Error Live Portfolio Prices

Live Portfolios trading prices for 1st trading day of the week after 4/25/2022 are wrong.

Prices should be avg(Hi+Lo+2*close) of the 1st trading day of the week. But the prices are the closing prices from preceding Friday.

Must have something to do with the recent update of Portfolio 123 to accommodate multi-country data.

Are there dates other than May 30th where you’re seeing a problem? If so, please provide an example. We’re seeing a problem relating to how holidays are being handled for this feature.

Attached is an excel file of “All Transactions” where I picked up the discrepancy.
It is not just the holiday which causes this error.

P123_SimTran_iMTop5XLYSelectfrom2016.csv (18 KB)

Yes, this is an defect I also observed in live portfolio, seems to me not caused by recent upgrading but longer time ago. This issue causes live portfolio’s have better return than actual return due to previous Friday’s close price used. Anybody should be able to see this issue in live portfolio transaction.

If I’m looking at this correctly, trade pricing code broke on 4/11/22. The 4/4/22 prices look good.

For example, one of my live books shows TGT being sold on Monday 5/30/2022 (Memorial Day) with the 5/27/2022 Friday price of $167.14. But on Tuesday when the trade actually occurred the closing price was $161.88. So the portfolio shows better returns than actual returns.

This problem has arisen since the launch of the recent P123 upgrade to facilitate multi-country trading.

So we need to go back and check all Monday (if holiday, Tuesday) transactions since April 11? Does that date correspond to when the recent update was launched, Georg? With all the changes, I can’t make sense of when it was actually implemented.

Was there ever a P123 investigation and solution to this problem? We can’t go forward with Monday’s prices potentially being incorrect every week, can we? Would like to hear how this is being / or was resolved. Thank you.

We’re looking into it and hope to have a fix soon. Thanks for your patience and for calling the error to our attention.

Geov: The problem was happening because the portfolio used formula weights and was reconstituting without rebalancing. There was a bug where that case would not trigger updating the Buy price to avg(Hi+Lo+2*close). The fix is in production now, so you should not see that issue going forward.

The other issue Aaron mentioned is that the the automatic rebalance creates the Buy transaction on the holiday (May 30th) using the holiday date and price. This issue has been identified and will be fixed.

Other uses mentioned issues with the price in this thread. Please let us know if you saw any other issues with the automatic rebalance Buy prices other than the cases mentioned above.

The 4/11/22 transaction prices look wrong to me;

WLK: 2022/04/11 114.94 117.43 113.98 114.35 (117.43+113.98+114.352)/4=115.03
BX: 2022/04/11 111.78 112.74 109.99 111.43 (112.74+109.99+111.43

If my calculations are right, everyone’s 4/11/22 data should be bad. Can someone else check?

I don’t think your calculations are right.
WLK’s high on 4/11 was 117.70, low was 114.24, and close was 114.61. So that averages to 115.29.
BX’s high on 4/11 was 114.13, low was 111.35, and close was 112.80. That averages to 112.77.
Maybe you’re using adjusted prices?

Without realizing it, I was using the adjusted prices from

Unadjusted prices;
WLK: 2022/04/11 115.21 117.70 114.24 114.61 445,928 → (117.70+114.24+114.61*2)/4 = 115.29 = happiness

Thanks for the correction. All is well.