I have built one. PM me and I'll send account details back. .. Seriously though, it's not impossible to do for a price/time priority market. Difficult, but not impossible.
I see what you mean, and i may be misinformed but i'm not sure how the order book would significantly impact trades (eg limit orders only) of an average duration at 1 hour and more on the top 20% most liquid markets (assuming we are allowed to make such a selection). In addition, those problems are modelled (approximated) in a good backtesting platform. So you're saying that even 1000 trades in OOS and an unchanged equity curve (compared to in-sample) has no statistical significance whatsoever - even with very robust and pessimistic modelling of real-world market dynamics in the backtester? Edit: Oh and, we're forgetting the possibility of constructing portfolios with machine learning. Hedge instruments, find correlations, using options, means more data and more statistical significance in one go, use order clouds, etc etc. It doesn't have to be applied to a single instrument.
I meant the filtering aspects, feature extraction etc. The hard bits. Anything post-trade is a dead end. Intellectual wasteland and career limiting quagmire. (Or a good job in Bangalore).
Which software available for sale for retail/semi-retail would do a "somewhat not too bad" job of simulating a limit order book? Would you name a few alongwith ballpark price. Thanks. Btw, I think braincell meant not long term trading strategies, but intra-day strategies with holding period of multiple hours. Also, I assume he would simply pay the bid-ask spread + slippage and would not worry about simulating fills at this stage in his trading strategy design and development endeavor.
Simulating fills is super-important, are you kidding? In most cases it's enough to count a limit order filled if the price traded above/below it (not at it). I don't see any flaw in such thinking. And of course, directional trading always pays the spread, who else would do it. Models for slippage can be approximated by simulating ticks, but again, with limit orders only you get the price you posted if it's penetrated (or better) so i don't see a problem with that.
that was my point the whole time: Such results may be highly statistically significant, however, they generally do not hold up to data snooping (alas my butter example) nor do they correctly account for the real cost of execution (spreads, slippage, disappearing bids and offers). The result of that study was highly statistically significant but how much belief do you put into the stability of the model going forward. Correlation trading, by the way, is pretty much dead. Hardly anyone survived 2008. Hmm, guess what happened, correlations were after all not correlations, sounds trivial but is anything but that.
Sorry but I didn't clarify. I am coming from the futures world, where size on inside bid-ask is sufficient for my purposes, that I don't need to simulate any fills. I understand that this is super-important when you are trading stocks with say < 10 million per day volume. So, you see as of now, I do not do any simulation as far as my fill rate is concerned. I just do a simple bid-ask + slippage and thats it. But I know that I will have to move to a more complicated world when I start trading stocks (not happening before 2014 imo). So, trying to ask if some good softwares do the work out there for me and how much do they cost. How do you simulate your fills currently? Your own software or something else ?
I think you did not fully get my post. There is none, you can offer 10mil and wont find a chance in securing a great order book simulation anywhere. My point was that you can't seriously declare victory of any intraday trading strategy until you actually implement it and expose it to market dynamics. Lol, intraday and not simulating fills? Paying bid+ask all the time? And how would he model slippage? I thought I invested time but I feel I wasted it...;-(
What's the matter with everyone i talk to that has as much as experience as you do, which i respect, thinks that every backtesting app built ever doesnt include real-world market aspects like spreads, slippage, etc. My one does, and disappearing bids would be important if i was an HFT, not a 1 hour holding period. But ok anyway, what you're saying is that future cannot be foretold. Ok now we get to walk forward testing. If we get two OOS samples, 25% of data each, we can find 100s of systems and see if positive OOS1 data correlates to positive OOS2 data and the scatter graph looks nice, what, you're saying that means nothing - again? My opinion is if there is correlation in OOS1-OOS2, we can safely assume that when we get only OOS1 that is positive, OOS2 (which now becomes live trading) should have a statistical chance of holding up which is equivalent to the previous OOS1-OOS2 tests. Anything wrong with that logic? I agree, i did some testing a year ago and the fuzzyness was profound. I'd rather trade the VIX heh.
Buddy you got an attitude I admit it! Please read my post I replied to braincell that I am talking in reference to futures trading. And mind you I am not trading for a fund wherein I need to buy/sell notional of 100 million within 2 minutes. I am running an independent trading account and for anything less than 10 million notional, my modelling of 1-2 tick slippage with spread is good enough. Don't jump to conclusions mate. and calm down if you want a proper discussion. I don't mind if your are massaging a bit of your ego with your posts here - thats fine - but everything in moderation Anyways, the point is you didn't fully get my post. I asked for a "somewhat not too bad" software that will simulate fills, not great top of the line product.