Apple/Google: Is the Pair Ready for a Contrarian Strategy?

The action in Apple AAPL over the past few weeks has been charged with energy as the stock has fallen off the tree for the momentum players. After hitting $704 last September, shares are trading near $423 Monday. That’s roughly 40% from the highs.

One way to take advantage of Apple's demise is to trade the Apple/Google GOOG pair as a spread. This “pairs” strategy would mean a long side entry in Apple and short side entry in Google as a dollar neutral position. This technique has been around Wall Street for over 20 years and is a form of statistical arbitrage. There’s no real “market” hedge. It is really dollar neutral in terms of the number of Apple shares long compared to Google shares short. The trade takes advantage of an anomaly in the spread between Apple and Google as the price differences between these two highly correlated stocks diverge to the point that the spread widens to levels not seen historically. But when I say historically, it’s not just time that is important, but the length of time two “similar” stocks have been under a high correlation. In other words, Apple and Google would move in the same direction at least 86% of the time for a very long time. Then it changes. We want to take advantage of this change in correlation. We can measure this change as a chart of the spread then overlay a simple linear plot.

Below is a weekly chart of AAPL. Notice the distance from the 50 period MA and price range. Roughly an 18% move from the 50 MA.

Below is the weekly chart of GOOG and the relationship of the price to the 50 period MA.

Below is the weekly chart with both AAPL and GOOG's divergence in correlation. We can see the time line I plotted on the chart when AAPL and GOOG divorced and went their separate ways. The stocks accelerated away from each other as GOOG trends higher and AAPL trends lower to historically wide levels.

The Trade defined as a spread between the “pair”.

Since we are setting up the trade based on the spread being wider than historic levels (Selling short the winner and buying the loser), we only care about the movement of the spread. In this model, we are looking for the spread to narrow. If history repeats itself, the prices will converge and the spread will narrow and we make money. We do not care which end of the trade creates the narrowing of the spread. We only care about the spread. How does this happen? Well, there are several scenarios and I’ll go through a couple examples leaving out any dividends or splits that would have a direct impact on our profit/loss. Scenario 1) is Google falls at a faster pace that AAPL falls. 2) AAPL rises at a faster pace than GOOG or 3) one stays in a tight range as the other increases in volatility. These scenarios might bring that spread tighter. We won’t try to guess what will happen or when, we just want a few scenarios too potentially work in the favor of the spread to narrow based on historical prices between the two stocks. Remember, this is a contrarian strategy and we are betting the ratio moves back to the mean.

We would want to keep the same dollar amount. Using the weekly close data of AAPL and GOOG, our set up shows a ratio of .53. Using AAPL as a base long of 100 shares makes this really easy to comprehend. For example, we might want to buy 100 of AAPL and short the same dollar amount of GOOG. So to find this, we plot the spread and see that 100 AAPL equals the same dollar amount of a short position of 53 shares of GOOG.

Sheet of a very basic ratio plot to see how it looks over a few week periods. Look at the last line item taht represents the current spread data point we would use if we wanted to put this trade on now.

Below is a chart of the spread: I plotted a simple linear regression over the spread and we can see the spread is at the lower band of the linear regression line. We want the spread to eventually migrate back to the mean.

What’s the risk?

There is risk in several areas. The spread gets narrow to a point of margin pressures and time is a risk. Remember, this is technically a contrarian strategy. There’s potential bankruptcy. If you trade pairs in other companies that might not be so fundamentally sound. (Very doubtful this occurs in AAPL or GOOG). Also, execution costs, margin costs and changes. Illiquidity or carry costs associated with short selling will alter the performance. All these factors can have an impact on your net profit/loss that’s not adjusted in the spread.

Advanced models:

I can take this strategy a few steps further by altering the ratio. The ratio can be altered through factors in a proprietary model that many professionals have in their statistical arbitrage models. For example, a weighting of volatility, dividends, liquidity or volume. But this example is the as basic as you can possible get and with the right risk and discipline, can work out if you want to take advantage of a contrarian bet in AAPL or GOOG. Be aware there are several other methods to plotting the spreads and ratios between two pairs which can alter your trading time frame dramatically. Take the time to do the research that fits your risk.