Zerodha Streak Issue: Backtesting Results are WRONG for Renko Chart for MIS Orders

Zerodha Streak shows WRONG backtesting results for Renko Chart for MIS Order type.

ISSUE: MIS order entry and exit done on different dates. MIS order Exit should be on the entry date only.

Now onwards i don’t know how to trust Streak Backtest results. Because of such thing the strategies that works on backtesting does not work on LIVE environment and retail traders losses money.

See the attached Pics for Reference:




@Streak can you.

Hi @Rajesh_Chauhan1 ,

This has been explained in the last paragraph here Charts - Streak Help

If renko brick size is large, and no bricks are formed after the 3:15pm(Square off time), currently the backtest is not able to find a brick/candle to square off on, which is why it gets carried forward.
Two approaches can be taken here, 1. Reduce the brick size and entry earlier during the day, 2. Backtest with CNC/NMRL.

If you need more assistance on using Renko charts in a certain manner or for requirement, please write to [email protected]

Ok, it means that Streak Team already knows that Renko Chart with MIS/Intraday results are WRONG.

And you people are still serving those results to you customers and when your customer points out mistake you will ask them to do CNC/Delivery testing.

Is it very hard technically to program your backtesting software to just squreoff at 3:15 LTP price for MIS testing?

This is not good Streak provide Renko Chart in the MOST Expensive Subscription Plan(Out of the 3 Available Plans) and it SAD to hear it is not TRUSTABLE.

@Rajesh_Chauhan1 ,

There are more complications than just squareoffing of at 3:15, it about figuring out the right price when the Renko brick is being formed without introducing a look ahead bias during backtesting(which can slip in if extreme care and attention are not paid to the designing of the backtesting system while maintaining the speed). Renko is a tricky chart type with a lot of caveats, we have also covered it in the webinar when it was released.

For most backtesting use cases it doesn’t affect the results in any manner, and also live and paper trading alerts are not affected and completely accurate as they don’t have to worry about any look ahead bais and just generates alert when the brick is formed.
You can also paper trade to see its more practical performance than just relying on backtest.

1 Like