Thanks for your reply. I must admit I am surprised. Without in any way intending this to sound critical, I would have thought that once you have sorted out a systems approach, you would just leave it run. I say this as someone who while accepting the validity of a systems approach is still grappling with the concept ala Mark Douglas that the path to success is simply to identify an edge and then trade in the full acceptance that you don’t know what is going to happen next and with no expectation that any particular trade will be successful, but that overall assuming you have a positive expectancy you will make money. My head believes this, my heart is struggling <g>.
Rereading my post, I'm struggling to see what might have surprised you. I believe I said that 95-99% of the days, my automated stuff runs without any intervention at all (in other words, my presence was not required, in hindsight). So, assuming ~97%, this would mean that my presence might be required for one day out of very 6 weeks or so. Noting that I make about 1500 order executions per day, and my presence may only be needed once every six weeks, it wouldn't appear that I am unnecessarily tampering with my system, would it? I do have full confidence in my systems, and virtually never do anything other than follow my systems (i.e. 99.999%+ of the time, systems are followed).
However, my presence is required to ensure that my systems can properly handle situations that have NOT been programmed. For example, assume that the system typically trades active intraday, as mind do, and further assume that all of your brokers order execution routes go down... Personally, I don't have that scenario programmed into the system currently. Or, possibly, one stock that you trade is halted, news pending. => In this case, I manually deactive my system from further trading of that stock for the rest of the day (I trade, not gamble => breaking news is not what that system was developed to exploit, and it would likely not work well). These type of situation happen infrequently, but they do happen. As a result, a human oversight presence is useful.
Another interesting thing is that thinking about my question afterwards (i.e. the question of why you would need to monitor your system intraday) it occurred to me that you might need to tweak parameters based on analysis from higher timeframes or based on opening activity – e.g. a higher bias for longs or shorts based on larger trends, that apparently you don’t see any need to do this. That is your systems work regardless of what the overall market trends are. That’s pretty amazing too. Do you have any comments on that.
I don't do any tweaking. I have my systems programmed to automatically tweak themselves, though. But, this is also automated.
-Eric |
