Jan 1 2008 Newsletter
Dividend Collection the Hedged Way
by
Rob Friesen
As we face a year of expected dividend cuts by some banks, brokers, REITs, homebuilders, and mortgage related companies, we will see the impact on those stocks. Stocks of companies that eliminate or reduce dividend can drop very quickly so factor in additional risk if you are trading these industry groups. Recently WM cut its dividend from .56 to only .15 and many analysts think that C with a current dividend yield of 7.7% is next. I hear that the options market is pricing in dividend cuts for WB and BAC also. Whether we can predict which stocks will be the ones cutting or eliminating, we know that the years of growth of dividends for the SP500 companies is now under assault and those companies that continue to thrive will be command a premium.
That being said, what are the opportunities for you as a trader to collect dividends from the consistent payers on an ongoing basis? Is it a risky strategy? How do you begin?
A dividend is a distribution of a
portion of a company’s earnings to shareholders, and can be in the form of
cash, stock or property. With mutual
funds, interest and dividend income from its portfolio holdings along with
capital gains from stock appreciation and trading activity is paid out as a
year-end dividend to the fund shareholders.
Dividend paying stocks have a dividend amount which can be paid out quarterly or annually. They also may have special dividends that come along annually or periodically. Traders and investors don’t just look at the dividend amount in dollars but also the dividend yield, which as a financial ratio is the annual dividends per share divided by the current price of the stock. This is our “bang for our buck” assessment. *Please note that very high dividend yields could mean very depressed stock prices which may lead into companies cutting or eliminating dividends altogether. What we are looking for is a high quality company, with consistent, stable and possibly growing earnings that will lower our risk and keeps that dividend flowing.
In pairs trading, we collect dividends two ways. The first is where the quality of the company is more critical as we are holding a core position in a pair for extended periods of time, as part of a portfolio. Not recommending it…but as an example, let’s say we are long BAC and short JPM in a pair combination. BAC forward dividend is 2.56 (trailing was 2.40) with a dividend yield of 6.40% whereas JPM’s is1.52 (trailing is 1.44) with a dividend yield of 3.60%. This pair could be held in a share neutral manner and you would make 2.80% for the year by doing nothing. (Assuming both stocks stayed in relative similar performance to each other). It would increase if you ran dollar neutral on the pair which means you would own 105 shares of BAC for each 100 of JPM. We have found interesting and rewarding pair combinations where the share size differences give us an additional advantage. For example on a different pair, we are trading it 100 shares by 200 shares (100 x 200) in order to capital balance. The stock we are short is 100 shares and we pay a .50 annually on that 100 shares but we collect 1.00 annually times our 200 shares (the stock we are long). This means we net out $1.50 for the year on this combination.
The second strategy is the collection of dividends by overnight swing trades, with no intention of long-term holdings. Unlike the first strategy, in this one the quality of the company many not be as critical if the trader is only holding overnight 4 times a year.
On Ex-Dividend Day (ex-date), the stocks previous close price is adjusted (marked down) by the dividend amount that would be paid on the upcoming Payable Date. As there is an upward bias to stocks over the long sample we note it assists slightly in offsetting the ex-dividend (XD) markdown. According to Ned Davis Research, Dividend-Paying S&P500 stocks gained 10.19% per annum between 1972 and 2005 vs. 4.39% for Non-Paying. We note that in flat markets, with quality companies, we should be able to keep some of the dividend; observations and experience reveal around 8%, by taking the stock long overnight going into the ex-date. In a bull market it may increase to around 12 to 15% as an average for all the stock dividends you would go after. In a bear market it gets more treacherous as stocks usually become much weaker after XD. Adding commissions and risk, one can see that you would need a lot of capital, a very high dividend yield and the right market conditions, in order of having any chance of making money on a naked stock for dividend capture.
So how do we increase our safety and our performance and make it a viable strategy? By pairing off our long stock with a strategic or normal short and taking the pair overnight into ex-date morning.
