Wellesley and Wellington show huge losses

[target2019]

It was a Vanglitch. Saw it in my tiny mobile screen this morning.

 

[RunningBum]

Originally Posted by sengsational
It could be corruption of data, but I think more likely a process problem where asset prices were left initialized (ie $0).

If the data feed for pricing fails, of course it should stop, but maybe the guy that got woken up every night on false alarms decided to bypass that bit of code because that process "always works".


That sounds plausible, as Wellesley "crashed" harder than Wellington, and the former had a higher bond AA than the latter.

Yes, and they were down by approximately the amount they have allocated in bonds. And I noted in another post that the bond fund closing prices hadn't been posted yet.

This shouldn't have been too hard for VG to explain. They had a problem with bond price reporting, and a second problem that they reported the mixed fund prices with only one component updated. It shouldn't be hard to put a sanity check in not to report a price if you haven't updated all of the components, like bonds showing $0.

VG is really slow today. Probably everyone logging in to check that everything is ok. Understandable delays but not reassuring.

 

[JoeWras]

Got logged in after a few minutes. Clicked on my portfolio and got this:

The server encountered an internal error or misconfiguration and was unable to complete your request.

​

Like someone else said, glad I'm not working. I pity the poor IT people working at Vanguard right now.

This kind of thing better not become a VG habit.

 

[sengsational]

This will be good for Vanguard's team to identify bottlenecks in the systems used to service web requests. Kind of a natural experiment of super high demand on their systems. I hope they use the opportunity to add appropriate components based on what is slowing things down today.

 

[JoeWras]

This will be good for Vanguard's team to identify bottlenecks in the systems used to service web requests. Kind of a natural experiment of super high demand on their systems. I hope they use the opportunity to add appropriate components based on what is slowing things down today.

You should come out of retirement and get into Customer Experience Management for VG. I hear Valley Forge is nice in the winter.

 

[target2019]

You should come out of retirement and get into Customer Experience Management for VG. I hear Valley Forge is nice in the winter.

Please do not wear a red coat.

 

[calmloki]

The conspiracy game is weak on this board.

The opposition doesn't need to hit us with a massive EMP and wipe out our records.
They don't need to infect our systems with a nasty virus that scrambles the records and causes billions to vanish.

All they need to do is tip the scales a bit by sowing distrust of the stability of the system, slowing or reducing American investment in American companies.

I've no understanding, but have decided that betting on our country beats burying silver bars and hiding in the woods.

 

[REWahoo]

Just for grins...

If the problem was caused by pricing the bond allocation at $0, which it appears was the case, then the "glitch" numbers for Wellesley and Wellington should have been only the equity portion of the funds. Dividing those numbers by the corrected fund values show the Wellesley AA to be 44/56 and Wellington 68.5/31.5.

While the Wellington numbers are very close to the fund's advertised 65/35 target AA, the AA for Wellesley is a little more equity heavy than the target 35/65 I would have expected to see.

 

[RunningBum]

Just for grins...

If the problem was caused by pricing the bond allocation at $0, which it appears was the case, then the "glitch" numbers for Wellesley and Wellington should have been only the equity portion of the funds. Dividing those numbers by the corrected fund values show the Wellesley AA to be 44/56 and Wellington 68.5/31.5.

While the Wellington numbers are very close to the fund's advertised 65/35 target AA, the AA for Wellesley is a little more equity heavy than the target 35/65 I would have expected to see.

Wellesley asset allocation as of 6/30/2019
https://investor.vanguard.com/mutual-funds/profile/portfolio/vwinx

Note that things can change in between quarters. It also would not surprise me if the fund manager went a little higher into stocks to try to boost return, and then settle back closer to the 36/65 target at the end of the quarter for reporting.

Asset allocation
as of 06/30/2019



37.65% Stocks
59.25% Bonds
3.10% Short-term reserves

 

[Go-NoGo]

Just for grins...

If the problem was caused by pricing the bond allocation at $0, which it appears was the case, then the "glitch" numbers for Wellesley and Wellington should have been only the equity portion of the funds. Dividing those numbers by the corrected fund values show the Wellesley AA to be 44/56 and Wellington 68.5/31.5.
35/65 I would have expected to see.

According to post in the Boglehead forum the problem was caused by "..The consolidated tape association which processes stock prices for roughly one third of the stocks listed in the United States, which had a major technical issue this afternoon into and past the close"


https://www.ctaplan.com/alerts#110000144324

 

[JoeWras]

According to post in the Boglehead forum the problem was caused by "..The consolidated tape association which processes stock prices for roughly one third of the stocks listed in the United States, which had a major technical issue this afternoon into and past the close"


https://www.ctaplan.com/alerts#110000144324

That may be true, but the problem with their website this morning shows serious load related issues. That's a different problem and potentially quite serious.

