Vanguard's awful "Portfolio Analysis" tool...

[nisiprius]

...isn't even accurate on Vanguard's own funds! This thing is an embarrassment to Vanguard.

I was using it for a quick check on my wife's account, which happens to consist almost entirely of Vanguard LifeStrategy Conservative Growth Fund (VSCGX).

Here is what the Portfolio Analysis tool reports for a hypothetical account that is 100% VSCGX and nothing else:


What's wrong with this picture? It's completely missing the "short-term reserves!" Here's what Vanguard's "Portfolio & Management" tab shows as the asset allocation for this fund:



These two pie charts are meaningfully different!

It's clear enough what's happening. VSCGX contains 20% Vanguard Short-Term Investment-Grade Fund (VFSTX). The official fund description for VSCGX evidently classifies VFSTX as "short-term reserves," while the Portfolio Analysis tool must be classifying it as "bonds."

Me, I don't know which way VFSTX should be classified. I do, however, feel strongly that Vanguard should classify it consistently!

(VFSTX has an average duration of 2.2 years. Vanguard gives it a 1 on its 1-5 risk scale, same as Short-Term Bond Index, and less than the 2 it gives Total Bond Market. Vanguard says VFSTX "Offers potential for more income than a money market fund. Appropriate choice for investors with short-term goals who can afford slight risk for the chance of higher returns.")

[dumbmoney]

To me "short term reserves" means a money market fund or the equivalent.

[Ron]

What does it look like on M* X-Ray?

Better yet (since I know you use RIP), run the Fido report and see what it shows, since RIP is interfaced to M* to calculate your actual holding breakdown, if you don't have X-Ray.

I don't know if VG uses M*'s data; I can only vouch for FIDO since I did some testing for them.

- Ron

[dognose]

I agree that Vanguard's portfolio analysis tool is a bad joke. This tool consistently makes mistakes in analyzing my portfolio. Vanguard needs to upgrade this monster or get rid of it entirely.

[nisiprius]

To me "short term reserves" means a money market fund or the equivalent.

Whatever. But then it's their official fund description that's wrong, and I assume they're being careful about that.

What does it look like on M* X-Ray?

This is interesting; Morningstar apparently splits the difference and counts short-term investment grade as if it were half cash, half bonds!



Curiouser and curiouser: an M* X-ray on VFTSX comes up 13% "cash," 87% "bonds."

[Sidney]

To me "short term reserves" means a money market fund or the equivalent.

Whatever. But then it's their official fund description that's wrong, and I assume they're being careful about that.

What does it look like on M* X-Ray?

This is interesting; Morningstar apparently splits the difference and counts short-term investment grade as if it were half cash, half bonds!



Curiouser and curiouser: an M* X-ray on VFTSX comes up 13% "cash," 87% "bonds."

If you look at the Portfolio and Holdings tab on VG it shows that 13.4% of the holdings are maturity of less than a year. That might be M*s cutoff for "cash."

[CABob]

It appears to me that the Vanguard tool does not properly handle funds of funds. It seems to look at the predominate fund holdings and consider that the entire fund has that characteristic. I use to own the LS Conservative Growth fund and experienced the same problem. I'm not sure how they handle target retirement funds.

[Beagler]

To addda insult to injury, the Portfolio analysis tool won't correctly analyze VG's own VA assets (eg, the % int'l included in the Balanced Fund). It's such a poor tool that long ago I created an Excel spreadsheet instead.

[Dry_Pillow]

How dare you guys question the almighty Vanguard. My 4th Degree BoogleKnights spit on you! For shame.

[sommerfeld]

...isn't even accurate on Vanguard's own funds! This thing is an embarrassment to Vanguard.
...
Me, I don't know which way VFSTX should be classified. I do, however, feel strongly that Vanguard should classify it consistently!

Have you reported this problem to Vanguard? I reported a separate problem with the portfolio analysis tool recently via the vanguard website "email" feature. The problem was fixed.

