Execution mistakes in the backtesting:
1- The long entry is triggered when the RSI closes above 50, with an exit occurring
when the RSI rises above 70 and the next candle closes red.
2- The short entry is triggered when the RSI closes below 50, with an exit
occurring when the RSI drops below 30 and the next candle closes green.
3- The RSI period used to check the previous candle's strength varies with each
trade, leading to inconsistencies.
4-In some cases, entries and exits deviate entirely from the first two conditions,
with no clear patterns or rules.
5- In certain instances, the entries and exits follow the first two conditions, but the
third condition isn't satisfied at all.
6- In other cases, entries and exits fail to follow the first, second, or third
conditions altogether.
Note- These errors are found on the 20 trades specifically on ETH coin only.
Suggestions based upon the current backtesting by using the ML model-
1- Need to backtest the strategies by using Binance API.
2- As per my experience so far, it takes time to run the backtest without any error
as there is limit of hit or we can say API hit is limited. For this, we can download
the data and store it local system so the backtest will run smoothly and fast.
3- We can create pre-defined criteria for scanning the conditions (like candlestick
and chart pattern conditions or indicators) for backtesting.
4- There are so many things we can do and we need to come with things
according to the priority because creating new things will take and cost. We need
to discuss those things on a large scale.
