I have been waiting for market halftime (June) to pen this article. I have a very simplistic algorithm (algo) that I call "buy on close" aka The Poorman's Algo. I use the word "poorman" as I do not employ anyone who holds a Ph.D., nor is it cloaked in a black box co-located at the NYSE. It can be deployed by anyone, from beginner to professional. I will get more into this later.
History Creates the Future
In 2007, I noted trading patterns in Apple (NASDAQ:AAPL) which were centered around options expiration. Options expiry used to be a monthly event until the advent of weekly options in July 2010. The end of July marked the beginning of The Poorman's Algo. It was in zygote phase at this time, but quickly developed. By the time 2010 came to a close, it was battle tested and ready for war. I wanted to prove to myself, with a little hard work, billion dollar firms can be outsmarted by a little ole amateur investor out in California.
Apple Influenced By Options
For the scope of this article I will bring up only 1 aspect of options trading -- buying calls. Apple is always the #1 leader in single stock option volume. Therefore, if any stock was influenced by options, it would be Apple. When you buy a weekly call that expires in 5 days, more than likely another retail investor did not sell it to you. The person or firm that sold you this call may have done so because they were legally obligated to. Since they never wanted the risk in the first place, they will offset the sale of a call with the purchase of Apple shares. This is what you would call delta neutral. Apple always trades more calls than puts in any type of market. If you have an equal number of buyers and sellers for Apple stock, i.e. 500K sell and 500K buy orders, the stock would not move. Let me now bring in the options side. Since Apple always trades more calls than puts, the stock will have to rise as buying bias comes in. Remember from earlier, when you buy a call, the seller will buy stock to offset it. This is what brings in the buying pressure. The opposite also has to take place later in the week as you will eventually sell this call. When you sell your call, the buyer of it will sell shares of Apple causing sell bias as long as the stock side is in equilibrium, i.e. 500K sell and 500K buy orders. In conclusion; people buy weekly calls in the beginning of the week and sell them at the end of the week.
Actual Results Forget to Lie
To fact check the above paragraph I have compared Apple to Priceline (NASDAQ:PCLN). I chose Priceline for its similarities in the following categories: stock price, 2011 gains, liquidity and earnings release dates. I also chose Priceline because it has a polar opposite category -- options. Apple may trade 500K options per day vs. Priceline's 20K; this will become a very important difference. In 2011 Apple gained $79.36 or 24.4%, where Priceline gained $65.46 or 16.3%. So let's see if options influence stock. Below is a graph of Apple on a Friday close through Tuesday close and Wednesday close through Friday close basis for 2011. I performed the same exact measurement for Priceline.
As you can see, on a Friday close through Tuesday close basis, Apple gained $168.52 or 51.7% vs. Priceline's $87.21 or 21.7% gain. This is a major difference but not mind-blowing. If you look on a Wednesday close through Friday close basis, things get out of control. Apple lost ($48.36) or 14.85%, while Priceline gained $62.12 or 15.4%. Why the major reversal between Priceline and Apple from Wednesday through Friday close? It's because of the options. Apple has a massive number of contracts trading, where Priceline has very few, not enough to influence the stock anyway.
The Fall Brings A Rise
I've mentioned before, buying calls at the beginning of the week causes buying pressure and selling them towards the end causes selling pressure. The algo I'll talk about in a minute was actually conceived in the reverse order. In 2007, I knew expiration Fridays brought selling pressure that wasn't relieved until Monday - Tuesday. This is how it began.
The Poorman's Algo
Now that we have discovered a tradable pattern, let's exploit it. The instructions are simple; you want to "buy on close" every Friday (or Thursday if Friday closed). You can enter various types of buy on close orders. The goal is to buy the closing print or as close to it as possible. The exit for this trade is a "sell on close" order every Tuesday. This is the first half of the algo or the long only side. A benefit of the long only is less holding risk. It's taking on 4 days of risk vs. 7 if you were to buy and hold all year. This means we are removing 156 days of risk per year. We already know the performance was much better than holding the stock all year in 2011, as it gained $168.52 vs. the stock's $79.36. Half the holding risk yet doubles the gain. Below is an image of 2012 year-to-date performance.
So far so good, on half the holding risk too, $140.41 gain vs. the stocks $151.59.
The second half of the algo is to exploit the fall from options closing and related reasons. The performance in 2011 was an additional $48.36 gain. The instructions are simple; you want to "sell on close" every Wednesday. The exit for this trade is "buy on close" every Friday (or Thursday if Friday closed). Below is the same image as above, only I introduced the short side or second half.
Year-to-date the stock has lost $98.53 points on a Wednesday through Friday close basis. Since the second half of the algo is short, this is a $98.53 gain. The image below I combined long and short side to create the full algo.
As you can see, this algo has dominated the stock, giving us alpha on Apple. In 2011, the full gain was $216.88 or 66% vs. the stock's 24.4%. Year-to-date (06/01/12) the full algo has gained $238.94 or 58.4% vs. the stock's $151.59 or 37%. In the above image you can see I have a line "Long Bias Algo". Because Apple is severely undervalued, I don't like the full short position on, but rather cut it in half. In other words, it goes long 100 shares and short 50 to give it long bias. From the results you can see it's best to go 1x1, but at least you know you can alter the lean (bias). Below is a year-to-date visual of Apple stock.
You cannot decide to trade this some weeks and not others or your results will severely be altered. It is designed to be traded as if you were a robot, there is no thinking. You have set instruction that should be followed at all times. It is imperative you trade the same size at all times. If you decide on 10 shares, then 10 shares it is for the year. You cannot trade 50 shares one week and 200 the next. You can write a code for your platform or manually click your mouse a few times per week.
What If This Stops Working?
The options market is just one of the many reasons Apple is undervalued. It is my goal to expose all the trends I know and see in hopes of them going away. My true goal is for Apple to trade at fair value and unabated. If these artificial waves are going to stay in Apple, I want everyone to make money off of them. They will either become so big that the whole world will start riding them, or they will go away. But, until then, surf's up dude.