Software bugs which aren't reported to someone who can do something about them won't be fixed.

[Sunny Sarkar]

I have found the tool very useful in analyzing my simple 4-fund portfolio, and use it regularly for rebalancing purposes. It even incorporates funds I have with another company (401k). To me, the AA pie chart on this page is one of the two most important pieces of information available on the Vanguard website, the other one being net cash flow.

I think the reason it is not splitting up the Life Strategy fund is because the fund's AA is not static - and the tool probably depends on static (strategic) AA data which is not available for funds that implement tactical AA.

[nisiprius]

...isn't even accurate on Vanguard's own funds! This thing is an embarrassment to Vanguard.
...
Me, I don't know which way VFSTX should be classified. I do, however, feel strongly that Vanguard should classify it consistently!

Have you reported this problem to Vanguard? I reported a separate problem with the portfolio analysis tool recently via the vanguard website "email" feature. The problem was fixed.

Software bugs which aren't reported to someone who can do something about them won't be fixed.

OK, fair enough, I'll do it.

I have mixed feelings, though, because submitting a high-quality actionable bug report is not at all cost-free, and I have, in the past, wasted a reasonable amount of valuable time reporting bugs to organizations that didn't want to hear about them. In the seventies and eighties, I felt that reporting bugs was well worthwhile; recently, I've felt that most organizations are not interesting in hearing about bugs from anyone but their own SQA departments.

But I said I'd do it, and I will.

Do you know of any better channel than a regular "secure email" via Vanguard's message facility?

[sommerfeld]

recently, I've felt that most organizations are not interesting in hearing about bugs from anyone but their own SQA departments.

Indeed. Many front line support organizations seem to be skilled at keeping information about software defects away from the people who actually can do something about them..

Do you know of any better channel than a regular "secure email" via Vanguard's message facility?

Unfortunately, no.

(If it's any consolation I spent about two hours on the phone this spring with someone from Intuit getting them to reproduce a bug in how TurboTax handled the Massachusetts health care mandate in non-resident/part-year returns. The bug was fixed a couple weeks later).

[beardsworth]

I am shocked, shocked I tell you. I thought all these problems had become things of the past when we were informed that henceforth we would be Vanguarding.

[Sunny Sarkar]

At one point of time the Total International fund of funds was not being split properly in the tool. I reported it to Vanguard via email on the site, got a polite response, and it was eventually fixed.

[coffeehubcap]

Don't get me going on this -- I'm still bummed out with the new and improved statements !!!

[nisiprius]

I've reported it...

[nisiprius]

(If it's any consolation I spent about two hours on the phone this spring with someone from Intuit getting them to reproduce a bug in how TurboTax handled the Massachusetts health care mandate in non-resident/part-year returns.

Oh, man, oh man, oh man. Reminds me of a similar issue I had a few years ago.

The bug was fixed a couple weeks later).

Well, your two hours were well spent, then, and I extend to you my sincere congratulations. It's not just anyone that can get Intuit to act on a bug.

[Munir]

To me "short term reserves" means a money market fund or the equivalent.

Whatever. But then it's their official fund description that's wrong, and I assume they're being careful about that.

What does it look like on M* X-Ray?

This is interesting; Morningstar apparently splits the difference and counts short-term investment grade as if it were half cash, half bonds!



Curiouser and curiouser: an M* X-ray on VFTSX comes up 13% "cash," 87% "bonds."

Nisiprius,

Does this pie chart state that VFSTX has 40% equities?! I thought it only has bonds and cash or cash equivalents.

[richard]

The last time I played with the tool I noticed that a change in stock allocation decreased my holdings in value stocks according to the Investment Style Analysis section (which shows value / blend / growth) and increased by holdings in value stocks according to the 9 Box section (M* type v/b/g and small / medium / large).

[droliver]

I really wish Vanguard would liscence Morninstar's X-ray tool and just incorporate it into their GUI.

