I’m working on a strategy with multi time frame conditions in zerodha streak.
Does the back tested results will differ from the live results.
I’m worrying because im frequently getting this notification"Multi-time frame can lead to unverifiable results due to partial candles"
And also the trigger which i got in live that are not showing in backtested results
Didn’t got your answer.
Can you elaborate ?
The response shared by you is correct @AnuragB.
The Multi-time frame (MTF) function by design checks your conditions on partial candles as the larger time frame candles take more time to be completed. Due to the use of partial candles, the Multi time frame results can be unverifiable on the charts.
Let’s consider an example of live deployment: If you want to enter a trade based on the following condition: “RSI higher than 50” in 5 min time frame. And you keep the base time frame as 1 min and the conditions are created inside the 5 min multi-time frame function as follows- “Multi time frame (5min, RSI(14,0) crosses above 50 )”
Now if we consider the scenario of the 9:15 5- minute candle during live trading. As you have used the base time frame of 1 minute your entry conditions will be checked after each 1-minute candle close. Suppose, if the 5 min RSI value goes above 50 at 9:16 itself (before the first 5 min candle close) then your entry will be triggered after the 9:16 1-min candle close. But if slowly the market goes down and by the end of the 5-minute candle say the RSI value has closed below 50. So on the chart, after 9:20 AM you would see that RSI is lower than 50. However, your entry has happened based on the earlier reading when RSI was above 50 on a partial candle, and ultimately since the market has gone negative the signal will be unverifiable in the chart later.
In the above scenario, the backtest will not consider this entry as in the data available for the backtest the RSI is lower than 50 at the 5-minute candle close. Thus accounting for lookahead bias in backtesting. Hence we have also added the disclaimer highlighting the same on the Create page. You can click on the below link to learn more about this:
https://help.streak.tech/create_advanced/#multi-timeframe-functions
If you want to avoid the look-ahead bias in the backtest and get verifiable results in the live trades, you can use the Multi-time frame completed function of Streak, instead of the multi-timeframe function. Multi-time frame completed function by design checks the conditions on the completed candle of the larger time frame.
ok got your point,
So how much difference can be seen in live trading with the same logics?
And how much can I rely on this back tested result to live trade