a forward basis, I would not expect to see another 10-15% increase from the $3.18, but perhaps something more like a 3-6% increase or around $0.10-$0.20 per share.
A High Yield Dividend Stock Should Provide a Yield of 4.5-5.0%
The third and final question was to look at the distribution of yields on dividend stocks. I looked at data from Zacks.com to source 7,805 stocks with positive market capitalization. Of these stocks, about 36% or 2,819 paid dividends. I then further segmented these by removing any stock with a market capitalization below $100 million, which left just 4,612 stocks, of which 55% or 2,554 paid dividends in the past year. The following table shows percentiles and dividend yields according to Zacks.com.
Source: Data provided by Zacks.com services, Author Calculations
So among dividend paying stocks, I decided the top quintile is a reasonable cut off for high yield. This gives around a 4.5-5% requirement. You can choose a different percentile. It is important to note the eliminations of non dividend paying stocks and also the lower market cap stocks, of which very few pay dividends. Thinking back to my earlier example with T, its forward yield would be $1.81 divided by the recent closing price of $36.69 or about 4.9% - right on my boundary for having a high yield. One can note that TTM, Forward and Current Dividend Yields are all very similar for T. This table would also put SPY around the 40th percentile. So only being slightly better than the market, does not really seem that good.
However, a high dividend yield is not always glorious when screening data. While T gives a high dividend yield with its steady progression of dividends and reasonably stable to increasing stock price, other stocks might show a high yield due to a collapsing stock price and historical dividend that might not exist in the future. These are, obviously, stocks to be avoided.
My follow up article will segment the data more finely to consider gradations on market capitalization and industry.
Disclaimer: This article is for informational and educational purposes only and shall not be construed to constitute investment advice. Nothing contained herein shall constitute a solicitation, recommendation or endorsement to buy or sell any security.