John Devine
Junior member
- Messages
- 13
- Likes
- 0
Hello dmilong. I'm using aps for about two months. I think you will have a hard time getting a used copy. It's bettter to order it through the company Tradingpatterns.com. If you need any hints on how to use the program let me know and I will try to find some time for you. APS is an excellent piece of software, it is very different from mainstream trading software but user friendly and it does what it is supposed to do.
John
John
Hello John and thank you for you offer. I purchased APS and i wonder if you can help me to develop for me a few support functions in excel or some other prog-lang to handle open positions on a consulting basis of course. The program is really cool and something really different as you said. I e-mailed you a week ago but no reply, maybe you are using some spamarrest service. Looking forward to your reply. Jim
Has anyone done any forward tests of the patterns generated by APS? I installed the demo version and it appears user-friendly and easy to use. I think there is a way to forward test the patterns in the demo version - the code is not necessary.
If this really works it's the way to go. Ron
If this really works it's the way to go. Ron
ronblack said:
I'm a new user and I'm working on this issue right now. I have two copies of the softwarte, one to get patterns and the other to forward test them (tradingpatterns.com will give you two passwords for running APS on two different machines if you ask for it).
Initial results for S&P 500 futures tell me that a good number of patterns pass the forward test. About 30% of the patterns I have tested up to now fail the test and half of these turn unprofitable out-of-sample.
Conclusion: the more patterns that show up in a scan the higher the probability of the signal hitting its preset target.
Next, I plan to forward test delay patterns.It's amazing what this program can do. Jim
dmilong said:
This reminds me of something like the Bayes rule in statistics. There is some true into that unless the market suddenly reverses due to some non-technical condition or other development. But overall, this is a sound mathematical approach worth looking into.
Good idea dmilong
Ron
Pat494
Legendary member
- Messages
- 14,614
- Likes
- 1,588
Jamy said:Hi,
check out www.tradingbasis.com
They offer much more tools. The website looks better. i have not tested the APS tools but i have purchased the tradingbasis tools and i am happy so far
Jamy
Are you still happy with tradingbasis ? Please could you say why you think it is so good.
Thx
Pat494 said:
I just checked the purchase page and the following info appears: Publisher: Beate Zmuck, Austria.
There is no clear address or any tel number/fax on that website. I wouldn't trade a black box of this kind. There is no historical performance analysis available for any of the signals the program generates. Actually, the user doesn't even know how the signals are generated in the first place. The price is about $270 for the complete package , which is reasonable although you can purchase a Metastock add-on at this price range that will do a lot more than this program and you won't have to switch to Amibroker.
Jim
Has anyone read about Ed Seykota from Market Wizards? He was one of the first and still only using complete mechanical trading systems. Considering he built his first profitable one in the early 80's it probably only consited of moving averages and some sort of trailing stop.
damo
damo
dmilong said:
Hello Jim, I have APS (version 4.9) and it has worked well for me but I need to get to the next stage of organising the output of the program and calculating probabilities along the same lines you suggested. If you are still around please send me a message to discuss possible cooperation. I have realized APS has many capabilities I am not utilising at the moment.
Alex
Pat494
Legendary member
- Messages
- 14,614
- Likes
- 1,588
dmilong said:
Hi Jim,
I see you are testing the program out on the S&P 500. Please could you say about how often does the program throw up a buy/sell signal on say <3 consecutive losers. Would it be about 1 a month or 1 a week perhaps ? and is the profitable trade ratio about 70% ( as per the sales blurb ) ?
Pat494 said:
HI Pat,
I'm through with the testing now and I.m using the program to position trade the e-mini. I used the S&P 500 to test the program because of the longer history but afterwards I got good results for the e-mini. I get 2 to 3 signals per week on the average but on certain occasions I get many more on a given close for the open of the next day. This allows me to do some statistical analysis based on the signal quantity and frequency.
Profitable trade ratio: if you noticed the program has a calculator of this ratio given the profit factor you want and the profit target/stop loss (the formula is also given in the support page of the website, see: http://www.tradingpatterns.com/Support/UseAPS/useaps.html#usefaq4 ). You can use the calculated ratio in the search for patterns. When I use larger stops then the profit targets I still get patterns with a ratio of 80% or more. I've done searches for a total four combinations of targets/stops and I added all the patterns I liked in a trading system in APS system tracking. After the market close I update the data file for the e-mini and I run system tracking to see whether new signals are there for the next day. You can play with many different strategies based on the number of new signals and signals with open positions. I'm now trying to figure out a good strategy.
I got an email from Tradingpatterns yesterday about a new article by Mike Harris that discusses the main purpose of the program:
http://www.tradingpatterns.com/synthesis.pdf
I think it's very interesting.
Jim
Pat494 said:
Pat,
You can convert a number of metastock generated ascii text files all at once. I'm sure you figured that out.
I spent a good part of this morning converting files too. It's pain but worths the trouble. If you run into problems or need help let me know.
Alex
