Semi-Automated S&R Trading

PetaDollar

PetaDollar

I understand markets through prices, support, and resistance. I found I could develop profitable trading techniques using only these classic ideas. My own personal desire of always wanting to be in a trade has always kept me from being successful in many different ways. Finally, I developed my own trading software which trades like I want to, but without my weaknesses. Today is a great day because I'm trading it live for the first time. My code base is covered with all kinds of automated tests, plus I've done a few weeks on the simulator, so I'm not expecting any trouble.

Today I have it looking to scalp the Sept. NQ at 2070.00, given that there is resistance if the price gets back up there. Today that means the price gets >= 2071.00 then comes back down to 2070.00, without having gone above 2080.00. The 2070 level corresponds to Friday's close and was also an important level overnight. That's where the "semi-automatic" part of my programs come from: I look at the chart, and feed them some basic information about what I believe is important. That's the fun, creative part of trading and I don't even want to toy with trying to teach a computer to do that. However, executing the plan and managing the trade is what I'm terrible at so that's what my programs do for me.

As I get comfortable and my account grows, I will be adding more possible trades to the daily programs. Should be fun! Go ahead and post any comments, criticisms, or questions here. I can't edit the posts. Hopefully we can get some good discussions going.
 
Very pleased with how the program handled the two shorts today. Had I been doing this manually, I'm sure I would have lost $100-$200. Sorry about the graphics... an automatic, good-looking daily report with plot is on the to-do list.

attachment.php
 

Attachments

Nice thread. Looks good. Semi automated is an enjoyable way to do things. Do you care to share what specific logic resulted in the automated exit of your positions today?
 
The first exit was a fixed 6-point ($120) target. The second was a 10-minute time stop. The program has some other exits too, which were not used yesterday- initial stop (which shouldn't be hit much), and +0.5 point stop once the unrealized gain gets to +2.25 pts. I can put in more stop moves as well but I only use a single move for S/R scalping. All of these exits work together with this specific kind of entry.

I love time stops. In my opinion they are one of the key ingredients to trading with technical analysis. If you are looking for a certain market behavior after your entry, you will hopefully find that most of the cases were complete in very specific time frame. No need to be in the trade for longer than that. You can see how it cut my loss on the second trade yesterday. This is critical for long-term profitability. The poker analogy is: good players don't build big pots in questionable situations.

Today's program: Buying around the overnight low (2586.00). It will do it up to five times.
 
The definition of support today was down to 2612.50 or below then up to 2613.00, canceling if we hit 2606.00. So we had two good entries and the third was just bad luck, with the time stop again saving us. Profits were limited by the time stop as well (all three exits were time stops today).

attachment.php
 

Attachments

I am running my trading programs on a remote Linux (Ubuntu) box from Linode. This took a little bit getting used to, especially with TWS Gateway needing a window.

See this link for a light, remote X-window install.

See this link to learn how to exit the terminal/logoff and keep your programs running.

Also, it took some tweaking both my code and the JVM to be able to run on an inexpensive (512MB RAM) package with low CPU. If anyone wants to know, just ask.
 
You must log in or register to reply here.
Back
Top