 

[Go-NoGo]

That may be true, but the problem with their website this morning shows serious load related issues. That's a different problem and potentially quite serious.

I agree, if the data provider had problems they should have a process in place to flag prices that are outliers and an alternate data provider to get correct end of day prices.

Re: logging in this morning - it just shows that they do not have the capacity handle large loads. I'd hate to think of what would happen if the market drops 10% in day. They should just go to AWS or Azure and get capacity on demand.

 

[Spock]

...
Re: logging in this morning - it just shows that they do not have the capacity handle large loads. I'd hate to think of what would happen if the market drops 10% in day. They should just go to AWS or Azure and get capacity on demand.

The last patent I was awarded was in secure cloud data storage.
If I knew my brokerage was using public cloud compute resources I would close the account immediately.

 

[NW-Bound]

... Dividing those numbers by the corrected fund values show the Wellesley AA to be 44/56 and Wellington 68.5/31.5.

While the Wellington numbers are very close to the fund's advertised 65/35 target AA, the AA for Wellesley is a little more equity heavy than the target 35/65 I would have expected to see.

Or is it possible that Wellesley is sitting on a bit of cash, or short-term instruments that are not affected by the bug?


About people not being able to log into Vanguard, I suspect that could be caused by their computer having a sudden higher than normal load. Vanguard account holders are not known as day traders, and they may not even log in at the end of every day to check their accounts. Now, everyone is trying to get it, and the door is not wide enough.

I myself rarely log into my Vanguard accounts directly, and rely on Quicken to download share prices and distributions. And I have not bothered to log in again since I logged in early yesterday and saw the wrong prices.

On the other hand, I log on to my other brokerage accounts daily to look for an opportunity to trade. These guys make a lot of money from transaction fees.

My brother who worked in IT at Schwab told me that a few times in the past when their computers were down or were targeted by hackers in a Denial-of-Service attack, they lost mucho money that day. And same as a hotel having empty rooms at night, they do not make up that money the next day. The opportunity is lost.

 

[Mr._Graybeard]

According to post in the Boglehead forum the problem was caused by "..The consolidated tape association which processes stock prices for roughly one third of the stocks listed in the United States, which had a major technical issue this afternoon into and past the close"


https://www.ctaplan.com/alerts#110000144324

I was ready to lay the blame on summer interns.

 

[GenXguy]

I noticed some odds things with my Vanguard balance last evening. I was down $36,000, then later it was showing the balance from last week, and then this morning, my balance was down $26,000. Well, better than $36,000! LOL. I should get some of that back today.

 

[Jackanory]

I was inspecting my Fido paper statement this weekend, as I consider my rebalance, and noticed that my VFIAX (Vanguard 500 index) beginning position for July 1st was unavailable! Ending balance was there, so all that was off was my monthly gain. I'd assumed that Fido tracked that amount, but maybe they get it from Vanguard, and had similar difficulties.

 

[pb4uski]

Yikes - add this to my absolutely horrible experience with VG over the past several days on the joint beneficiary issue and it's enough to make me bail on these guys once and for all.

Unfortunately, you can't buy VG Admiral funds at any other brokerage (Fido, TD, etc) so if you want to own VWIAX or other similar funds you pretty much have to put up with VG's total incompetence when it comes to systems, processes and customer service.

Yeah, but you're making a mountain out of a molehill on the joint beneficiary issue.... there is an easy solution that you don't want to do... so no sympathy.

 

[SecondCor521]

It could be corruption of data, but I think more likely a process problem where asset prices were left initialized (ie $0).

If the data feed for pricing fails, of course it should stop, but maybe the guy that got woken up every night on false alarms decided to bypass that bit of code because that process "always works".

Having written production firmware code for two different Fortune 500 companies, I'll just say that it is quite a bit harder to write (and test) the error cases than it is for the run-of-the-mill everything-is-working case.

It's also hard to maintain software systems which have the production demands like the US stock market. It's a real time system which needs to have pretty much 100% up-time, near instantaneous response times, incoming and outgoing data from hundreds or thousands of other computer systems. And, I'm sure they have minor changes going on all the time that need to be seamlessly integrated without breaking anything else.

As far as Vanguard's site being slow, I would not expect them to build a system that can handle 10x or 100x a normal load and still maintain typical response times. That would be a waste of money in my view.

