You should then also have an additional minute or tick based price history for a realistic backtest. Brokers won't fill your trades at Renko boundaries.
Thanks for the response, but I'm confused by your suggestion that brokers won't fill a trade at a Renko boundary. Provided there is volume in the bar, how is it any different from the boundary of a minute bar?
Both are abstract buckets for a stream of ticks. I can't see how an arbitrary time boundary is any different from an arbitrary price boundary in terms of getting filled.
Obviously if the Renko bar is in the middle of a gap and has no volume, you'll see slippage. But that is trivial to allow for - you simply skip the bar and don't run your exit rules against it. A time-based bar could also close in the middle of a gap, though admittedly this is less likely.
In the end, it's obviously a good idea to do a final check at tick resolution, but that is slow, even in Zorro. For fast iteration of ideas we have found Renko bars as reliable as anything else on our previous platform.
If a strategy is so marginal that small inaccuracies in slippage are critical, it's probably not tradeable anyway. You could say that of any bar type, given that we can't rely on 100% data quality.