An Alternative to the 50-200 Moving Average Crossover System

The stock market has historically had a bullish bias, yet commonly used moving average crossover systems such as the 50-200 EMA crossover do not take this into account when a buy or sell signal is changing. For a buy signal to be generated, the stock market has to make a move of a similar magnitude as it does to signal a sell. In my opinion, instead of using a buy signal with the reverse criteria of a sell signal, a buy should be triggered more easily than a sell.

The EOT (Either of Two) moving average crossover system is designed to address this issue. It uses three moving averages, either simple or exponential. The system triggers a buy when either the short-term moving average is above the medium-term moving average or the medium-term moving average is above the long-term moving average. I like to use the 20, 50, and 200 EMAs with this system instead of the 50-200 EMA crossover method.

The following chart shows both the 20-50-200 EOT crossover system and the 50-200 crossover system applied to the S&P 500.

Click to view larger chart.

EMA Crossovers on SPX

Although the performance is essentially the same, when I apply this strategy to the NASDAQ Composite or the Russell 2000, a meaningful difference in performance turns up.

russell 2000 EMA crossover test

NASDAQ Composite EMA crossover test

S&P 500 (starting 10/23/1950) B&H 50-200 20-50-200
CAGR 7.02 7.18 7.73
SD of rolling annual returns 14.6 10.5 11.6
Number of trades 0 54 104
Max Drawdown 56.8 33.2 33.2
NASDAQ (starting 11/23/1971) B&H 50-200 20-50-200
CAGR 8.68 8.54 10.44
SD of rolling annual returns 22.8 16.8 18.4
Number of trades 0 45 75
Max Drawdown 77.9 43.5 42.8
Russell 2000 (starting 6/23/1988) B&H 50-200 20-50-200
CAGR 7.37 6.29 8.41
SD of rolling annual returns 17.6 11.6 13.6
Number of trades 0 32 46
Max Drawdown 59.9 27.4 30.7

What does this tell us?

The EOT trend system’s returns have been higher and more consistent across several indexes than those of the 50-200 crossover strategy. However, its volatility, as well as its number of trades, have also been higher.

A few more thoughts…

This strategy should also perform well when applied to international markets and emerging markets. Emerging markets often snap back fast after a bear market so it makes sense to use the EOT system.

The EOT system can also use weekly 4, 10, and 40 EMAs. It will probably lower performance slightly, but it will take less time to implement.

SDividend Yield: A Simple Indicator to Add to the Dividend Investing Toolbox

In their research paper, “High Yield, Low Payout”, Credit Suisse analysts Pankaj N. Patel, Souheang Yao and Heath Barefoot found that the combination of a high dividend yield and a low payout ratio performed better than high yield alone.

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=946448

To be able to follow a strategy such as the one in the paper, I created a simple indicator that combines both the dividend yield and the payout ratio. It does this by “adjusting” the dividend for the payout ratio. The formula is as follows.

SDividend Yield = (Dividend*(1-Payout Ratio))/Price
Where
Payout Ratio = Dividend/Earnings

The SDividend yield can be used to compare the dividend yield and payout ratio of stocks at the same time. This is useful for sorting stocks by yield while ensuring that the yield is sustainable.

SDividend yields and normal dividend yields cannot be compared to each other. The median SDividend yield for dividend paying stocks is slightly below one, whereas that of the regular dividend yield is closer to three. The SDividend yield for a stock can also be negative. This is caused by a payout ratio of over 100%. These companies are ones with which extra caution should be taken because there is a low chance of the company continuing to pay dividends.

Although I do not think this should be the only indicator used in dividend investing, I believe it can be a very useful addition to the toolbox of a dividend investor. It is also not a replacement for the dividend yield because the dividend yield IS the actual yield whereas the SDividend yield is not. Nevertheless, by using the SDividend yield instead of the regular dividend yield, we can gain a more accurate depiction of which stocks have high dividend yields combined with low payout ratios.

The 10 stocks in the S&P 500 with the highest SDividend yield are as follows. This is not a list of stocks to buy but rather a list of stocks to do further research on.

Name Ticker Price Stock Rank SDividend Yield (%)
Merck & Co. Inc. mrk 37.71 80.910 2.93
Eli Lilly & Co. lly 36.15 81.523 2.705
Ameren Corporation aee 26.45 51.465 2.61
Constellation Energy Group, Inc. ceg 36.05 88.550 2.57
Exelon Corp. exc 44.28 65.428 2.31
Cincinnati Financial Corp. cinf 29.08 80.309 2.22
The Chubb Corporation cb 52.05 83.490 2.17
American Electric Power Co., Inc. aep 34.49 55.907 2.15
FPL Group Inc. fpl 48.84 62.396 2.13
Lockheed Martin Corporation lmt 83.88 70.308 2.1

Welcome to the VectorGrader Blog

I’ve started this blog to publish and discuss my research on financial markets, investing strategies, and trading systems. I have a quantitative bent in me, so much of what I post here will be similar in nature. My current intent is to make one or two posts per week as well as to update strategies I talk about. This may, of course, change as I do not yet know how much interesting stuff I will find to post about.

I create quantitative investing strategies for VectorGrader.com and this blog will hopefully help drive traffic. In the past, I posted research and strategy updates at VectorGrader.com for subscribers, but to make it more available I will in the future place it here instead.

I’ll probably start by posting about several simple indicators and strategies I created and how to use them. Look for the first few in a couple of days.

You may want to subscribe by email to keep up with new posts.