GDX Robot maturity

[mklein9]

Between the two robots, the IWM robot seems to be further along maturing as a system. It also seems to attract almost immediate attention when it behaves in an undesired way (as shown by the in-depth work in August, and the recent filter work by Pascal to resolve LT/ST edge conflicts). On the other hand, the GDX robot just concluded two rather poor trades combining for an approximately 18% unleveraged loss, and not as much attention appears to be given to researching why. Work on changing subscription systems, developing a real-time system, and simply running a single robot sounds like plenty on the team's plate.

This may be my impression looking from a distance, but it brings up the question: should dual-robot subscribers be putting much emphasis on the GDX robot at the moment? I know from my own experience that such complex systems take time and patience to mature. I am not complaining here (for sure!) but wouldn't mind a little comment or two about the relative maturity and confidence between the two robots.

Thanks for all you do, everyone. The IWM robot seems to be turning out just awesomely well. I hope the GDX robot is moving in the same direction as well.

-Mike

[Pascal]

Between the two robots, the IWM robot seems to be further along maturing as a system. It also seems to attract almost immediate attention when it behaves in an undesired way (as shown by the in-depth work in August, and the recent filter work by Pascal to resolve LT/ST edge conflicts). On the other hand, the GDX robot just concluded two rather poor trades combining for an approximately 18% unleveraged loss, and not as much attention appears to be given to researching why. Work on changing subscription systems, developing a real-time system, and simply running a single robot sounds like plenty on the team's plate.

This may be my impression looking from a distance, but it brings up the question: should dual-robot subscribers be putting much emphasis on the GDX robot at the moment? I know from my own experience that such complex systems take time and patience to mature. I am not complaining here (for sure!) but wouldn't mind a little comment or two about the relative maturity and confidence between the two robots.

Thanks for all you do, everyone. The IWM robot seems to be turning out just awesomely well. I hope the GDX robot is moving in the same direction as well.

-Mike

This is a good question Mike.

The issue with the GDX robot is that as you pointed out, we had two bad trades in a very volatile environment. Hence, an 18% loss. The problem is that I could not find an issue with the model. Maybe I did not search enough.

Some months ago, I compared the IWM Model as modeling the changing of the guard in front of Buckingham palace.
If one guard sneezes, you notice it, because he usually should not do that.

However, the GDX Robot model is like catching the moves of a teenager riding a wild horse in the forest. It is very difficult to detect "wrong" moves out of the general activity. GDX is linked to only a small number of stocks which can have a great influence on the ETF itself. For example, a few months ago, ABX overpaid for EQN.TO. ABX is the largest component of GDX. ABX lost 10.7% and GDX lost 4.7% during the two days that followed the announcement of end April 2011. This sort of cascading fall due to one component issue will never happen on IWM, because it is based on 2000 stocks.

The real advantage of GDX is that historically it has a low correlation to IWM. This is why I like trading this instrument. However, in times of high volatility, I drastically reduce position sizing and I NEVER use the leveraged ETFs on GDX.

Anyway, I will revisit some aspects of the GDX model, especially the opportunity to trade GDX when its correlation to IWM is high, which seems to have been the case in the recent weeks.


Pascal

[mklein9]

The real advantage of GDX is that historically it has a low correlation to IWM. This is why I like trading this instrument. However, in times of high volatility, I drastically reduce position sizing and I NEVER use the leveraged ETFs on GDX.

Anyway, I will revisit some aspects of the GDX model, especially the opportunity to trade GDX when its correlation to IWM is high, which seems to have been the case in the recent weeks.

Thank you for the insight, Pascal. I think the position size reduction under volatile conditions is one of the key observations. At the time of the last two GDX trades the volatility-based stop was 10-12%. Given that at least a partial role of GDX in the EV trading system is to hedge against IWM, a correlation measure against the broader market might be useful.

-Mike