The lure of dividends is always there. Even the novice stock trader notices that certain stocks move down suddenly on their ex-dividend date, and realizes that holders of that stock just received a tidy cash payment. It is no surprise that every trader, perhaps tired of fighting win-one lose-one battles in the volatile high-beta stocks, is eventually tempted to try the slower moving dividend stocks with their predictable payments. But to the trader, a strategy of simply buying and holding is not interesting enough. The trader wants to trade.
Last time in Part 1 I wrote about the general behavior of dividend stocks, and looked at the strategy of simple buy-and-hold. Look at what happens to a company which pays a 10% dividend like clockwork -- and has no fluctuation in stock price due to such distractions as investors buying or selling their stock.
Stockholders of the imaginary IDC get a 10% per year income. They get it regardless of when they buy and sell. If they buy before the ex-date and the stock price falls, it is exactly balanced by payment of the dividend. On the other hand, there is no way to shortcut the process - you cannot get more price appreciation or dividend per day's hold time by buying on certain days and selling on others. The futility of choosing which day to buy or sell makes this scenario uninteresting to the stock trader. Traders choose times and prices, and all choices give the same yield in this case. Very boring!
Real dividend stocks do not work like this. The existence of reliable cash payments to stockholders causes a distortion to the ideal price behavior which is generally, within some bounds, predictable. Nothing is more exciting to a stock trader than spotting something predictable in the behavior of a stock's price. Take a look at a more typical [but imaginary] dividend company, RWDC.
Imaginary company RWDC sees its price fall by the amount of the dividend on the ex-date, but the price movement is exaggerated, frequently with overshooting movements to the downside and the upside both. What is most often evident is a rapid price recovery following the ex-date to the pre-dividend price. The mechanism at work is one of supply and demand and is easily understood.
For instance, anyone holding RWDC for a long period of time will tend to not want to miss out on the next dividend before selling, if selling is planned. This is one of several psychological explanations. The sudden drop in price post-dividend is exacerbated by the lack of demand for the shares -- after all, it is a long wait to the next dividend date, and few buyers formulate new plans to buy a dividend stock immediately following the ex-date. The sudden low price attracts the attention of buyers looking for a bargain, despite the fact that the true drop in stock net price is low. Once the downward price movement slows, the rising price creates price momentum and attracts buyers who may have no intention of holding the stock to the next dividend, especially daytraders. Other phenomena are also at work, some of which I'll mention in a later article. The only thing important to notice is that a typical dividend stock often recovers its pre-dividend stock price before the next dividend ex-date.
This leads to the strategy of Dividend Capture. In this strategy the dividend trader seeks to obtain the full dividend payment while minimizing the time during which the stock must be owned. The dividend stock, like any stock, has a risk of ownership, and the fewer days the stock is owned, the lower the risk. The least time is one night -- that is, the stock is bought in the closing minutes of the trading day before the ex-dividend date (the mo-date, or "must-own date"), then sold at the first profitable opportunity, perhaps even early the next morning. Owning the stock on the midnight preceding the ex-date is sufficient to be a shareholder "of record" and thus receive the dividend. The trader attempting dividend capture will typically hold the stock until the purchase price of the stock is regained -- selling will then give the entire dividend as profit. If the hold time is small, there will be time to take the money and put it into a 2nd stock before needing it again for another capture in the first stock. In this way the same capital can be used to collect dividends from more than one stock at a time, possibly increasing the total gains above that of simple buy-and-hold dividend investing. What a great idea! But does it work?
Here is an actual example of a dividend capture. On market opening of June 11, you purchase EOD at $9.21. June 11 is the mo-date, and ownership of the stock is required on midnight that night. The next morning, you put a good-til-canceled order out to sell EOD at $9.21. The stock opens at about $8.90. Because of the dividend, the market makers tried to open it at 28 cents under the previous day's price (which price -- open, average, or close -- is a mystery to me), which would have put it at $8.93, so it was very close.
