I have written a number of articles on the dividend yield model of the S&P 500. The most recent article is here and the original article is here. The model has been reasonably reliable since the Fall of 2009. Using the model, the fair value of the S&P 500 Index (the Index) has been a price which produces a dividend yield of 2.0% on trailing dividends. The Index has traded at prices producing yields between 1.8% and 2.2% and buyers of stocks at times when the dividend yield on the Index is at or near 2.2% have generally done very well. In August of last year and again early this year, the Index broke out of the range indicated by the model with price declines drove the dividend yield over 2.2% but the market snapped back quickly to price levels within the range indicated by the model.
The trailing dividends for the Index were $44.32 as of the end of May. The dividend level has consistently trended higher so that by the end of this week it will probably be a higher number. Using the $44.32 dividend level, the table below provides the dividend yield as of Monday's close as well as the price levels needed to produce various other dividend yields.
|New Fair Value||$1970||2.25%|
Using the model that has worked since the Fall of 2009, Monday's close was slightly below the floor level and a strong buy signal. The middle of the range would produce a yield of 2.0% and a price of $2216 for the Index which would be a big move up from here. If the Index traded up to $2462, that would constitute a strong sell signal. I have added a "new fair value" row to represent a shifting of the model to reflect the recent increase in short term rates implemented by the Federal Reserve.
I think that the model works reasonably well because yield hungry investors start to pile into reliable dividend paying stocks as soon as the prices dip and the yields automatically rise. This has started to happen this week. Investors have been rewarded because dividends have increased steadily since the recovery began and the market has simply moved up with the dividends to keep the yield relatively constant.
Risks to the Model - I have consistently said that there were two things which could derail the model. The first of these is higher interest rates. If rates increase, yield hungry investors will migrate to bonds and will "insist" on higher yields from equities by driving equity prices down to produce those yields. The second risk is a disruption of the pattern of higher dividends through either actual dividend declines or a cessation of dividend increases.
Interest Rates - After some 6 years, the Federal Reserve finally raised short term rates in December. It is unclear whether or not this will disrupt the model. Long term rates (rates on 10 and 30 year Treasuries) are now actually at some of the lowest levels since the Crash. Put another way, investors have had lots of opportunities to desert equities and earn much higher risk free interest rates on bonds over the past 7 years than are available now. In the past, investors have tended to switch back to stocks when the yield on the Index reached 2.2% even though the interest rate on 10 year Treasury bonds was much higher than it is now. I added the "new fair value" numbers to the table to reflect the possibility that the model will reset based on higher short term rates, but I do not think that this is actually what will happen. The higher short term rates have not produced yield opportunities in bonds which really compete with stocks at the current equity price levels. A recent commentator pointed out that some two thirds of the stocks in the Index have dividend yields higher than the interest rates on 10 year Treasuries.
Dividend Increases - I take the second risk much more seriously. We have definitely seen a slow down in dividend increases. In the early stages of the recovery dividends for the Index were increasing by more than 10% a year and I even suggested a 1% per month rough rule of thumb. In the past year (through the end of March) dividends increased by only 6.6% - still a solid trend but definitely not the powerful pattern of increases seen in earlier years. On closer examination, the problem has been concentrated in the energy sector where dividend decreases and suspensions have been common. Because the dividend number is an aggregate for the entire Index, the downward trend in energy stock dividends pulls down the average. If we exclude energy stocks, it is likely that the dividend increase rate for the rest of the stocks in the Index would be near 10%. An owner of large cap stocks who stayed away from the energy sector very likely saw year over year dividend increases of nearly 10%.
As time passes, the potential for more energy stock dividend decreases and suspensions and its impact on the Index will lessen for two reasons. Stocks which have already suspended their dividends cannot produce further dividend declines. In addition, the energy sector has become a smaller and smaller percentage of the Index (which is market cap and liquidity weighted) - dropping from 12.27% in 2011 to 7.1% as of the end of this May.
On the other hand, there does seem to be some evidence of a slowing of dividend increases - which would make sense given the fact that earnings are also slowing or even declining. The table below compares 2016 year to date (as of the end of May) statistics with statistics for the full year 2015 concerning the number of companies in the Index which have increased dividends, decreased dividends, stopped paying dividend and started paying dividends.
Given that - as of the end of May - we were not even half way through the year, the dividend increase and dividend start numbers look encouraging and suggest that we may even exceed the numbers for 2015 by the end of the year. But the dividend decrease numbers are already near the level they reached during the full year 2015 and so are the dividend stop numbers. Almost all of the stocks with dividend decreases or stops were energy stocks - most notably ConocoPhillips (NYSE:COP) which cut its dividend from $2.96 a year to $1.00. We seem to have developed a bifurcated market with the energy sector pulling the dividend numbers down and with everything else performing consistent with post-crash patterns. I think that we will see the pattern of dividend increases continue with less downward pressure from the energy sector but with more resistance in other sectors due to subdued earnings.
Conclusion - I think that the model still has legs. If the price declines in stocks drive the dividend yield near or above 2.2% (an Index price of 2015), there will be a strong tendency for yield hungry investors to pile into the market. This will show up initially in strength in those stocks with high dividends and with a reliable history of dividend increases, or at least not dividend decreases. It will tend to push the Index up and other stocks will follow - with the exception of the energy sector, which will be driven by the oil price. It is, of course, possible that in a panic prices will decline to drive dividend yields above 2.2% but I do not think that these price levels will persist until we have either a pervasive pattern of higher interest rates or a cessation of the pattern of aggregate dividend increases for the Index.
Disclosure: I am/we are long COP.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.