Help with Renko Backtesting

Oct 02, 2017 at 15:47
7 Replies
Member Since Sep 29, 2016   264 posts
Oct 02, 2017 at 15:47

If there is anyone here on myfxbook that is experienced in making high quality backtest for RENKO charts kindly comment or send me a PM.

I have developed a simple renko strategy that looks promising on demo forward testing, and I have tried backtesting with many methods but the backtest quality delivered is very poor and not representing accurate results. I suspect that this is because MT4 does not have quality tick data or is not efficient with renko backtesting.

I have seen other platforms that can achieve this (Sierra Charts) and other softwares for doing this but they are quite cost prohibitive!

I am more than happy sharing the MQL4 file with whoever is able to help me with this issue.
Member Since Jan 05, 2016   1189 posts
Oct 06, 2017 at 21:04
MT4 Renko charts are generally 'Offline' generated charts with custom time frames and brick sizes.
The back tester doesn't work properly with custom time frames and defined brick sizes.

If you want to test Renko strategies my suggestion is to run the EA on a demo account for at least a month and just let it process as it normally would. If it looks profitable on demo, then I would run it on a live micro or cent account.
If it looks too good to be true, it's probably a scam! Let the buyer beware.
Member Since Oct 23, 2012   349 posts
Oct 13, 2017 at 17:15 (edited Oct 13, 2017 at 17:16)
az-invest offers custom backtest scripts for 99% accurate renko backtesting, i have them all they work great!

if you need me to backtest for you just shoot me a PM
if you follow the flock like sheep you always end up stepping in shit!
Member Since Sep 29, 2016   264 posts
Oct 14, 2017 at 16:47
@Professional4X I am running some forward tests in demo for a few months now and they look ok, but really want to see how they would have done in the past. I have found other providers like Sierra Charts that have backtesting, but they require monthly platform fees and paid tick data! No thanks :P

@sirius1fx Many thanks for your advice, will PM you shortly!
Member Since Mar 20, 2011   44 posts
Oct 17, 2017 at 07:57 (edited Oct 17, 2017 at 07:58)
I found FXT Builders which use the bar and tick data from your charts to create Renko, Mean Renko, Range, PointO, Flex Renko and Range Bars, to working very well.

I use Tickdata to get my history and drop the script on the chart and end up with Renko charts in the Tester.
Member Since Sep 29, 2016   264 posts
Oct 17, 2017 at 12:26
@OnTheEdge Many thanks for your advice! I had been considering about the OVO product you mentioned but could not find any feedback regarding the quality of these products. The price also seems reasonable so will be giving it a try.

Have you got reasonable results so far using the OVO product?
Member Since Mar 20, 2011   44 posts
Oct 17, 2017 at 18:37
What I liked was the ease of creating the charts and since it's time consuming I could reasonably create charts for multiple currencies at the same time.

I created a legend so M1 = 2pips, M5=4pips etc. or something along those lines so when I tested different pairs I wasn't guessing which time frame was what size Renko.

As far as results my backtest and forward testing seem to correlate. Although, I did develope a nice strategy using az-invest renko charts but could not get the backtest to match with OVO results, not sure why.

One thing I found is setting the start Date &Time when creating the charts made a difference in backtesting results and in forward testing. This is a result of how the Renko charts are created and I think it's important to keep this in mind when creating your charts. This should also be your starting point when you go live. You could use your backtest data and than pick a Date & Time to start your live charts that correlates to a bar open/close with your backtest data.

I also always set the Rest Open on New Trading Day to True, if you don't you end up with these beautiful runs when the market gaps and in backtesting you will have trades that couldn't have existed but do and are profitable. Really skews your backtest results.

One thing I thought about trying was using the ATR option to create the brick size as I think this would help when going live as the bricks would increase or decrease in size as market volatility increased or decreased. However, backtesting this theory is very time consuming and at some point my patience ran dry.

I'm not in a position to help you backtest but if you have any questions feel free to ask.

All the Best

Member Since Oct 23, 2012   349 posts
Oct 18, 2017 at 12:18
When using AZ-Invest backtest scripts you must include the skip first tick on bar function on every tick or it will give you imaginary trades inbetween bars..
if you follow the flock like sheep you always end up stepping in shit!
Sign In / Sign Up to comment
You must be connected to Myfxbook in order to leave a comment
*Commercial use and spam will not be tolerated, and may result in account termination.
Tip: Posting an image/youtube url will automatically embed it in your post!
Tip: Type the @ sign to auto complete a username participating in this discussion.