SAMEPERIODLASTYEAR(DATESYTD( )) different from DATESYTD(SAMEPERIODLASTYEAR( ))??

[AlB]

Hi all,

This came up as I was trying to answer this post by @hbolo (no need to read it to answer this question). Here is the pbix

We have a Date table with dates between 01/01/2018 and today (Jan 3rd, 2019).

We place Date[Date] (down to the day level) in the rows of a matrix visual and these two measures in values:

YTDPreviousYearRevenue =
CALCULATE (
    SUM ( Revenue[Revenue] );
    SAMEPERIODLASTYEAR ( DATESYTD ( 'Date Table'[Date] ) )
) 

YTDPreviousYearRevenue_2 =
CALCULATE (
    SUM ( Revenue[Revenue] );
    DATESYTD ( SAMEPERIODLASTYEAR ( 'Date Table'[Date] ) )
)

where we are simply calculating the YTD revenue for the previous year.

For January 3rd, 2019 (the last day on the Date table),
[YTDPreviousYearRevenue_ 2] yields the correct result (i.e. the revenue for 1-3 Jan 2018).

However, [YTDPreviousYearRevenue] yields an incorrect result: the revenue for 1-31 Jan 2018. [YTDPreviousYearRevenue] is thus considering the full month of January 2018 while the YTD of 2019 is only 1-3 Jan, the first 3 days I would expect both measures to yield the exact same result.

Note that [YTDPreviousYearRevenue] yields the correct result for the 1st and 2nd Jan 2019. The only issue seems to be on the 3rd Jan 2019, the last day on the Date table.

Does anybody know what is going on?

Many thanks for your help
 
Note: I suspect it has to do with the Date table finishing on Jan 3rd, 2019 but do not know why. If the Date table is extended to include dates up to the end of 2019 this problem disappears.