Perishabull Posted February 3, 2011 Author Report Share Posted February 3, 2011 This a a chart of the ratio of GDXJ to GDX covering the last year. I'm looking at this as it turns out it would have been more profitable (so far at least) to have acted on my GDX buy signal by trading GDXJ as a proxy for GDX. It's really a moot point whether I would have opted to trade GDXJ had I seen this at the time, I probably wouldn't have since on 27th Jan it would have looked as if the ratio had topped and was on it's way down. Anyway I digress. The potentially interesting aspect of this is that today I thought I would try out my indicator on a ratio...just to see what might happen. Here is the resulting signal chart for my DBDT indicator on the GDXJ/GDX ratio. DBDT is my main indicator, the one I have been trading and dissecting over the last weeks and months on this blog, as it's by far the most interesting and accurate so far. I cannot plot the indicator signal line with the value of the ratio as an overlay so I have marked double bottoms on my indicator working on the GDXJ/GDX ratio and the buy signals are on 25th August and 30th September 2010. Those are the red lines I have marked on the ratio chart above (the first chart). So these were clearly very good entry points to buy this ratio. This is possibly another interesting dimension of my indicator I have discovered today but really it's not possible to make a valid assessment on the basis of a solitary chart. I'll need to do a full study and I already have a backlog of different studies that I am going to complete, time just seems to evaporate these days. I've never been so completely absorbed in my life as I am now with this. I worked out some new numbers on the performance of my indicator over the last year on the top 100 stocks in the S&P 500 and the results are slightly dissapointing in one sense but enlightening in another. This is an assessment of all the buy signals as a group and all the sell signals as a group. On the one hand you have an 81% for buy signals that were correct, on the flip side it's rather obvious the hit rate for the sell signals tells me my system is mostly a long only system when it comes to index based stocks (S&P 100). 54% sell signals being correct is marginally better than a coin toss. My intention was to develop a system that could work both ways. Clearly it is only accurate in one direction when it comes to stocks that form an index (S&P 100). I think that for commodity correlated stocks this doesn't apply since the work I have done on the gold complex stocks and oil based ones seems to be very good on both the long and the short side. I'm certainly not in any way unsatisfied with the accuracy of the buy signals. Another number I worked out was that 28.5% of the buy signals were at or one day following an intermediate low that developed into an uptrend. The real test of this indicator / system will be over a period of time. Link to comment Share on other sites More sharing options...
This topic is now archived and is closed to further replies.