Lost $349 today..... another bad day..... I think the method is due for a good day soon... maybe it will be tomorrow....
Not looking good again, currently down $120... my bot is having a misserable week... I knew this week was going to be slow, I should have went with my first insticts of not running it this week... Everything is still within normal boundaries from the back test... its just when its loosing money like it has the past several days it really makes me want to tinker with it or pull the plug.... hopefully things pick up soon before I loose too much confidence in the program...
frost, I just went live with an automated system, too. I started yesterday even though it was Thanksgiving week. I thought about waiting until next week, but I looked back at my historical data and there was no major difference in performance on days before holidays. How far back is your system tested? Can you just look back and see if taking off days before holiday weekends is the right thing to do? If it is tested back less than 6 months then I would be careful. I wrote about 5 systems based on what I saw by looking at charts for the past couple of months. All of them did well for the past 3 months to a year. But all but one performed terribly prior to that. Only one of my systems made money consistently for the past four years with manageable drawdowns. That is the one I am trying live now. So, what I am saying is be careful if it is not tested back very far, but just let it do its thing if it is. Cpt Power
cpt power, I have data for about 3 years. The system has remained profitable throughout all 3 of those years, although there are 2 quarters where it was negative....... the system has performed MUCH better over the past 6 months than it had been performing in the years before that..... but has remained profitable the whole time..... However, I have not done any tests for holidays.... System currently down -$250 as we speak...... hopefully it catches a good trade or two at the end of the day.. then I will shut it down until next monday
frost, Is your system one that you expect to make money almost everyday or one that nets out positive over a few weeks or so? Mine is more like the latter. It loses lots of days, but comes out profitable over time. For this reason, I will take every trade and not make any decisions overriding the system. My system is market neutral and there is no telling when it might win or lose. Many times from my backtesting data has a drawdown been erased and profits piled on out of nowhere. I'm just saying this because it seems like you are tempted to interfere a lot. If you are confident in the programming and backtesting that you have done then you need to just let it be. I'm just going to let my system do its thing unless a larger than historical drawdown is seen. Mine so far (net amounts): Tue -190 Wed +80
Cpt power, Yea my system is one that is net positive over time not day to day.... It is positive somewhere around 60% of the days..... Everything that has been happening is perfectly normal and is rather MILD compared to some of the times it went through even in just the past 2 or 3 months... its just when my real money is on the line it makes me very anxious....
System has closed out its last trade of the day and is now officially done for this week. Here are the totals thus far; Day 1: +$1,450 Day2: +$40 Day3: -$344 Day4: +95 Total: +$1,241 at end of Week #1 Day 5: -$35 Day 6: -$349 Day 7: -$229 Total: -$613 at end of week #2 Total thus far: +$628 after 7 trading days System is performing below historically daily average at this point... expecting next week to get things back on track.
Maybe I'm not a very good system developer, but I wasn't really under the impression that creating a system that is substantially profitable every trading day is possible. If you've managed to build a system that is profitable on a weekly basis and manages to take only a few trades per day, I'd say you've pretty much found the holy grail... RoughTrader
Was Day 1 an aberration or a normal expected day every few days. If Day 1 was an outlier or rare occurrence then your system has a very bad losing record. Try not to let one good day blind the results. Study what happend on Day 1 to make sure it is something that can be repeated and not just on strong huge trending days since those are more rare then the sideways chop we have had which led to a lot of those losses. Taking the time now to study this will ensure you are net positive and not just covered by one unusually good day. Good Luck!
optioncoach, Day 1 was somewhat rare due to the fact that it made 1450... but days where it makes 800+ are fairly common and happen once every week or so... It is those days that make it net positive over the long run EDIT: Also, today I have been backtesting with a new filter. It basically looks at the range ER2 has been moving in the previous several bars before the entry signal was given. Thus far it is showing a lot of promise......