The deal duration suggests same combo bot deal ran before and after the spike
Settings are covering +133% short deviation. Max funds of combo bot are 2.2k.
How does going up +500% cause a drawdown of -292$ only if we are using 2.2k in all.
It should be much more in my opinion. Thereby causing me to question the accuracy of the backtest
I’m not sure that I understand the problem. If he wondered why equity drawdown not so big, this is because we didn’t check it every candle, but only fixed number of times during backtest. Possibly that this spike wasnt in equity check candles.
We cannot check every candle, backtest a year on 1 minute and then he browser won’t be able to display that equity chart. So we decided to check only 100 times, so all charts have 100 points independently of how many bars.
As for liquidations, liquidationns checked each candle.
One more note, backtester work with isolated margin only, even if you will select cross in settings.
I think if you want the backtester to inform for futures strategies then those 100 independent points have to be top 100 the price lows or price highs depending on the direction of the strategy.
For spot strategies max funds used can be checked on top 100 lows unless they are checked for every candle currently.
Make a Feature Requests of it. Apparently once more it’s not-bug but a Feature Requests, even though an improved/corrected implementation isn’t only nice to have but quite essential.
The deals calculates their own drawdown on the lowest candle, that’s not an issue. What I’m saying is for displaying the equity curve we need a fixed number of data points that we will use for the graph. We decided on 100 becauw most browsers handle that can comfortably.