First I will talk about the normal short and then we will get to the strategic. Let’s choose LLY and PFE for the normal example. LLY pays 1.88 per share annually for a 3.5% yield. LLY has raised its dividend 41 years in a row and as a bonus in having the wind at your back raised its guidance 3 times and projects a banner year in 2008. It projects that 7 of its drugs will reach a billion dollars in sales this year. This company’s stock in and of itself could be a defensive play for investors. Why I like this situation for what we do in dividend collection is that any dips on the stock have the probability of investors and traders stepping in to buy. We have also observed that the Wall Street machine supports dividend collection. Many of our traders inform us that the stock they took overnight into ex-date had an analyst upgrade on it that morning. The stock opened up significantly. They could find no other news behind the stock. As a trader, you should not count on this happening to you each time, but just note it as a positive benefit to our strategy in general. PFE pays 1.28 annually for a yield of 5.5%. The company, like LLY, has also hiked its dividend 41 times in a row. They are in fact correlated competitors with competing drugs like Lilly’s Cialis against Pfizer’s Viagra. There is room to trade this pair for production on a regular basis, as well as carry a core position to accomplish the dividend yield spread I mentioned in the first strategy, as well as the opportunity to have the pair on overnight 4 times a year for PFE and 4 times a year for LLY. There respect ex-dates arrive at different times. PFE is first on FEB 6 this year and LLY is next on FEB 13, so mark your calendars…
In practice: On Feb 5 you take PFE long and LLY short and take the pair overnight into Pfizer’s ex-date. PFE previous day’s closing price would be adjusted down by the quarterly dividend amount; 32 cents. This is a real adjustment to your holdings. If you had purchased 1000 shares at the close on Feb 4, you would now have lost $320 as it is .32 lower than what you paid for it. (Note that you have an offsetting dividend coming)
In observation: Traders see that PFE is discounted, many are not even aware that it is ex-date morning. Pair Traders see that the pair has shifted to a discount of PFE to LLY and again many are not aware that it is ex-date. This is also a Pair that is widely traded and has proven itself to be elastic and mean reverting over the years. Sector traders may also notice a discount of PFE to LLY.
The theoretical: Traders and investors whether they know about the dividend or not see a discount and buy PFE. Pair Traders see the pair discount and buy PFE and short LLY. This brings the spread between the two stocks in slightly which allows the Pair Trader to keep some of the dividend. So why not just put the pair trade on in the morning and forget about the overnight altogether? Excellent question! It’s a case of opportunity lost versus the risk. If a trader combines two good stocks together in the form of a pair, the market risk is reduced although there is still stock specific risk. Our experience has been that the trade is rarely a loss, often a wash, and often a significant profit for a short-term trade. In other words, the money lost on the mark down, is made up by the dividend so there is no loss except for transaction costs and time (that is the time until the dividend payable date) You are without the dividend until this date. More on the opportunity lost: In a perfect world the spread comes in all the way allowing the Pair Trader to keep 100% of the dividend. In an “alpha perfect world”, PFE gets upgraded by 3 analysts that cover the pharmaceutical sector and the stock jumps $1.00 allowing you to profit on the pair trade as well as keep the entire dividend.
So this strategy supports a variety of scenarios like; lose slightly or wash or get some of the dividend or get the entire dividend or get your cake and eat it too. You may even make money on the spread before the close on the day before ex-date and not even take it overnight. In order for any strategy to work properly the losses must be contained. I recommend that just as you initiated the trade with validity, you continue to ask the question: is this trade still valid? Remember, your purpose in the overnight trade was to collect a portion of the dividend; NOT to marry a losing position!
Now for the art-form: If everything were as mechanical as buying the div stock and shorting something against it and making money on an overnight trade, we could automate it all and kick our feet up. The truth is that in time has to be spent in research and pairing the best choices of stock together for maximum yield. Also, in order to minimize your downside you must get the spread on for a great price. Even if the dividend play does not pan out, you still want to exit that pair with minimum damage. Conduct your research like it is the only thing that matters and at the same time choose your entry and execute as if that is the only thing that matters. Some stocks have incredible runs on the day before the ex-date. I can recall one trade on CVX – XOM where CVX had a 2.00 move up. I was going to try and collect the dividend of .50 but by the end of the day, after the spread went in my favor 2.00, I decided to take that profit and not hold it overnight for the .50. It looked to me that CVX run up a bit too much and that could open up my risk the next day. The goal in trading is not to build a perfect widget (system) but to make a profit, and in this case I had entered the trade for the purpose of making money from the dividend, and the stock decided to give me my profit that day instead of the next…and I am o.k. with that. J
The Strategic Dividend Play: This strategy involves finding a short against your dividend long that provides the maximum potential bang for your buck. It’s different than just shorting PEP against KO every time the KO dividend rolls around or visa-versa, but rather involves looking for the most strategic short. It could be the recent trade short C against long BAC, JPM or WFC that some of our pair traders were doing. The flexibility here is that if your trade goes right, you may be able to extend this trade out for many days, weeks or months. So what started as a valid dividend trade provided extra validity for a longer term trade. You might look for post-event set-ups, extremely overvalued red-flagged stocks, or use an ETF as some traders have, although you must reduce expectations of mean reversion using ETF’s against a stock. It may work better with something like XOM or CVX against XLE (Energy Select ETF) as XOM as a component is19.57% of XLE while CVX is 12.33%. On the other hand if you were trading WFT; it only makes up 2% of XLE so it would be more strategic to short SII against WFT, benefitting from a higher correlation than the use of XLE as the hedge.