[Ed 2]

I really wish Vanguard would liscence Morninstar's X-ray tool and just incorporate it into their GUI.

But you will need to pay more for all these extra services for it [ ER], agree?

[nisiprius]

To me "short term reserves" means a money market fund or the equivalent.

Whatever. But then it's their official fund description that's wrong, and I assume they're being careful about that.

What does it look like on M* X-Ray?

This is interesting; Morningstar apparently splits the difference and counts short-term investment grade as if it were half cash, half bonds!



Curiouser and curiouser: an M* X-ray on VFTSX comes up 13% "cash," 87% "bonds."

Nisiprius,

Does this pie chart state that VFSTX has 40% equities?! I thought it only has bonds and cash or cash equivalents.

No, the pie chart is an X-ray of Vanguard LifeStrategy Conservative Growth, which contains 20% VFSTX. An X-ray of VFSTX, which I didn't show, indicates that VFSTX is 13% cash, 87% bonds, no stocks.

[blevine]

[quote="droliver"]I really wish Vanguard would liscence Morninstar's X-ray tool and just incorporate it into their GUI.[/quote]

etrade licenses m* tools, and I put in my portfolio there,
as well as Vanguard's tool. I find Vanguard's tool more intuitive and easier to use. M*'s give too much data, not enough information.

[blevine]

reported bugs in the interface on vg.com to Quicken and they fixed it fairly quickly.

On the other hand, a simple question as to the holdings of a fund, nothing to do with online tools, took MONTHS to answer....

[blevine]

Only includes VG funds, ignores your "outside holdings" even if they
are vanguard funds held elsewhere (I hold my ETFs at etrade but they are all VG funds).

[ncaraway]

Deleted by author

[SpringMan]

Today Vanguard's Portfolio Watch tool shows Sector allocations for the U.S. stock market as 0%.
Information economy
Sector Your Stock Portfolio U.S. Stock Market Difference From Market
Hardware 3.6% 0.0% 3.6
Media 0.6% 0.0% 0.6
Software 1.5% 0.0% 1.5
Telecommunications 1.8% 0.0% 1.8
Service economy
Sector Your Stock Portfolio U.S. Stock Market Difference From Market
Business services 2.1% 0.0% 2.1
Consumer services 6.2% 0.0% 6.2
Financial services 13.0% 0.0% 13.0
Healthcare services 13.3% 0.0% 13.3
Manufacturing economy
Sector Your Stock Portfolio U.S. Stock Market Difference From Market
Consumer goods 9.8% 0.0% 9.8
Energy 21.9% 0.0% 21.9
Industrial materials 5.7% 0.0% 5.7
Utilities 1.8% 0.0% 1.8
Uncategorized
Sector Your Stock Portfolio U.S. Stock Market Difference From Market
Other 18.7% 100.0% –81.3
Total 100.0% 100.0%

Looks like a brand new bug

edit: Reported to Vanguard using their interactive chat feature. It is being referred to the web support team.

[ruralavalon]

My personal gripe with Portfolio Watch -- the Portfolio Watch feature incorrectly characterizes Vanguard's Precious Metals and Mining fund (VGPMX) as a domestic allocation. Actually it is very predominately international (about 82% non-US).

Yes, I have just now reported this to them.

[Random Musings]

At the very least for Portfolio watch, Vanguard should have a strong handle with respect to categorizing all their funds and ETF's. It really is not that complicated and they have the data.

RM

[DartThrower]

Vanguard is working on a redesign of their website. Let's hope they are putting significant effort into addressing the problems that plague the current Portfolio Watch tool.

There will always be gray areas, but let's hope they can get the basics right. Their customers have a right to expect it.

[stratton]

If you really want to get confused on Morningstar look at the portfolio of Pimco funds. Their software can't handle the swaps.

