FinancialData return varying values for a specific day?

Questions
, ,
1

A simple query returns varying values. I assume Mathematica is pulling data from multiple data vendors and cycles through them, perhaps randomly.

Example:

1st Call:

FinancialData["bond", "Close", {{2015, 12, 31}, {2015, 12, 31}}]

{{{2015, 12, 31}, 102.52}}

8th Call:

FinancialData["bond", "Close", {{2015, 12, 31}, {2015, 12, 31}}]

{{{2015, 12, 31}, 102.703}}

10th Call:

FinancialData["bond", "Close", {{2015, 12, 31}, {2015, 12, 31}}]

{{{2015, 12, 31}, 102.52}}

If this is indeed how FinancialData works, one will never be able to achieve a repeatable result. Without repeatable results, this function is nearly useless for any sort of serious work.

Can anyone confirm that this is indeed what is going on here?

2

I can’t confirm that WRI is using multiple data sources (it would be very odd if they were), but I can confirm the behavior that you are seeing…

Counts@Table[FinancialData["bond", "Close", {{2015, 12, 31}, {2015, 12, 31}}], {100}]

<|{{{2015, 12, 31}, 102.703}} -> 27, {{{2015, 12, 31}, 102.52}} -> 73|>

Strange…

3

Thanks David!

4

“Close” for historical values is adjusted close and for “BOND” (PIMCO Total Return Active ETF) dates before {2016, 11, 1} are seeming to vacillate between two values from the source. This can be avoided by using “RawClose” for dates in that range.

FinancialData["bond", "RawClose", {{2015, 12, 31}}]

{{2015, 12, 31}, 104.22}

5

Yes it can…but raw close is just that, the raw closing stock price. If one wants the split and dividend corrected price, one needs to call “Close”. In this case, I need the corrected price. One can also purchase corrected end of day data, but that kind of defeats the purpose of this particular function.

Thanks for the suggestion Christopher! I appreciate the comment.

Cory

6

This seems to be unexpected behavior. Please submit this as product feedback at: