Hello,

I was exploring the accuracy of Mathematica's TimeSeriesResample[ ] function trying to transform 15 min financial time series into daily "BusinessDay" data time series.

Dealing with french stock data, I use this command :

tsr = TimeSeriesResample[ts, "BusinessDay", HolidayCalendar -> "France"]

Sadly, it seems that the french HolidayCalendar used by Mathematica may not be accurate.

Indeed, Euronext official Calendar of business days 2022 (https://www.euronext.com/en/trade/trading-hours-holidays) gives Friday 15 April 2022 and Monday 18 April 2022 as closed trading days. As such, my 15 min source time series "ts" doesn't contain any values nor timestamps for those dates, iterating from 14 April 2022 straight to 19 April 2022.

But "tsr" has those dates, creating interpolated values.

Is this behaviour normal ? Isn't the whole point of using "BusinessDay" specification to suppress datapoints from time series when market is closed ?

Going forward, I did read the official documentation for HolidayCalendar (https://reference.wolfram.com/language/ref/HolidayCalendar.html) which says that we can define a custom holiday schedule under the Scope section.

Does that mean that I can use a list of closed market dates of my creation and feed that list as argument in TimeSeriesResample function ? I did read the documentation's examples, but couldn't find a way to make it work at all.

Thank you in advance for any help you can give me.

Cheers.