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)
I asked a question but Big Mike closed it and told me to ask it here, so here goes:
Before we start, a few things I'd like to point out:
- I know how renko bars are formed, I am not an idiot. Please don't just say that they can't be used in real time, because they can under most market conditions.
- I will not execute any strategy that evolves in NinjaTrader, so any Limit/Market order problems in the NinjaTrader backtesting language are not applicable. I have slippage under wraps too.
- As I mentiond in the original question, I will be (or at least try to) using tick (trade, Bid X Ask) data for a significant period (say 1yr backtest, 5yr out of sample). In all liklihood I will be doing this in Matlab. Any stategy that does evolve will work from Matlab througbh to TT (after I have trded it live for a while for sanity).
So, the question is "has anybody tried this?".
CPI
Can you help answer these questions from other members on NexusFi?
Then it'll all be in Matlab I guess (that should be fun eh!)...
Oh wait, you mean like C#, Jave etc?
I'll probably get someone else to do all the TT / Matlab integration stuff. I'm not a programmer, see. I just want the data in renko so I can do all my trader-stuff before handing it over to the boffins!
The basics I will do myself, because it is just a case of getting the data into matlab (I say that now, I have only really used matlab for econometrics etc so far, so I expect it to be a bit of a challenge).
Initially I will be using tick data so that won't come from TT, probably tickdataplus or somewhere. Then if I get somewhere I will get someone (mates of mine) to build up some simple program that uses TT data from Excel (i.e TT -> Excel (via RTD) -> Matlab based application -> My mince pies -> back into TT)
Then if we are still GOGOGO it'll go tt -> matlab -> Autotrader via COM (I think, I'm not a computer boff).
The computer bits of it I am not worried about, I've seen it done loads - its the strategy thats the important bit!!
I have coded strategies for Renko Bars in Easy language, but the bottom line is that what you have to code, tick by tick, is a series of "bars" which extend up to one brick-size above and below the bricks shown on your chart, so they really are not "Renko" bricks which are to be intercepted by your "set target" and "set stops" but TICKS! You can plot the renko bricks, but where the price actually moves can be in the open spaced above and below the "open" and "close" (until triggering a new brick by exceeding or touching as well one level past the "high" or "low" of the prior bricks. Since a backtester built using "reserved words" gets wrong information given said highs and lows, it cannot "see" those phantom non-defined price moves above and below the plotted "bricks", and it will give outrageously wrong estimates of your actual real time trades. It is more a matter of the stop orders (stop loss) and target orders (limit) which will execute outside of the bricks in real time that make your actual coding have little to do with those simplified visual "bricks" on the chart.
So, my little programming adventure with Renko ran way overtime, and was really a tick based strategy which happened to use renko construction retrace and continuation "rules" to give trading bar close setups only, and tick by tick monitoring between those new setups for the stop and target hits, and the cancelling or resetting of the stop and limit orders whenever each new brick closed out and the new brick started forming. That "forming" could finish, or close, one size above the prior high, or one size below the prior low before it completed, and was ambiguous the entire time in between. The chart may look like Renko, but reality will not look like Renko. And the profit will not look like renko profit ...at least it was a big disappointment for me in the end, but not a total waste of time, either. Since I do not know the programming of Ninja Trader, I do not know how it handles "look inside bars". I do know that EasyLanguage does NOT do it at all, let alone properly with Renko charts, Point and figure charts, and line break charts and TradeStation says it (back testing properly) is not supported for those chart types..
So to sum it up, if your renko strategy is working properly in the back test, your hypothetical executions, if they involve stops or targets, can occur, as they would in real time, anywhere in a three brick range before you see one little completed setup brick appear on the chart.
