I have around 25 strategies for stock trading that I created myself and I would like to autotrade, but I need some advice on the best way for me to do that. I am willing to use any broker and any trading software (Tradestation, Ninjatrader, etc.) but I already did autotrade a strategy in TradeStation back in January 2013 and everything worked fine so ideally I would like to keep using Tradestation.
The main problem I had with TradeStation was that there was no easy way for me to backtest my strategies on multiple stocks all at once and see a combined report of the backtest results. I tried Portfolio Maestro but could never get it to work with any of my custom program strategies (it gave me an error compiling them). So, I had my programmer create a custom built backtesting system for me. He used Python and C++ and it works great. Now I want to start doing live trading for some of the strategies I backtested. The programmer wants to build his own trading module on top of the backtesting module, since that way he could use the same exact strategy code for backstesting as for live trading, but I would much rather use off the shelf automated trading software such as TradeStation. For each of the 100 or so strategies I am backtesting, there are hundreds or thousands of optimizations and variations I am also backtesting, but in the end I will narrow that down to around 25 actual strategies that I will trade, so it seems easiest just to convert these "winning" strategies to EasyLanguage (or whatever auto trading platform I use) and not worry about programming an auto trading system from scratch. I also looked at using NinjaTrader or Tradelink for this, but I like TradeStation best, mainly because the brokerage component is fully integrated instead of using an API.
Here's my basic plan for automated trading:
1. I will run TradeStation from a Windows VPS (probably speedytradingservers.com), and my programmer and I can both login to the VPS remotely to view the TradeStation screen.
2. I will run around 25 strategies on a portfolio of 40 stocks (no futures, options, or Forex). None of the strategies use high frequency trading, just day trading using 1 minute time intervals, like for example buying all stocks at the open that went down for the past 5 days in a row and exiting for a profit target or at end of day.
3. I will run some script to make sure all the automated entries and exits get executed like they are supposed to. Trade Watch at Trading Platform Servers Hosting VPS Forex Stocks Metatrader Tradestation Ninjatrader Multicharts | Triage Trading ( seems like it would be good for that.
4. To start with I am not too concerned with money management. I have $1 million in my account and will do $100,000 trades, so I can have up to 40 trades a day using day trading leverage. Most of my strategies trade at most several times a week, so there will only be on average 10 trades a day. If for some reason my strategies wanted to enter more than 40 trades in a day, it would be fine if it just uses the first 40 trades that are triggered and ignores the rest. I can also maybe set a limit for each strategy so it only trades x times per day, so that way there is room for each strategy to have at least a few trades.
5. All my strategies are 100% automated. I do not want to do any discretionary trading.
6. All of my trades are market orders. I do not use limit orders.
7. I will not be using TradeStation for backtesting, other than a small backtest for each strategy to make sure the trades in TradeStation match the results in my custom backtesting system. I want to make sure I am using the same exact strategy for live trading that I backtested.
Based on what I am looking to do, should I use TradeStation?
If I do use TradeStation, these are the issues I need help with:
A) Is it better for me to always keep 40 charts open (one for each of the stocks I trade) and use the built-in automatic trading, or should I instead use object oriented programming in Easy Language and not use charts? The only problem with using placing …