I'd love to set up for automated trading and have tried to do it for a long time but have never had any success with it. I couldn't see much problem in setting up an automated system. I can do that with Microsoft Excel receiving data and sending signals to Interactive Brokers. I've tested that in simulated trading. The much bigger problem I see is finding a system that is actually profitible. I've come up with hundreds of ideas both simple and complicated but when I test them, they all just come close to breaking even. For instance, I've tried all kinds band trading systems, where you buy at the lower band and sell at the upper band with various filters. I've tried breakout systems. Moving average crossovers. Etc etc. Even if I could find a system that loses a lot of money, that would be perfect, I could just reverse the buy and sells. Does anyone have consistant success with a mechanical trading system? Could you give me a hint which direction I should look? Do you have more success with futures or with stocks?
I run ninjatrader on my home pc over a fast dsl line and remote desktop into it while I'm at work to monitor it. It is a fairly simple system that trades ym and generates 1-4 trades per day. Not ideal but with my system stops and targets get entered with my entries so even if I completely lose connection the downside is limited and I can also log into my broker directly from work and close trades if necessary.
woodsidetrader, In case of a loss of connection with the broker, does NT automatically try to reconnect, download any missing data, and synchronize positions and strategy logic? tks
Not 100% sure on that. Fortunately I haven't run into that issue yet. With ninja you can setup multiple datafeeds with failover. I run esignal with my broker's datafeed as a backup. Not an ideal setup doing it remotely but hopefully I'll be trading full-time before too long. The ninjatrader forums are pretty active with the developers so they would be able to give you an accurate answer.
Sam, You might check out RightEdge. It is a .NET trading platform. See also the rightedge-ats Yahoo group. In particular, there is a strategy being covered on the group that might be of interest. It does not sell a position at a loss, so ultimately success depends on stock selection. The strategy itself is no big deal; it just tries to take advantage of minor up excursions from among a selection of stocks. In this up market it has been profitable (no surprise there either). It is an interesting way to get into autotrading. Because of the no-loss rule the strategy itself can't get you into much trouble; you do no worse than buy-and-hold generally. Also contributing are manualy entered (partially) covered calls on positions that move adversely. It is a quite unsophisticated approach to autotrading. (Disclosure: I am the owner of the rightedge-ats Yahoo group and a novice at autotrading.)
Another option would be to change your schedule so that you can trade the overnight markets and then go to work during the day. If you can manage it, the CME EuroFx makes some incredible moves between the hours of 2am and 6am, ET. Just a thought. Good trading, JJ
well, i've been building an ats(s) for a bit..but i haven't put a successful system online live, so take that fwiw. i would suggest looking at successful methods used by discretionary traders. an example would be the AHG thread. there are others of course. also, when looking at your system results that failed.. look deeper. why are the failing? is it a trend following system thats getting whip-sawed in choppy markets? (or vice-versa). sometimes, it's not the perfect entry you seek, but weeding out the less than perfect ones. it is a fun puzzle really. and imo .. it can be solved.
I trade a few systems that at first glance are perfect for automation. But my efforts so far have been frustrating because of data spikes, system glitches, etc. At some point data and systems may be perfectly reliable. But we aren't there yet (may never be), and I need to be involved in pulling the trigger on every trade. While that means remaining within hearing distance of the computer for the trading day, there are certainly worse working environments.
I'm not automated but I have set up the next best thing. I have all my systems programmed in esignal. I have set up email alerts on esignal. Esignal sends the trade signal to my outlook. I have my email forwarded to my pda. I get my emails pushed through to my phone every minute. When I get an email on the pda, it's set up to bark like a dog. (So I know it's a trade signal). The only emails that get sent to my phone are tradesignals so the data usage is quite low and cheap. I only get 4-5 signals a week. I can also access my broker account from my phone to put a trade on or close one out. However I prefer to use my desktop if possible. The 3g network connection I have is very fast. If you can't use your work computer for trading, another option is to carry a small notebook with a wireless connection. Or even use your phone as a wireless modem. I prefer to manually place my trades just so I know what my system is doing and I can watch my trades, particulary early on when they are not yet in profit. Runningbear
Technically, all the broker software and retail trader software, maybe no software on earth really, is ready for prime time regarding unattended trading, bugs can show up long after release when something stresses a system... If I was going to go unattended I might colocate and let somebody else handle the IT end of things to destress my responsibilities. Having an account stop residing with the brokerage would be a good thing, do any offer that service? Keeping some insurance in place in case of sudden moves the wrong way with no way to get stops filled, as in a 9/11 type of event, is a good idea anyhow...