Are these forms of Dividend collection new strategies for Traders/ Pair Traders?
No. I have been teaching these strategies in classes in public since 2001.
They are also written about in the Art of the Arb manual and in previous newsletter publications. In my upcoming classes I will teach on increasing your edge further by the pairing of multi-strategies together and how to identify high probability pairs and trades.
Please note that the Advanced Pairs Class in March, conducted by Dave Sheldon and I will include software designed to collect, analyze and implement these strategies.
I recommend you start by becoming aware of the dividends that are available and then forming basic relationships of common household name stocks and putting those together in an effort to collect some. Start with small size and learn the structure, nuances and the art form that is required. We have many combinations of pairs in our database on www.pairtrader.com along with their dividend amount, date and yield. Remember to always check 2 or 3 sources of data for the ex-date and dividend amount, as errors have been noted occasionally.
Let’s harvest these while we have an edge! We have always tried our best to collect these but over the last 18 months traders have noted increasing productivity and yield. What is a healthy expectation for this currently? As I need to live in the percentages and not the absolutes, my current expectation is to retain 35% of the dividend on a basic, no frills, and overnight dividend collection trade. Let’s put some very conservative numbers to this: If you have a dividend yield of 1% quarterly (4% annually) on a $50 stock you would have a dividend amount of .50 quarterly ($2.00 annually). Pair this stock together with another $50 stock and take a position of 2000 x 2000 shares. Let’s say you do not realize a significant profit before the close on the day prior to ex-date and therefore take it overnight. Using the 35% expectation you have a gross profit of .175 per share or $350.00. Using costs of $15 on the 2000 shares and round-tripping the pair, we have approx costs of $60. Some traders have to add in overnight holding costs also. So on a basic no frills strategy, our expectation would yield $1160 annually on this one stock being in the trade for 4 to 10 nights a year. I added a few nights incase you take a weekend trade. Do note that this will reduce your yield due to higher weekend risk and increased holding costs. Let’s drop our yield down to 3% as an annual average but expand our business to include an average of ONE trade per day. Not too hard to do from the universe of stocks we have on the NYSE, NASDAQ and AMEX that pay healthy dividends. Using the $50 stock, 2000 shares and 35% of the 3% yield, on trade a day would profit you $50,625 after commissions for the year. You can increase your earnings through higher dividend yield, increasing size on lower priced stocks, doing more than an average of 1 trade per day and adding additional strategies to your trading business.
Many traders try to predict direction, or pick that home run stock, or just chase stocks around all day, whipped around by the noise. The structured traders are looking for a defined strategy where they can clearly identify when it is working and when not, and using the noise and chatter of the market rather than being emotionally affected by it.
If you choose not to use this strategy, at least be aware that the good, growing, dividend paying stocks have an upward bias along with the market and if you short the stock and pay the dividend without having specific reasons and an overcoming edge from your strategy, you have are trading with a negative expectation over the large sample.
If you are concerned that in sharing this, it will reduce its potential…it might, but on the other hand there are plenty of skeptics and those that will not take the time to develop the strategy properly. Due to the efficiency of the strategy, sloppy trading, carelessness and mistakes must be eliminated.
This was from March
2009 Newsletter
Prior supporting
articles: www.pairtrader.com
, January 2008 Newsletter (download PDF from our website – free to all)
BTM archives: Presentations by Mitch Weiss and Rob Friesen
(need to be in BTM to access)
PairTraders in our trading circle have been actively collecting dividends since 2001. Back then we figured out that with the high correlation of pairs within the same industry group, the oscillation of supply and demand between them, the upside bias of a stock prior to the Ex-Dividend Date (X-Div), and the concept of premium of discount in putting spreads on, we were able to keep a substantial amount of the dividend being paid.