During the day the stock spikes up and down, and someone buys a large block and the price spikes up to about $9.25, taking out your GTC order. You now have all your cash back, and your 28 cent dividend is on its way to you. This is the best result you can hope for -- in most cases you will not be able to sell like this, on the ex-date for the same price as the previous day, but it does happen. It happens surprisingly often, too - my informal surveys indicate perhaps 15-20% of the time you can do a 1-day turn of this kind.
Now take the money and find a new target. The next available home for your cash is Apollo Investment (NASDAQ:AINV) the next day. Buy it on June 15 for 10.57 and put out a GTC order the next morning, as you did with EOD. The next day you have qualified for the 28 cent dividend, but the stock walks away from your price. You hold the AINV, and your order is not taken until a month later on July 27. Once again, you've gotten all your money back and you have another dividend too. Congratulations, you captured two dividends in less than 90 days using the same cash! This kind of efficient use of capital is the object of dividend capture.
To see how much one can make by dividend capture, I set up a paper trade simulation covering late May 2009 to about 90 days later in August, utilizing the following high yield stocks: Encore Energy (NYSE:ENP), BOE, Walter Investment (NYSE:WAC), Blackrock Kelso (NASDAQ:BKCC), Whiting USA (NYSE:WHX), ETY, EXG, BDJ, BGY, China Infrastructure (NYSE:CII), EOD, AINV, Solar Capital (NASDAQ:SLRC), ETV, ETW, NY Mortgage Trust (NASDAQ:NYMT), Kohlberg Capital (NASDAQ:KCAP), ANH, MFA Mortgage Investments (NYSE:MFA), ETB, and ETJ.
I divided my cash into 15 bundles of $10000 each for a total of $150,000. When one of the stocks was about to post a dividend, I would buy the stock at the opening price on the mo-date, hold it overnight to qualify for the dividend, then sell it as soon as the stock returned to the price I'd paid for it. If the stock did not recover in price by the next mo-date 90 days later, I would give up and sell the stock for whatever I could get, at the opening price on the mo-date.
Did I make money? Yes, it was profitable, but not impressively so. Of the 21 stocks, Five failed to recover their purchase price and I had to sell out after holding the stock 90 days. Four recovered immediately, right on their ex-date, and no hold time was needed. The rest were in between. Six of the bundles of cash were able to collect more than one dividend in the quarter; the other 9 got one each.
Bottom line: the original $150,000 was now $152,219, for a gain of about 1.5%. Not at all impressive, considering that if I had simply divided the cash between the 21 stocks and bought and held each for 90 days, I would have gotten 5.3% gain -- about three times as much. I repeated the experiment with a different group of about 45 stocks yielding about 6% and got similar results, this time showing slightly less profit than the high yielders.
Dividend capture as a strategy can work, and it does work, but it can't be recommended for routine use because it simply doesn't make enough money. There are two main reasons for this: (1) In order to get out of the position, you sell as soon as you get your price. But, stocks which recover quickly may be making a strong up move, and selling immediately causes you to miss the gains you would have gotten. (2) Waiting for long periods of time to get back your purchase price tends to tie up your money in negative balances waiting to earn a relatively small dividend. (3) The worst case loss is large, the best case gain is small. You are betting against a disaster, playing the short odds, and one disaster can wipe out the gains from many, many dividends.
You can modify the dividend capture strategy to improve its returns. Some possible mods are obvious - for instance, never selling on the ex-date even if the price recovers, because the stock is likely to be headed much higher. The thing about modifying strategies, though, is that one never stops modifying. A very highly modified dividend capture strategy is no longer dividend capture, it is outright free form stock trading of a stock which happens to have a dividend. That's not the same at all.
There are a lot of ETF's which hold dividend stocks and which pass on the dividends to you each quarter (after taking their fee). Some popular ones are AOD, VIG, and BDV. You can bet that the managers of these funds do a bit of dividend capture, sometimes. How much, is hard to guess -- this is not information they will want to reveal in the prospectus. I'm sure that when they do use dividend capture, it is opportunistic in nature and occurs when the market is favorable.
Next time in part 3 of this series, I'd like to write about the third strategy for the aggressive dividend trader: Reverse Dividend Capture. Until then, the best of luck to all of you.
Disclosure: Long BKCC, BDJ, AINV, ANH