There has always been a great deal of attention to price earnings ratios in calculating whether the market is expensive or cheap and in predicting future market direction. However, since the Crash of 2008-09, the S&P 500 appears to have settled into a relatively narrow range with respect to an entirely different ratio. Come hell or high water, the S&P 500 seems to want to trade with a dividend yield between 1.76 and 2.20.
The table below provides monthly S&P 500 dividend yields for the period of time since the market emerged from the Crash. After trading at relatively high dividend yield ratios, starting in September 2009, the S&P 500 has settled into a fairly narrow range. I am providing data regarding dividend yield on the first of each month for the past 4 years (the February 2013 entry is based on data as of February 12). There was fluctuation in the yield within each month, which is not captured by this table, but the table does show that, even after such fluctuation, the ratio returned to being range bound rather quickly. My source is the website, multpl.com, and the data can be found here. The ratios are based on data from S&P itself.
The dividend yield of an Index can be complex to calculate. The S&P 500 experiences entrances of new stocks and exits of old stocks all the time, and the addition of a REIT to the Index can increase the dividend yield of the Index even though none of the stocks in the Index have increased their dividends. Similarly, a change in the relative weighting of various stocks due to price fluctuations can change the dividend yield of the Index. The Index contains some very large cap stocks that pay no dividends -- e.g., Google (GOOG) and Amazon (AMZN). It also includes quite a few REITs and electric utilities, which tend to pay high dividends. In that sense, the dividend yield of the index is a fairly ambiguous ratio not necessarily tied to the value of any stock or group of stocks. Given this diversity, the stability of the ratio is all the more surprising.
In following individual stocks, I have found that -- with the exception of special situations like BDCs, REITs, MLPs and utilities -- it is getting harder and harder to find stocks yielding more than 4%, unless there is some special "problem" with the company. There are some exceptions -- like AT&T (T) and Verizon (VZ) -- but not many. A number of stocks seem to gravitate around a yield of 3%, with the stock trending up as the dividend is raised. This is not an absolute rule, but it appears that yield-oriented investors may be beginning to "clear the market" for yield in a narrower and narrower range.
Of course, the general trend since the Crash ended has been for dividends to increase. Some of the large banks omitted dividends for some time during and after the Crash while they were rebuilding capital, and the resumption of these dividends will tend to give the dividend payout of the Index quite a boost. In addition, steady "dividend champion" stocks, which increase their dividends through thick and thin, will push the payout higher. The fact that the recent changes in the tax law have a degree of permanence and continue a somewhat advantageous tax treatment for dividends also may encourage higher dividends.
I think more research on this phenomenon is warranted. Historically, the dividend yield on the Index has been much higher than the current range reaching an all time high of 13.84% at the bottom of the market in 1932. It stayed at levels considerably higher than the current range for years until the 1990s, and reached a low of 1.11% in 2000. Is there a plausible reason for the range to be lower than it was, for example in the 1950s, when it traded in the 3-5% range? Has the migration of REITs into the Index had a major effect? How much potential is there for a fairly sudden increase in dividends if banks resume traditional dividend levels? Is some other factor at work here?
As long as the market stays within the above described range, we can identify certain phenomena with a degree of mathematical certainty. If the trend of higher dividends persists, then one of two things must happen. Either the dividend yield of the index will increase and exceed the range, or the price of the index will increase and the dividend yield will stay within the range. Recent experience suggests the latter. Right now, we are actually above the middle -- 1.98% -- of the range, suggesting that the market may trade up by about 3% to 1550 to hit that mid-point in the range. Watch dividend action closely; it may signal the direction of the market with surprising accuracy.
Now, for the bad news. If this analysis is correct, then stocks are increasingly behaving like bonds and trading on a yield basis. To the extent that this is true, higher interest rates could move that dividend yield range much higher. Of course, if dividends themselves increase sufficiently, the market might support both higher dividend yields and higher share prices. But there is now somewhat of a risk that the stock market itself is subject to many of the same "interest rate" risks that spook the bond market.