Where are we now in
this strategy with so many changes in the market?
It certainly is disappointing to see so many companies cut or eliminate their dividends. This has really cramped our income from this strategy when applied more in a mindless way of trading. It has forced us to combine more Art & Science and it brings out the need to trade ‘em requirement. Let me explain; as there are fewer dividends to pursue, we find that there is more competition to purchase the Dividend paying stock on the day before X-Div. This can cause this stock to be at a premium to its industry group and we do not like paying premium…we want to buy things at a discount (on sale). So, the need arises to move our time-frame forward. In the last 6 to 9 months, traders that collect dividends have been scalping the pair created for the purpose of collecting the dividend, as much as 5-days before X-Div. This enables the trader to have time to do his research, compile his correlated pairs, devise a plan for trading them, and go about waiting for a good entry point where he can put on the pair (with being long the dividend paying stock of course).
This front-end loading of the strategy allows for repeated opportunities to be in and out of the pair locking in a profit each time. Use the following schedule to adapt your expectations on scalping according to the upside bias experienced.
X-Div Date: Whatever the pair will give you. We do not marry the pair if trade is not working on this day.
Day before X-Div: 100% + We find the Dividend stock is often quite strong this day, outperforming its peer group and often can trend up. You will not get as many trades if this occurs so it is necessary to maximize your current trade. (Investigate PEP vs. KO performance on the day prior to X-Div, March 4, 2009) (PEP Div. was .42)
2-Days before: 80% + Usually the pair performs quite well and you may get multiple scalps.
3-Days before: 60% +
4-Days before: 40% +
5-Days before: 20% + This just gets you warmed up waiting for good entries and executing quicker scalps. This day helps you to get to know your pair.
Investigate TR vs. HSY in the 5-days prior to X-Div. this will give you some insight into the behavior of two stocks created for the purpose of capturing repetitive trades leading up to the X-Div date. (TR Div. was .60 March 6, 2009)
Remember that if you want to hold the pair overnight into X-Div date, you will receive the dividend on the payable date.
Here
is an idea…
by Rob Friesen
In light of the possibility (not the guarantee) of a bounce
back, or significant multi-day rally in the market, let’s see if we can be
prepared to capitalize on that by utilizing the dividend strategy combined with
relative strength and quality stocks.
Few Canadian companies have slashed dividends. Canadian banks have been listed as the
strongest and safest in the world and pay significant dividends. However, we
see that these stocks have sold off along with their
Quarterly Annual
Symbol Price Dividend Dividend
Yield (Forward)
BMO 21.49 .57 10.30%
BNS 20.35 .40 7.20%
CM 29.20 .72 8.50%
RY 22.99 .50 8.20%
TD 27.60 .50 6.90%
Seeing as we already
learned how to use Finviz as a great tool…click on
this link: (press CTRL key and then click
your mouse on link)
http://www.finviz.com/screener.ashx?v=341&ft=1&ta=1&p=d&r=1&t=BNS,CM,RY,TD,BMO
Once you are on the
site, scroll down and you will see charts and a great deal of financial
information, news and profiles for the stocks.
Let’s go through some more companies that are actually
raising dividends…what a novel idea!
While digging around to find these companies it I discovered that most
of them seem to be in consumer staples, utilities, health care and energy. These are things that everyone consumes
unlike the purchases from say Best Buy that are often financed. It is in those “borrowing to buy” sectors of
the market we see companies hit the hardest and then they eliminate or reduce
the dividend to stay afloat. Below is a
list of companies that have had a history of raising dividends. (Of course, that can always change).
Quarterly Annual
Symbol Price Dividend Dividend
Yield (Forward)
ABT 46.89 .40 3.40%
ADP 33.28 .33 3.90%
CB 35.73 .35 3.60%
CL 56.05 .44 2.90%
FPL 42.82 .47 4.20%
ITW 26.80 .31 4.50%
JCI 9.13 .13 4.60%
KMP 41.61 1.05 4.20%
KO 39.10 .41 4.00%
MMM 41.83 .51 4.50%
OTTR 16.77 .29 6.80%
PEP 47.10 .42 3.50%
PX 55.24 .40 2.80%
SHW 42.98 .35 3.10%
SYY 20.23 .24 4.50%
Once again…here is the Finviz
link for you to investigate these further:
In summary,
these stocks may provide opportunity for the dividend strategy as well as
perform nicely if the market rallies.
They may also be more resilient to the downside unless their dividends
get cut.