SPY (SPDR S&P 500 Stock Index ETF):
As promised I took at look at how the neural nets would consider the stength of the $SPX rally using the Spider ETF $SPY.
Here are the basic statistics. The thing you should note is that the model, which looks for Bullish Gartley like patterns on a daily basis and allows the neural nets to optimize for small deviations in price pattern symmetry, show that the gross profit ratio (or profit factor) was at the minimum acceptable barrier around 1.6/1 and the win percentage was above the 60% minimum I would accept (at 67.8%). 10 years of market data from 2003 to 2013 were used with close attention paid to “price low to price low” cycles and 2 years of real-time walk forward data was used to test the model so that it would trade on data the model had not seen before. This is A LONG ONLY MODEL, so the model will not snap back or stop and reverse under any set of rules based instruction.
Results of the model trained by neural nets:
A look at the prediction chart for SPY daily data shows that IT TOO (like many traders) tried to by the dip in mid-August and would have gotten flushed. (I would have traded a stop underneath it if I had actually traded it, and I can explain how that works should signals be generated on a daily basis again). One notable item is that this model in its current configuration still wants to be long on Monday August 31, 2015. Though I typically will not buy any non-momentum low reversal in this model, it does indicate based on the buy that it could indeed rally to the resistance points I mentioned in the previous post (the equivalent of 202 to 204 on $SPY. For that reason, I would not necessarily go long this model but rather wait until we see a momentum reversal at resistance. If it fails there, then odds increase dramatically for a reversal to the downside that I mentioned in the previous post. AGAIN, no guarantees that this will happen, but all I can do is interpret what I have seen from these models over the last decade or so. That seems to be what is in store next. We will very soon see.
I will post this daily chart but wait until I see a test of price resistance before I comment further. Beacuse of a difficulty with a data vendor I have used for years, I could not apply the same amont of analysis on WTI Crude Oil Futures as I could with SPY, but I can at least make a short comment. Last week, deep short covering became the order of the daily sessions, and it quickly rallied above the previous support I used to measure the 33.78 target low. Before establishing any new target, I want to see where the reversal point lies. I do believe that inventories are still excessive on any rational basis for at least another 9 months to a year. What the great question will be is the desperation for cash that OPEC, Saudi Arabia and other Middle Eastern and African countries to generate profits. That will drive the continued stockpiling of oil, or it could lead to balance in pricing and inventory reduction. World leaders are generally a bit nuts, so all we can do for now is watch for key support and resistance in the charts.
The chart basically shows the daily cascade of the 38.2% bullish counter-trend swings in the weekly CL #F Charts. The chart time line is obscured by both 1) my mistake of sliding the chart editor underneath the toolbar. I cannot repeat the chart as present, but when I take one computer off the data crunching for forex, I will the time coordinates to the chart. A closer look at the daily chart shows the clusters of price action that would act as resistance to any rally. If the rally fails around the 38.2% retracement at 49.00, then I think the next move favors a correction close to the one I had estimated in the previous post.
The point is, we seem destined once again to be rendezvousing with that same trend line is a few days. I will reset the new target lows at that time. I still think the ultimate destination could be in the low $30s/bbl.
That is it for now. As always, that you for supporting this blog!