VERY disappointing day. 5 trades for -$344..... 2 hours before market closed the bot was up only $30... and then proceeded to get whipped around on several trades to close out the day... Sad part is, after realizing the other day that tues and wends are the BEST days for the method I nearly didn't turn it on today..... Hindsight is 20/20 I guess... I am still undecided about if I will turn it on tomorrow or not yet, since friday is the WORST day for the system. Days like this make me want to try and tweak the system so much, but I must resist this earge. Overall nothing wrong with the day, everything historical limits.... just the way it goes, can't make money EVERYDAY...
ER2 had some great swings today. I always find the channel days easier to trade than the trend days. Something to think about: Was your negative return today due to it being a Thursday or because today was a channel day vs trend day? If your bot is looking to catch a trend and you don't have a bias filter, you may want to look into that. That should keep you out of some trouble spots. You said your system is a SAR (stop and reverse). On a channel day like today; if your rules put you on the wrong side only once, you are going to be fighting the market all day.
I am afraid to ask this. were the trades at a loss straight from open? or was the market moving up and down too quickly?
More than likely today's results was due to it being a range day. The strategy is basically a trend following strategy, so days like this aren't ideal. However it does have the possibility to profit on days such as this one. I have a few things in place to try and keep it out of the market in these situations but today it just failed..... Nothing for me to get TOO upset about, but would be nice if I can find a way to avoid days like this even more. Identifying the type of day it is, is more than half the battle..... Its just not very easy to identify whether the day will be range or trendy... I am open to any suggestions someone may have of ways to stay out of the market on these types of days
walter, No, not all of the loosing trades were a loss imediatly following the open of the trade... several of the trades had times where it went rather positive, just did not get any exit signals to take the profit out of the trade.. However, a few of the trades were losses imediately
Why not create a database of economic events and have the bot go flat & offline 10 min prior and 30 mins after a major event such as an Fed minutes. I don't have the data but I would bet there is 100% chance of a choppy market immediately following fed minutes.
It would be interesting to see if fourier analysis can tell the difference between range days and trend days. Id imagine range days would have a mix of frequencies, grouping around a relatively highfrequency or maybe fairly random while a trend day would probably have mostly extremely low frequencies .
I will look into this "fourier analysis", If I can somehow find a way with any degree of certainty to predict the type of day it probably would greatly increase my profitability.
I am looking for an opion here. Lets say I have the method I trade now, it produces X amount of money over a period of time. Now lets say you have method B, this method is a more restrictive version of the origional. However this method produces 30% less total profit than the origional does. However it has a few perks over the origional method: #1 makes money a higher percent of the days. #2 The equity curve is a bit smoother. #3 Expectancy is more than double that of the origional method per trade. #4 Max drawdown is around 1/2 that of the origional method. This method only makes about 1/3 the number of trades the origional method made. Would you sacrifice the 30% extra profit in this situation?