Paul

[EyeDee]

Vanguard is working on a redesign of their website. Let's hope they are putting significant effort into addressing the problems that plague the current Portfolio Watch tool.

.
For those having problems with the Portfolio Watch tool I hope they do "improve" things, however, your comment that they are working on a redesign of their website has me concerned. I find Vanguard's web site to be one of the easiest fund company site for me to use. For example, I always have trouble finding what I am looking for on Fidelity's site.

I really hope they put their effort into fixing features such as the Portfolio Watch tool and leave the overall layout/feel of the site alone as it is one of the best, in my opinion.

[Doc]

If you really want to get confused on Morningstar look at the portfolio of Pimco funds. Their software can't handle the swaps.

Paul

Paul, please clarify your comment. It seem to me that the swaps represent long and short positions in some underlying FI security that mostly expire in less than a year and therefore are considered as "cash" instruments. M* reflects this in its portfolio view as a long/shot cash position. Now it might be confusing but I'm not convinced that the swaps aren't being properly "handled" by the software.

[nisiprius]

Sommerfeld, thank you for chivvying me.

I reported it via Vanguard's message system, under category "other," and received a reply saying that action will be taken:

Thank you for your e-mail and screenshots. We apologize for any difficulty you've experienced with the Portfolio Analysis tool.

We are investigating the classification information that you provided for us concerning the Portfolio Analysis tool. Once we are able to determine the cause of the inconsistencies within this tool, we will notify you by e-mail. Your patience in this matter is appreciated.

[ddb]

Out of curiosity, does anybody here actually use these Vanguard web tools to make investment decisions, or do you use them solely for hobby/entertainment purposes? I'd suggest that only the latter is appropriate.

I wouldn't trust an online tool to make decisions, even if the tool comes from the fund company that owns the funds that you input into the tool.

- DDB

[nisiprius]

Out of curiosity, does anybody here actually use these Vanguard web tools to make investment decisions, or do you use them solely for hobby/entertainment purposes? I'd suggest that only the latter is appropriate.

Well, let me put it this way. It wasn't a very important decision, but yes. Well, sort of.

The first time I ran Portfolio Analysis it gave me a red warning message about the risks of foreign stock investing, and said my foreign stock percentage was something like 23% or 24%. I was surprised, because I thought it was 20%, and 20% happened to be what I wanted.

With a little tinkering, I found that it was quite correct. I had been including only my actual foreign-stock funds (e.g. Vanguard Total International Index). But I was also holding Pax World Balanced Fund and had overlooked the fact that this fund had a significant foreign-stock allocation.

So I trimmed back to 20%. Of course before I took action, I did my own spreadsheet calculations. And of course it was all silly and inconsequential anyway.

But if you accept the premise that I wanted 20%, the tool alerted me to the fact that I didn't have 20%, and I did take action to tweak it to the exact magic arbitrary meaningless round number I'd decided on for no very good reason.

When I got it back to 20%, the Portfolio Analysis tool was happy, by the way, and quit giving me the red warning.

[ddb]

Out of curiosity, does anybody here actually use these Vanguard web tools to make investment decisions, or do you use them solely for hobby/entertainment purposes? I'd suggest that only the latter is appropriate.

Well, let me put it this way. It wasn't a very important decision, but yes. Well, sort of.

The first time I ran Portfolio Analysis it gave me a red warning message about the risks of foreign stock investing, and said my foreign stock percentage was something like 23% or 24%. I was surprised, because I thought it was 20%, and 20% happened to be what I wanted.

Thanks for sharing. I'm glad you confirmed the results yourself prior to taking action, but I do see that the tool is potentially useful for giving alerts. Then again, if the tool had incorrectly classified, say, the PAX fund is 100% domestic, then you wouldn't have known to take action and still would have been better off relying on your own spreadsheet analysis.

- DDB

[rustymutt]

I believe that it's right on. I keep 50/50 US/IntL and I must say that it agrees with my math. I'm not sure why others would be having any issues.

