Auto trading newbie

Quote from WyckoffTrader:

Canned programs are usually as static as you can get and when the environment or volatility change than you are going to see a different sloping curve than you see now.

I understand you are starting out but this business is all about the hours and mistake you make and if you have capital after that than you are better than 80% of those that try ATS.....good luck and look into trading a portfolio of systems to cut down risk and also learn about correlations of your systems against each other and you are ahead of the game at that point.
More...

I agree, got rid of the pre-canned strategy over the weekend and now running on my own approach.

I have been trading for about 10 years, started out with about $700 and over the years with some growth and lots of paychecks deposit I am trading with about $34k now (was $28k 12 days ago had some good trades since then). But this is my first adventure down a purely technical trading path and the first time with ATS. I have lost all my money I 2 or 3 times, it has been a while and learned from each time. I hope now with that experience behind me that the ride with ATS wont be as bumpy as my first year trading.

I am looking to try and get once strategy built and running at a time. So far one down, I will work on the next one soon to build up a portfolio of systems. As for the correlations between systems that will be something I will have to work on after I get a few strategies working for me.
 
Quote from Rabbitone:

Back testing is only part of the answer. I perform extensive forward testing to eliminate curve fitting.

If forward testing is not available to you than long periods of paper trading may give you an idea of how your strategy is performing in various market conditions. In this paper trading period the strategy must maintain a reasonable expectancy through out.
More...

What, you mean like monte carlo simulations with shifty volatilities and such?
 
Well, i think i'm running into the case that the computation of values for bars are different between backtesting and live testing.

When backtesting, the test is only run every 10 minutes (assuming 10 minute bars), and each bar is 10 minutes long. But it seems while live testing, the current bar is anywhere from 0-10 minutes long and it throws things off.

I could be wrong on that as i am still testing, but everytime i try to do a live test, and then for comparison, back test the day after closing, things never come out the way you think they would.
 
I have found that when backtesting with Strategy Desk that the if I set the Interval option to a smaller interval that I have in my strategy without checking the Override box the results are very misleading. The intention to set the Interval lower is to check for the Strategy paramaters more frequently but as reality has shown when comparing with backtesting the two results are completely different. I have not had any luck successfully backtesting strategies with smaller timeframes than 10min with Strategy Desk. I am using 60min and 15min bars.
 
Insert foot in mouth. I just got off the phone with Strategy Desk support staff and realized learned how incorrectly I have been using the intervals. Back to the drawing board to make strategy better. So far learning a lot about ATS and still making money which is good.
 
Quote from iancand:

Insert foot in mouth. I just got off the phone with Strategy Desk support staff and realized learned how incorrectly I have been using the intervals. Back to the drawing board to make strategy better. So far learning a lot about ATS and still making money which is good.
More...

Good to hear you checked with them. I share your good results before and after checking with them they turned to crap. Did he tell you the horror stories about folk that do it without checking with them? 100's of trades in a day or hour even trying to get refunds but they can't do it! I feel bad for them.
I didn't check the 1min interval box and used open open instead of close close for buys and sells. I gave up on ATS and using the trade alerts for the indicators I like so I can pinpoint the best time to enter and exit. I am good at exiting at the high but early on the entry by a day or two due to too short a time frame. The screener is great. All things considered in SD I think its great. Better than Stockcharts and free!!
 
I have had some good success with ATS strategies but I found most of it was luck. I had left out backtesting the 1m intra-bar signals. I had had some good trades that just by chance when there were no intra-bar signals. My original strategies looked fantastic on backtests but that wasnt real data, I needed to also test with intra-bar data to see if any false signals were to trigger inbetween. The intra-bar test for the most part were very similar to the actual trades that were fired off.

Most of these strategies I was using breakdown with intra-bar signals, it just seem to easy at first now I am grounded back into reality, and am working on ways to optimize them. A few look pretty promising and will try them out on Monday.
 
I did my first test with them which instead of trading, would send me an email of a trade. And due to a bug, at the end of the day, it sent me 300 emails every minute. I suppose this is what happened to others, so i'm glad i tested first.

My strategy runs every minute, but i still can't replicate my back testing results and i'm not 100% sure why. I can get close, but not close enough it seems. If only we could backtest to the tick level.
 
That is exactly the issue that I discussed with the SD support staff. So with the inability to properly backtest at the tick level and that slippage can affect shorter time frame based strategies much more so than longer time frames strategies I am staying away from strategies based on intervals less than 10min. When I backtested with the intra-bar data set to 1min for one of my strategies I used over this past week that used 60min periods it was almost always within a few minutes to when the actual trades were fired off. This weekend I am working on some new strategies with my new found knowledge and will roll it out on Monday to see what results I can get.
 
You must log in or register to reply here.
Back
Top