QuantConnect Community

Hi Pier,

This can be achieved by creating an indicator object not tied to the algorithm (e.g. using self.sma = SimpleMovingAverage(...) instead of self.sma = self.SMA(...)) and manually updating the indicator under the “Manually Updating” section here. Also, since the indicator needs to be updated data a few minutes before market close, we need to change our resolution to Resolution.Minute.

I’ve shown an implementation in the attached backtest.

Best,
Shile Wen

Thanks Shile

by my understanding this sma indicator has a minute resolution, so it's not going to calculate a 20 day moving average and use the last value to precalculate the sma

is it correct?

My aim is trying to precalculate the value in order to anticipate the entering in the market , but it's important to calculate the average on the last 19 close plus the latest 10 min to close" value and do this everyday..

 when I use a manual indicator, will it calculate sma on the "10 min to close" value or will it use the correct close value in the following days?

I hope my question is clear .. it's a bit complicated to explain

Hi Pier,

Because we use a Scheduled Event that executes once per day, it is very similar to being on a Daily Resolution, the only difference being the close value used for the computation uses the price 10 minutes before Market Close as opposed to the price at Market Close.

Please note the solution will not use a LEAN indicator, and other indicators using similar data will be more difficult to implement.

To achieve “the average on the last 19 close plus the latest 10 min to close value and do this everyday”, we can use a RollingWindow of the past 19 market closing prices, sum these prices, then add on the price 10 minutes before Market Close of the current day, then divide the result by 20. I’ve shown how to do this in the attached backtest.

Best,
Shile Wen