Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Non Day Trade: Bought DOW today at 45.79 to complete that spread. Gave up buying JNJ and bought UNH instead to complete that spread. Bad idea UNH got beat up. Tried to swap COKE for KHC in that sector but only got filled on 1 share of COKE. Yes, 1. I am not kidding. So kept the KHC for now. I think I was up about 213. I said "I think" because what IB shows on my screen only seems to have a small similarity to what shows up on my equity run tomorrow. We shall see.
Day Trade Side: Another day where Russell just carried the ball. Only 2 of my legs involved Russell, but those two legs were nicely profitable and offset the loss I took on the middle leg. For the week the model has been up 2190.20, 2472.20 and today 1402.08 per unit. Doing 2 units each day I have captured most of the last two days after missing out on quite a bit Monday because I was away from my desk. Finished today +2636.
Tomorrow I am going to go live with the Ninja Trader version of my day trading program. It's a little nerve wracking, but also exciting. I am making progress with my C# skills, but certainly any flaws will be revealed at the cost of real $$
So far I have not been able to translate all of my Excel algorithms into my C# code. This causes two problems, one I can work around, one I can't. I'm just going to have to learn more C#.
The first is dynamic position sizing. My algorithm uses volatility primarily as a means of position sizing. I (so far) cannot duplicate IB's Historical Volatility readings. I think I'm close, but not there yet. So I wrote the code to make contract size a parameter and I will just reset it at the beginning of each day if necessary. It doesn't bounce around too much. In fact position sizing has been the same for 8 straight days now, so changing it when needed will be a simple task. I'll just calculate it in Excel, like I always do and then enter it into the Strategy's parameters.
The 2nd issue is that this same input is needed to determine which market to buy and which to sell. I've always thought that %change would be about the same but looking at backtest results, I see a fair number of differences. Until I can write the code to duplicate what I am doing in Excel, I am going to use %Change for the selection process. It's an interesting question: Is my way really any better? I guess we'll find out, though the smallness of the sample size will perhaps invalidate any conclusions.
One thing I know will be way worse is the fills. In my trading I always enter (exit) on a limit order, and generally get positive slippage on one side or both. Once in a while it gets away from me and I have to capitulate and just get in (out) and I end up with negative slippage that day. But overall my slippage is positive because I am willing to wait it out if the market isn't moving fast. With Ninja there is no way I have the skill to write that kind of logic, so I will be buying the ask and selling the bid every single time, both in and out. I expect that to cost me about $100. per day and is something I am going to track. Maybe it will be better than I am expecting, and maybe it won't and I will have strong incentive to create computer readable logic to getting better fills.
I only have 4400 in the account, so if we start with any kind of a draw down I will have to add money, which I am willing to do. It has done exceptionally well these past 3 days, so odds are it will have a few losers out of the gate, but I am not going to guess - I am just going to get in and see how it does.
I am not going to touch these trades at all (other than a program crash, or I see a blatant error with my code), so no adjustments, early exits, adding or subtracting throughout the day, none of the hands-on stuff I sometimes do with my IB account. Just going to plug it in when I get up and let it do its thing all day. I have SIM traded it the past couple of days and it seems to work. My biggest concern is how it handles a negated trade. I've gone over that code time and time again and it looks good to me. It also seems to work correctly in backtesting. But no trades were negated these past couple days in live SIM trading, so we shall see.
I will report on my daily summary how it does and any differences I note between it and my IB trading.
Good luck with NinjaTrader. An option, which I'm sure you have considered, is to use another platform that integrates well with Excel, and use VBA code to fully automate your strategy. I do this with R|Trader Pro.
Yes, I know there are other, perhaps easier ways to skin this cat. And I am quite proficient with VBA, but a lot of things I want to do require a more robust programming language. As I get older I want to automate more and more of my trading, if possible. I did consider learning Java and the IB API, but decided to try Ninja and C# first. I also had planned on building an Option model with Python and the numpy library, but my options trading is somewhat on the back burner right now. So Python goes on the back burner too. For now.
As always, thanks for the input. Always appreciated.
Non Day Trade: After a couple days of moderate gains, the trades lost ground today. All 4 spreads were negative on the day, though I have no idea how much as IB never bothers to print settlement prices. I'm going to guess somewhere between 200-300 down for the day. I am about done with the equities, at least until/unless I find something that I like and think can work.
Day Trade: Well, my Ninja trade was a bust. It was rollover day and I got everything set up with the Dec contracts, and fired it up. And then it came time for leg 1 and nothing happened. I have no idea why. Maybe because the spreads were gigantic. I don't know. All I know is nothing happened. I entered the first trade manually, but after about 30 minutes I decided that that defeats the whole purpose and I closed the trade. Got lucky that the trade made about $100. but I would have closed it either way. I am going to try again tomorrow.
As for my regular account, as I mentioned, the spreads were huge. Over $100. for each leg. I have never seen that kind of spread with the e-minis on rollover day, so it appears that the micros are not ready for prime time in that area just yet. I looked back at last rollover and saw that I traded Jun on the Thursday and Friday before going to Seps on Monday. I may have to do that, but I will try the Dec tomorrow and see how it looks. I did manage to get the first two legs on, but it was very difficult because of the massive bid/ask spreads. And then I was worried about getting out if it went haywire. We got a Trump bump that helped my trade and I bailed with a 768 dollar profit and called it a day. Then I spent the afternoon complaining about my equity trades. I'm tired of them and will probably bail tomorrow even though I'd originally said I'd give them 1K to work with.
I changed everything today to the December contracts as well, and even the e-mini's (particularly YM and RTY) were very sparsely traded with big spreads, so I can imagine what the micro's must have looked like. I'll stick to December tomorrow (Friday) as well, but next time will rollover on the following Monday. Lesson learned.
Non Day Trade: The equities continue to not work for me. I didn't close. Yet. Down another few hundred.
Day Trade: A pretty good day. Trade was mostly sideways all day, had one real good spurt, and then back to mostly side ways. Finished+1480 to end a very nice week.
Ninja: Well, I initiated the system with the Dec again today,and what do you know? This time it worked. It took the trades at the proper times and executed properly. It did, however take a different trade than I took with my Excel based selection. Super small sample size of 1, but so far my way is significantly better. Ninja lost, after commissions and fees 314.40. Howver that does not count slippage, which was both worse and better than I had anticipated. I was guessing I'd lose about $100. per day and I lost 232, so in that sense worse. But 187.50 of it was on one leg of one trade (on the M2K) so without that it was "only" 44 dollars. I'll have to see what happens going forward. Obviously 232 would not be acceptable on a daily basis. So with the model's loss, commissions and fees, and slippage, the total damage was 546.50 and my Ninja balance now sits at 4041.92. So not an ideal start, but we'll see what happens next.
I am going to spend some time this weekend thinking about how to define a limit order entry strategy. In my manual trading, I always enter on a limit. If I don't get filled, I a) wait it out b) bump my price a little c) capitulate and get in at the current price (though still on a limit) or d) say forget it and close the side of the trade that did get filled, usually at a loss. How to define that? How to codify it, if in fact I can define it? Stuff to think about.
