Quote from Jordan:
I'll wait until the gap stops filling before I re-evaluate and tweak. The old patterns like the 30 minute break which has worked for years or the older patterns like Edwards and Magee that continue to work? Or even Fibonacci from the 13th century is it... that still works too despite a massive increase in volume?
I didn't say that all working methods/patterns will be rendered useless as a result of increased volume on 24-hour trading, only that you need to
be aware of the potential impact of this gradual, but
fundamental change to the market's opening behavior.
Some methods should very well continue to work regardless of this change, while other methods may not continue to perform. For example, Tony Crabel, often revered by many traders as a role model for going from trading book author using objective systems analysis to $2B fund manager with no losing years, and as the king of opening range trading, is now taking a beating. Perhaps the combination of increased volume on overnight sessions has fundamentally changed the impact of the opening range/gap analysis. Perhaps that combined with the different character of the current bear market.
Who knows the reason. What I do know is Crabel uses very strict pattern based system with extensive backtesting showing statistical significance. But backtesting
any methodology, regardless of statistical significance (i.e. confidence intervals), only works so long as the underlying dynamics that you are modeling do not change beyond the ability of the system/method to capture that trading edge. This is why so many system traders and other prediction models such as NN's fail. Backtesting alone in-and-of-itself, even if done with proper discipline for robustness and significance, are always at risk of changing dynamics. You have to find ways to identify enough of the dynamics a priori and build tests to validate the behavior of these dynamics separate from your system. In case of dealing with only observed data (e.g. NN's) there are still ways to identify the effects of shifting dynamics because you at least know what the inputs are, even though you do not know how the inputs are utilized. Some of this is addressed with chaos/complexity theory. The vast majority of the work in this area is all proprietary and I have seen almost no public information on this topic.
As far as Edwards & Magee and Fib are concerned, I am specifically talking about the
opening range which is what inandlong specifically discussed. As you know, there are many strategies based solely on the behavior of the opening range. Do a search on all the opening orders trading on ET, for example. All of Crabel's revered work (and his out-of-print book selling for $1500) deals specifically with ORB-related patterns.
Now what happens to all of those specific strategies if the fundamental nature of the opening changes? These strategies will have to be tweaked. Like I said, this change will be gradual.