I must remind everyone that there is a truth in computer data bases,
bad information in, means bad information out. Could some of the issues be user related.

I'll sit back and wait for the backlash.

[kcyahoo]

I reported the Vanguard Portfolio Analysis problem years ago. They showed no interest in even understanding what I was telling them. I too created an Excel model to understand my true allocations.

[Sunny Sarkar]

Want better tools? Go to Fidelity.
Want lower fees? Go to Vanguard.
Simple

30 years later what will matter to me is all the money I saved from the lower fees (compounded), not what tools I used along the way. Hence I choose option 2.

I'll put up with anything that Vanguard throws at me (without getting me in trouble) as long as they keep their fees low.

Best,
Sunny

[Ron]

Want better tools? Go to Fidelity.
Want lower fees? Go to Vanguard.

That's why we use both. It dosen't have to be either/or, IMHO.

Our total fees are lower, as compared to just FIDO.
FIDO tools (which I use extensively) are much better than VG.

- Ron

[Doc]

Want better tools? Go to Fidelity.
Want lower fees? Go to Vanguard.

That's why we use both. It dosen't have to be either/or, IMHO.

Our total fees are lower, as compared to just FIDO.
FIDO tools (which I use extensively) are much better than VG.

- Ron

Kind of an aside. I was just comparing two funds and noticed that Vanguard had a 0.8 % e/r for a non-Vg fund. I know that the e/r is only 0.4% so I decided to check Fido. Fidelity's "research pages" are currently unavailable.

So with Fido I get high cost and no data and with Vanguard I get low cost and wrong data. Which to choose? Decisions, Decisions.

(BTW the e/r difference is due to Vg using prospectus info rather than what it actually is. Very misleading.)

[hollowcave2]

I agree that this problem should be reported.

The Life Strategy Conservative Growth fund has the Short Term Investment Grade fund as part of its holdings, and this is what's being listed as short term reserves, which is not accurate. The allocation showing 55% bonds is more representative of the actual holdings.

[ruralavalon]

Here is Vanguard's somewhat confusing response to my report to them mentioned above, concerning mis-characterization of VPGMX as a domestic rather than international holding.

Dear Mr. xxxxxx:

Thank you for your recent e-mail. We appreciate the time that you have
taken to contact us regarding our Portfolio Watch tool.

On 7/26/2010, you had sent us an e-mail requesting that we research an
issue with our Portfolio Watch tool and how it is characterizing the
Vanguard Precious Metals and Mining Fund (VGPMX) as a domestic allocation.
Our technical team has concluded their research on this issue. I have
provided the following information that will explain this issue:

Our Portfolio Watch tool categorizes funds based on their target
allocations. The Vanguard Precious Metals and Mining Fund is a domestic
stock fund, from a category standpoint. Our methodology is that domestic
stock funds are treated as 100% domestic in the Portfolio Watch tool, even
though the fund has 84-86% in International Funds. While we recognize that
this methodology is not really the best way to categorize this fund, the
tool is currently functioning as designed.

Our technical team is aware of this issue and is currently researching how
we can properly configure this tool to display a more accurate
categorization for funds such as this one. We do appreciate your feedback
on this issue. Your suggestion will be forwarded to our management team for
their review.

If you have additional questions, please call Vanguard Voyager Select
Services at 800-284-7245 and ask to speak with a web technical support
specialist. You can reach us on business days from 8 a.m. to 7 p.m.,
Eastern time.

[SpringMan]

It is bad that these problems exist but what is worse is that they won't fix them even when bug reports are submitted. Shame on Vanguard. If they were as sloppy on the rest of their website as they are in their portfolio analysis tool, I would leave them. It must not be the same computer programmers that handle both.

