Hm, I found another issue in my code. I'll try to sort that out first to see if it's related somehow.

Sorry, here is the code.

self.cerebro.addobserver(bt.observers.BuySell)
self.cerebro.addobserver(bt.observers.Value)
self.cerebro.addobserver(bt.observers.TimeReturn, timeframe=bt.TimeFrame.NoTimeFrame)
self.cerebro.addobserver(bt.observers.Benchmark, data=self.data_benchmark, timeframe=bt.TimeFrame.NoTimeFrame)

If I activate cheat-on-close the difference is even bigger.
self.cerebro.broker.set_coc(coc=True)

Hi everyone,

I am a novice regarding backtrader and also python but it is great fun to learn and use the platform. Great work by all involved!

Today I tried the Benchmark observer and noticed that the return from the strategy is different than when using the TimeReturn observer. Is this a bug (missing negative sign for return of strategy in benchmark observer) or is it possible that I am missing something. See attached plot image.

Thanks in advance!

Best regards,
Torbjörn