Suddenly can't trust M* X-Ray (or Instant X-Ray.)

Crash

I made one switch on Friday: exited BHB in favor of BLX..... BHB was 5.9% of my total. Now, X-Ray tells me, suddenly, that I'm holding 16.49% of total portfolio in Latin America. Ya. I should not be surprised, I suppose. I liked to use that tool, but how useful is it to me, now? ORK!

....This is to say nothing about the fact that the tool recently cannot classify over one-third of my bonds. Even for an automated, unthinking tool, that's just hilarious. Because nothing has changed in the portfolio re: bonds in many months. DOUBLE ORK!

a2z

i suspect m* has long bought data from another source. (possible for everything other than in-house analysis reports, maybe their quant score)
if a holding is unclassified, they dont bother much to fill.
why not multiple sources? that means effort to reconcile.

i still rely on xray overlap to generally point me where any single equity may be over\under represented for my personal preference.

msf

I made one switch on Friday: exited BHB in favor of BLX..... BHB was 5.9% of my total. Now, X-Ray tells me, suddenly, that I'm holding 16.49% of total portfolio in Latin America.

To clarify your numbers: what was your Latin American holding percentage prior to the portfolio switch? If, for example, your portfolio had been 10% in Latin America prior to the switch, the new figure would not be out of line.

Please also confirm that you are reporting, both before and after change, percentages of total portfolio, not percentages of foreign holdings. I know you wrote "total portfolio" but I find it's easy to slip between those figures when looking at World Regions under x-ray interpreter. At least that's something I tend to do, especially since I'm interested in the regional breakdown of my foreign holdings (disregarding US).

the tool recently cannot classify over one-third of my bonds. Even for an automated, unthinking tool, that's just hilarious. Because nothing has changed in the portfolio re: bonds in many months

Again, a question of clarification. If you're talking about individual bonds, I've never seen M* have a problem classifying, so long as I input all the data columns: quantity, current price, credit quality, maturity, and coupon. You can see how each bond is classified (or not) by going to x-ray detail/bond style. If you've got a bond that shows up there as unclassified, that would be interesting. PM me if you'd like.

If you're talking about bond funds, then their holdings change upon each new fund report, even if your own portfolio doesn't change. Here too, the detail/bond style page can help identify the culprit(s). In my portfolio I see one bond fund that is 100% unclassified.

Morningstar relies upon funds themselves for bond data. "Morningstar asks fund companies to send ... information on a monthly or quarterly basis for each of their fixed-income and allocation funds."
https://advisor.morningstar.com/Enterprise/VTC/FI_Survey_Guidelines_cashRevision_FINAL.PDF

Not only raw data, but some calculated figures are provided by the funds in these fixed income surveys. For example, "Morningstar asks fund companies to calculate and send average effective duration (also known as option-adjusted duration) for each of their fixed-income and allocation funds."

I also vaguely recall that M* discards some fixed income data that's over a month old. I not having success in locating a statement to that effect now. But this could be another reason why the "unclassified" percentage increased.

Crash

-Before the switch into BLX, my Latam position was nil. My bonds are all in funds. TUHYX PRCPX WCPNX PRCFX PRWCX.

msf

Which percentage table are you looking at? If it's the world exposure table on the X-ray overview page, that's showing percentage of total stock. I ask in part because that would explain the apparent discrepancy, and in part because I'm having trouble finding an X-ray table that shows regional exposure as a percentage of total portfolio.

That would be nice to have (I'd like include foreign bonds in calculating foreign exposure by region).

In contrast, the tracking page shows individual securities as a percentage of the whole portfolio.

Suppose a portfolio is 1/3 stock and 2/3 bond. Then increasing the Latin American stock holdings by 16.5% of your equities would be the same as increasing your Latin American holding by 5.5% of your whole portfolio.

Percentage of total portfolio:
Latin America: 5.5%
Other stock: 27.8%
Bonds: 66.7%

Percentage of total stock:
Latin America = 5.5% / (27.8% + 5.5%) = 16.5%

I'll take a look at your bond funds.

Crash

@msf Ah, yes: total stocks, not total portf.
Here's more, and more particularly:

X-Ray shows:
Cash 6
Other 4
US stocks 41
Foreign 1
Bonds 49.
************
North America equities: 82.62
Latam 16.5
UK 0.17
Europe Dev. 0.14
Australiasia 0.56
******************
% of total assets:
PRWCX 39.8
TUHYX 20.4
PRCFX 13.61
PRCPX 7.83
ET 6.55
BLX 5.91
WBALX 4.34
WCPNX 1.93

Appreciate the attention given!

msf

FWIW, the culprit is TUHYX. Since most of these funds are from T. Rowe Price, and TRP seems to have provided data for all of them on the same date (M* reports portfolios as of Oct 31), I'm at a loss to explain why M* can't fully analyze this one fund.

Did you know that 1.07% of the TUHYX bonds are in Latin America

Crash