[sommerfeld]

It is bad that these problems exist but what is worse is that they won't fix them even when bug reports are submitted.

huh? what part of

Our technical team is aware of this issue and is currently researching how we can properly configure this tool to display a more accurate categorization for funds such as this one.

did you not understand? they're going to fix it but (for reasons they don't disclose) it's not a quick fix. If I had to guess, they likely need to rewrite how information percolates through the code and it's likely a bigger job than it appears.

I've been in software long enough to know that it's best to not rush bugfixes; better to wait for the correct fix.

[SpringMan]

It is bad that these problems exist but what is worse is that they won't fix them even when bug reports are submitted.

huh? what part of

Our technical team is aware of this issue and is currently researching how we can properly configure this tool to display a more accurate categorization for funds such as this one.

did you not understand? they're going to fix it but (for reasons they don't disclose) it's not a quick fix. If I had to guess, they likely need to rewrite how information percolates through the code and it's likely a bigger job than it appears.

I've been in software long enough to know that it's best to not rush bugfixes; better to wait for the correct fix.

I reported this bug on July 21 and it is still not fixed, see my earlier post. It remains broken and I will be surprised if it is fixed anytime soon.

Today Vanguard's Portfolio Watch tool shows Sector allocations for the U.S. stock market as 0%.
Information economy
Sector Your Stock Portfolio U.S. Stock Market Difference From Market
Hardware 3.6% 0.0% 3.6
Media 0.6% 0.0% 0.6
Software 1.5% 0.0% 1.5
Telecommunications 1.8% 0.0% 1.8
Service economy
Sector Your Stock Portfolio U.S. Stock Market Difference From Market
Business services 2.1% 0.0% 2.1
Consumer services 6.2% 0.0% 6.2
Financial services 13.0% 0.0% 13.0
Healthcare services 13.3% 0.0% 13.3
Manufacturing economy
Sector Your Stock Portfolio U.S. Stock Market Difference From Market
Consumer goods 9.8% 0.0% 9.8
Energy 21.9% 0.0% 21.9
Industrial materials 5.7% 0.0% 5.7
Utilities 1.8% 0.0% 1.8
Uncategorized
Sector Your Stock Portfolio U.S. Stock Market Difference From Market
Other 18.7% 100.0% –81.3
Total 100.0% 100.0%

Surely they know what sector percentages are in the US stock market, they had it right in earlier releases. My comment was not specifically directed at the precious metals and mining being 100% domestic which I did report several years ago by the way. By these problems, I was referring to the entire thread not one specific post. Also, I had a 40 year career in software development and no place I worked would tolerate their lackluster support.

[tibbitts]

It is bad that these problems exist but what is worse is that they won't fix them even when bug reports are submitted.

huh? what part of

Our technical team is aware of this issue and is currently researching how we can properly configure this tool to display a more accurate categorization for funds such as this one.

did you not understand? they're going to fix it but (for reasons they don't disclose) it's not a quick fix. If I had to guess, they likely need to rewrite how information percolates through the code and it's likely a bigger job than it appears.

I've been in software long enough to know that it's best to not rush bugfixes; better to wait for the correct fix.

Similar problems with this tool have been reported to VG many times over the past several years, and the response has always been a form of "we know and we're working on it." While it's ok for VG to simply not fix the problems, it should post a prominent disclaimer regarding the glaring inaccuracies. Clearly the tool misrepresents an account's actual allocation, enough to cause an unsuspecting person to believe (s)he has a dramatically different allocation than is actually the case.

One of my specific complaints in the past has been that my non-VG funds are correctly categorized, while my VG funds aren't. The last I checked that was still true.

Paul

[Ping Pong]

It's not so much a bug as it is a lack of desire to enter the thousands of holdings for each mutual fund into a database each quarter and then categorize all those holdings to get the appropriate percentages. It's a fairly labor intensive process when you figure there are thousands of funds out there and no standardized way of reporting holdings.

Though you would think this would be easy enough for Vanguard to do for their own funds. Perhaps they get the data for all funds from a third party, and the third party doesn't do anything special for Vanguard funds.