 

[24601NoMore]

Yeah, but you're making a mountain out of a molehill on the joint beneficiary issue.... there is an easy solution that you don't want to do... so no sympathy.

And what, pray tell, is that "easy solution"? I don't have time for a lawyer to create a trust or a will before travel - and don't want to break "our" money out into separate "his" and "hers" accounts as that's not what it is.

What solution would you propose? Far as I can tell, there is no solution that's even remotely acceptable / workable.

ETA - I also don't think it's a "mountain out of a molehill" as EVERY OTHER large financial institution and even most small ones (eg: CUs) support beneficiaries on joint accounts. I should not have to change my entire account structure around and divvy up money between "his" and "hers", which it is not, because VG is obstinate - and without a single "good" reason that I've heard from them.

 

[gwraigty]

And what, pray tell, is that "easy solution"? I don't have time for a lawyer to create a trust or a will before travel - and don't want to break "our" money out into separate "his" and "hers" accounts as that's not what it is.

What solution would you propose? Far as I can tell, there is no solution that's even remotely acceptable / workable.

Do the DIY will, not naming the specific Vanguard account. Include the account in a list of supplemental info. Enjoy your travel. When you return well rested, give the matter some serious thought. Consider transferring assets to another brokerage. There are suitable ETFs, many commission-free. Or, pay a small $4.95 commission to buy Vanguard ETFs.

But whatever causes you the least anxiety is the right solution for you.

 

[pb4uski]

And what, pray tell, is that "easy solution"? I don't have time for a lawyer to create a trust or a will before travel - and don't want to break "our" money out into separate "his" and "hers" accounts as that's not what it is. ...

You just said it... you don't "want" to separate the joint account into separate his and hers accounts.... that would be an easy solution... you each name your spouse as primary beneficiary and as contingent beneficiaries those who you would name if they allowed joint beneficiaries.

If one of you dies then combine the accounts into one... if you both die then the contingent beneficiaries get the money as you want. I don't see any downside other than you don't want to do it.

... and without a single "good" reason that I've heard from them.

I hypothetized why they would rather not do it in that other thread.

 

[JoeWras]

Having written production firmware code for two different Fortune 500 companies, I'll just say that it is quite a bit harder to write (and test) the error cases than it is for the run-of-the-mill everything-is-working case.

It's also hard to maintain software systems which have the production demands like the US stock market. It's a real time system which needs to have pretty much 100% up-time, near instantaneous response times, incoming and outgoing data from hundreds or thousands of other computer systems. And, I'm sure they have minor changes going on all the time that need to be seamlessly integrated without breaking anything else.

As far as Vanguard's site being slow, I would not expect them to build a system that can handle 10x or 100x a normal load and still maintain typical response times. That would be a waste of money in my view.

But I'd expect better behavior than a simple "server got an error" message. Users shouldn't even know or care what a "server" is. I'd also expect that error to come up quickly. If they are having back end issues, throw everyone off to a page explaining the issue. In the rare cases where Amazon has had issues, they do this with a nice, calming web page. Not just a white line of type.

Quality code typically has 50% or more of error cases. It is hard and costly to write such good code.

It is part of my issue with a lot of consumer goods these days. They clearly don't care about that stuff.

 

[RetiredAt55.5]

It's also hard to maintain software systems which have the production demands like the US stock market. It's a real time system which needs to have pretty much 100% up-time, near instantaneous response times, incoming and outgoing data from hundreds or thousands of other computer systems. And, I'm sure they have minor changes going on all the time that need to be seamlessly integrated without breaking anything else.

As far as Vanguard's site being slow, I would not expect them to build a system that can handle 10x or 100x a normal load and still maintain typical response times. That would be a waste of money in my view.

Actually, they do load test their software for high demands - they've been doing that for at least the last 20 years. They have an entirely separate "production like" system that they load up prior to every major software release, to test that it can support high logins and the highest accessed pages and transactions.

You're 100% right, they are making minor changes all the time. And of course a developer will swear on a stack of bibles that their change will have no impact on the health of the system. Been there, done that, and have sworn on that bible lots of time (usually right, but not always .... ).

It's a constant battle to monitor this. And you're right, it's not cheap - either the separate hardware/software licenses/dedicated test team/etc.

Vanguard knows what they're doing. It is a trade off on how much money/resources to devote to this though. Everyone wants the expense ratios to be almost $0.

 

[JoeWras]

DW worked in IT in the finance industry. This problem is giving her flashbacks. Their management went through huge pains to avoid any kind of mismatch of account balances as presented to the customers. It was a huge, huge worry.