By Lipa Roitman Ph.D.
September 14 2012
) is now the most valuable company of all time. Rumors of new iPhone and the release of the iPhone 5 have been running rampant of late, helping to give shares their latest boost. Being in a high-tech field, AAPL stock has a different dynamics than the rest of the market. Thus playing AAPL against S&P 500 is an interesting proposition. Both AAPL and SPY
"Spider" have large market capitalization and are considered relatively ‘safe’ investments. Playing one against the other would be like speculating on the high-tech VS. the
whole market, or NASDAQ VS. NYSE. Which one will outperform the other?
Let’s see if the dramatic rise of AAPL against the S&P 500 over the last year (see FIG.1) could have been predicted.
The IKNOWFIRST stock market forecast algorithm provides daily forecasts for stocks, indexes, currencies and commodities, among them Apple stock forecast
and the S&P 500 index forecast. For each market (For example the AAPL) the following data is calculated daily by the self learning algorithm:
1. Signal – The movement direction (increase/decrease). Six such signals are given for each market for six time horizons, from three days ahead up to a year ahead. The signal strength is the absolute value of the current prediction of the system. The signal can have a positive (predicted increase), or negative (predicted decline) sign. The signal tells how much the last price deviates from what the algorithm thinks is the "fair" or "equilibrium" price.
– The "strength" of the prediction, which indicates how confident the system is about the prediction. Predictability is calculated by measuring the past performance of the system for that market and for that time horizon. Theoretically it could range from minus 1 (actual move opposite to predicted) to plus 1 (perfect prediction). For the 150+ markets in the system the predictability generally ranges from 0.3 to 0.7, depending on the market and time horizon.
We have analyzed the performance of the system in order to develop a profitable low risk strategy. In this article we report the results of the performance of IKNOWFIRST algorithms for predicting the AAPL and S&P500 index. We have also compared the signals of these two markets in order to see whether the algorithms were sensitive enough to correctly predict the AAPL / S&P 500 ratio.
Figs 2 and 3 show the AAPL / S&P 500 ratio in red thick line. The blue signal line shows the algorithm prediction (signal) for that spread, Fig. 2 shows the 30 days horizon signal, Fig. 3, the 90 days signal. Each point in these lines was taken from the forecast produced daily by the system over the last year. To make this chart the AAPL signal was divided by the S&P500 signal. The resulted AAPL / S&P500 daily signal was added to the corresponding AAPL / S&P 500 daily ratio. Thus if the signal blue line is above the red actual line, that means a BUY signal, if below, SELL signal.
One can see from Figs 2 and 3 that from the beginning of January to the end of March the signal was BUY, and if one were to follow up on that signal and be long AAPL and short S&P 500, he would gain up to 45 percent, depending on what stage he closes the position. This is one example of "market neutral" strategies, which means being simultaneously long one equity and short another. This strategy makes the position less sensitive to general market fluctuations.
As a side remark, Figs. 2 and 3. show that the system have correctly predicted the positive market reaction to the new iphone3 model.
Fig. 4 shows the average predictability for the 30 and 90 day forecasts. It was generally positive, with the 90 days predictability higher than the 30 days. That is in line of what we have generally seen in most markets in the system, the longer term forecasts have higher predictability. We think that is because the long term forecasts are less affected by the daily market noise.
Discussion and Conclusions
We showed here that apple stock latest rise against the S&P 500 index was predictable.
Along the way we have come upon a possible market strategy that utilizes AAPL and S&P 500 forecasts to create a spread that is not dependent on the general market trend, but only on the high-tech VS. the general market play, i.e. a kind of "market neutral" strategy. It could be possible to apply this to other market pairs,
There are now two ways to trade the market using the algorithm. One is to simply trade long or short the individual equities. We propose today the second method, to trade spreads of equities such as AAPL-S&P 500, NASDAQ-S&P 500, DJI-S&P 500 and Russell 2000-S&P 500. All trading is risky, but this lower risk strategy has the advantage of being somewhat insulated from the general market ups and downs. Again, one needs to use judgment when trading. Going long on any equity after a long uptrend is risky. There are always safer opportunities.
Apple stock forecast for today: to buy or to sell?