I have not run into this problem...yet. Maybe the stuff you are doing is more complex. Personally, I have to try and keep things simple (complex coding and webs of rules give me headaches and have yet to improve my system results). So far, every rule in my systems are all in one EasyLanguage (the mother of all oxymorons ) document (other than called functions, for simplicity). So, you can program Ensign to do keystroke macros? Is this like getting a buy signal, and having Ensign, through your macro software, doing an alt-tab to IB, then entering the hotkey for a buy and then the transmit key? Do I understand that correctly? That seems pretty neat, as long as all the components are working correctly. I suppose it would take a lot of faith to actually leave the room while all this was happening, lol. Kirk
Great stuff Tymjr - very helpful indeed. I too appreciate you taking the time to put it all together and for the detailed charts. Very instructive. Very thought provoking. Michaelday - while RavenQuote has some nice features, it does not have backtesting capabilities. StockWatchPro, which also uses the Qcharts data feed, advertises backtesting capabilities but i have not used the product extensively (other than a 1 week trial, but i didn't even look into the backtesting component). I personally have switched to Ensign - it is the most powerful of the three, but it is also probably the least user friendly - meaning that while the capabilities of the product are superior (in my opinion), one could definitely use a background in computer programming to be able to fully utilize its functionality. Just my 2 cents. Tntneo - very cool.....
tntneo: âI like to use a combination of targets and trailing. A part of my position will be assigned a target and another part will wait until it is time to pull out.â Yea, same here. I scale into positions, at times, and I almost always scale out. I will unload portions of my position at different points. Depending on the circumstances, I will leave somewhere between a quarter and 10% of the original position on âjust in caseâ. My target vs. trailing stop discussion revolves around the fact that, although I work to avoid it, I will be stopped out on most of my original full position if the trade never moves far enough into my favor. This is not a problem as I bring in more, on average per contract, in successful trades than I originally risk, even on those contracts closed at conservative targets or shallow trailing stops. Iâve worked hard, though, to try to maximize my winners by working on different trade management systems that are capable of keeping the majority of my contracts on for as long as possible to increase my average.
Oh yeah, you are right. I switched to that approach afterward. but still got trouble with entry at nextbar stuff (buy stop for instance were tricky). but the biggest problem for me was to have a realtime feed working. After a while I gave up. Ensign (and I used AIQ before) is straightforward. I think the same way the new Tradestation online is (my TS experience is only with TS2000!). yes, it is like that. almost. Ensign creates a small file with some parameters, send a keystroke and another script program (specialized in this kind of task) gives all the orders to IB (even multiple accounts !). Yes, you are also right : I NEVER LEAVE THE ROOM ! or someone does check for me. My role is to check the machine and software work properly. if not, I can take over within a few seconds. The goal is to automate things, not to switch on a machine in the morning and take back the cash at 5:00PM ... well, maybe that would not be that bad after all... hehe tntneo
I agree. I was software architect in a previous life, and Ensign is very easy in that sense. If you don't know programming you can still use a lot of the software, just as you would do with RealTick or QCharts. But the real power of the system is when you know programming, it goes much further than TradeStation. If someone knows EZlanguage from tradestation though, Ensign is not more difficult. You can grow into the more sophisticated features easily. tntneo
Neo, I have heard that Ensign (becuase it uses Esignal I guess) only has about 6 months of intraday data to use for backtesting, and an even smaller amount of tick data. Is this true? If so, how are you able to test your systems? Can you import data from other sources? With TS 2000i you can get data files for an almost unlimited amount of back data. Since I started using TS Pro (the online version) I have run into limits on the amount of back data available (they no longer allow for you to use external data like ascii files...I'm really bummed about this). For the mini's, I have been able to pull a continuous contract of 30 min intraday data back to February. I would like to go back over the entire length of its trading life, but I don't know yet if this will be possible (email to support still pending). Thanks, Kirk
Kirk, yes Ensign is limited in back data. This is the major and only issue I have with them. This is a legacy 16 bits problem : you can see only 2338 bars (whatever resolution, so 1 mn bar would go back in time much less than 60 mn bars). (esignal gives back 2 months of intraday, tick data is 10 days I think, but still with the 2000 bars limitations.. that gives you 2 hours of tick data on NQ ). I did backtest with TS in the beginning actually. But since I moved to Ensign I can do backtesting with saved files (it is tedious but you can keep files and backtest them later. esignal data is used only in realtime, the chart is saved on disk so you can archive them). This is a pain, but shows how much I appreciate Ensign Also, my system is very active so I get statiscal validity pretty quickly.
Neo, you said "If someone knows EZlanguage from tradestation though, Ensign is not more difficult. You can grow into the more sophisticated features easily. " I've tried both and I have to say EZlanguage is much, much easier than Ensign. I spent about 10 minutes with a EZlanguage example and quickly came up with all kinds of signals that I could test. I printed and read all the Ensign manuals and examples from their website, and still could not make sense of its language. Do you think there's any way one could learn it quicker? I just regret that I didn't spend more time on computer classes in college... Francis
I am not a good benchmark to judge if it is easy or not because of my background. And I was also lost in Ensign in the beginning. What helped were the samples.spt file. You have there a few examples of backtesting. primitive but it gives you the idea of how to use this feature. The problem is the lack of samples (and better ones). With Tradestation you have plenty of them with all kind of indicators. All garbage but what one would expect with free samples. The point is, you can quickly find out how to do it. I bet without the samples you would be lost in TS as well. I could give you some pointers and you may read the help on the web site, it is very good because it shows how to do things as opposed to what things do. http://www.ensignsoftware.com/espl/espl.htm I know I sound always so pushy toward Ensign, while I don't get a dime from them or not even better support but if you read my past posts you will see I tried many before finding and using them. There are not perfect but they offer a very powerful software for just a few bucks a month. Maybe for non-programmers, Tradestation is the better way to go. tntneo
Hi Kirk, with regards to your previous mail (How much data?), here are just few words of my experience in this area.. As Neo mentioned few times, Ensign may only backtest 2338 bars of data at a time, so typically about one week of 1 minute day session bar for Nasdaq stocks. But, with the quite extended build in import functionalities, it is very easy to import an ASCII file (from various sources) into the Ensign quotes database. On my side, being as well a Ravenquote user, I got access to qfeed. So, when I can't retrieve data because of the limited esignal databank, I use the Qcollector program to create custom ASCII files from the Qchart database and then import them into Ensign. This way, I can backtest up to two years old 1 minute bar data (but always 2338 bars at a time). In fact, as I am mainly interested by the first 90 minutes of trades, I just delete the other bar data so that I end up backtesting about one month at a time. And I am further working on ESPL scripts to run automatically in a row backtesting batches for different kind of stocks, periods and trading system variables. That's one of the area where Ensign programming language brings significant improvements versus Ravenquote. However, for example, for scanning stocks against specific alerts, I still prefer the standard functionalities of Ravenquote quote sheets... Today, while daytrading, I run both programs simultaneously : after one year in quest of the ideal software, I finally settled with this dual combination : => Ravenquote for the market environment charting (indexes, futures, PREM, TIKI, etc..) and a 1st stock scanning and selection => Ensign for charting the individual stock indicators and providing the live entry and exit signals coming from my trading system. The next step could be to have an automatic link to my broker, but's that's for sure beyond my present programming abilities. Let's hope Neo will be able very soon to offer us such a first class system Happy trading Eddy