I am developing a Trading Simulator that uses a datafeed from esignal for ES futures traders. The simulator currently features: * Realistic market feel [few second delay] * Market Orders that buy on ASK and sell on BID * Limit Order Capabilities * Stop-Loss Capabilities * Trailing-Stop Capabilities * Real-Time Profit-Loss Updating (Very fast) If anyone would be interested in trying this out, please let me know. I am programming some pieces tonight and it should be ready in a few days. If there are any other features, please let me know. I will eventually be programming a full featured autotrader that will have volatility adjusted trailing stops, etc. Is there any interest for this? aphie
if there is a few second delay then it isnt realistic market feel.. of course, you wouldnt know that would you.. but, really if you want a simulator there are some good ones out there.. im currently using ensign and they have a simu-broker and a market playback feature which will play back any previous days data for the ES and NQ as if it were real time.. and they have an integrated simulated broker.. its the best thing since sliced bread imo.. and there are other simulators out there as well.. -qwik
Alright, I'm about to go to bed for a few hours so I can catch the morning action -- but here is a screen-shot of the program thus far: I am thinking about programming some form of advanced probability indicator that scans previous market-data and comes up with a probability score of a long or short shot into the market -- but that would require some serious mathematics and would take me a little longer. I'm not sure exactly how I would approach that, except to maybe combine some 1, 5, 10, 30, 60, daily level stochastics and find extremely oversold or overbought levels across multiple time-frames. Meanwhile, I'm going the path of a hybrid trading system that uses mechanical / intuitive user input. I'd like to get the best of both worlds as much as possible. Here's the screen-shot with explanations.
This project is getting a lot more ambitious on my part. Not only have I successfully been able to interface with esignals Turbofeed, but I have also been able to come up with some interesting ideas on an auto-trader function built-in to the program. I am thinking about making it switchable between simulator and real-time autotrader with IB + esignal. After pouring through a lot of tick data, I often see that when there is a sudden increase in buying or selling, it can last up to 5 or 10 seconds before subsiding (sometimes longer of course). The "Momentum Scalping" feature takes 10 snapshots a second of the market and when it detects a sudden surge in buying or selling, it goes for a .50 scalp. I am going to put this in a simulated environment with IB and see how it does, but it can react FAR FAR more faster than a human ever could -- another advantage. I am also looking at a Volatility scalper "autoscalper" that works in ranging markets. Here is a new picture: aphie
Deep, I am using Visual Basic Version 6.0. There are API's to interface with IB, and esignal has a DDE connection to their data streamer. I put my program right next to esignals and it was just as fast, even though DDE is ... what .. a 14 year old technology? I asked them about any future enhancements for developers but they apparently have their hands full with "normal" users, much less "advanced" ones. Thanks -- I'll be happy to send you a copy to play with when I get it further along. aphie
DDE only eh? It might get clogged up during high volume times of day like the open. Speed and response would be critical for a micro-scalping system. DeepHiney
Deep, I'd like to discuss with some of the people at Esignal a more advanced and robust method of data delivery into a program outside of esignal's resident program. However, I was using it this morning around 9:45am on the ES e-mini's (just testing out some of the interface gadgets), and my program was tearing through the data like a champ. I only have an 800 megahertz Pentium III, too. I minimized my program and put it right next to e-signals, and it was exactly indentical to their updates within their program. Not too shabby for an older technology. Whatever speed advantage I would need to lock into a short-burst momentum wave would probably be found in reducing the latency between esignal to me and me to IB brokers. However, even without any adjustments, I'm sure that my program would be able to buy or sell into a fast momentum spike quicker than 90% of the manual off-floor traders, and as long as some of them buy or sell after my program does, I'm good for the scalp. We will see in the coming weeks -- and if I have the guts to take it live with real money -- but only after successful trial runs in a simulated mode! aphie
Just make sure you are conservative with your fills on your sim. Fill longs at ask and shorts at bid. Good luck. Deep