AutoTracker Returns Off By 0.01%

[tsptalk]

Don't forget that the bond market closes at 3:00 PM ET, an hour earlier than stocks, so there is a little guesswork when there's late action in stocks.

 

[Boghie]

Could TSP and the AutoTracker be wrong on the F fund today? I was planning on gaining a few penneys today.

Handballer,

I noticed the same thing. I looked at the TSP Share Price page and found that the F Fund did decrease.

Friday: 13.5807
Monday: 13.5589

I don't know. Maybe the index we are tracking is not perfect - like the I/EAFE.

 

[Handballer]

Could TSP and the AutoTracker be wrong on the F fund today? I was planning on gaining a few penneys today.

 

[tsptalk]

+0.01% could be the winning return in 2010. :laugh:

 

[Boghie]

Tom,

The way things are going in the market now I might want that 0.01%:nuts:

Uuuuggggghhhhhhhhh...

 

[PessOptimist]

I used to watch Lost in Space as a young person becoming aware of females in a different light. Timmy and Lassie's Mom (June Lockhart) looked pretty good, especially in some of the views shown. Why were her breasts and cleavage suddenly so fascinating?

Tom, a lot of us have had to wrestle with spread sheets and database rounding errors in the past so good luck with that. My recollections go back to dBase and Lotus and having a math co-processor plugged in to the computer. I thought the problems were fixed for the most part by software upgrades but found that when we moved to Office 2007 a lot of our access databases and excel spreadsheets developed these problems again. I have no idea what programs are crunching your numbers and what those programs are based on but IIRC truncating the displayed numbers may be reducing the decimal places used in the calcs. I don't remember what the fix was. Of course going out to 22 decimal places will slow things down a lot and is unnecessary. using 4 decimal places and displaying 2 should be OK.

Not that you don't know all this.

 

[grandma]

Poolman,
On another note: the "Lost in Space" bumbling Dr. Smith is character actor Jonathan Harris. Here is an interesting link: http://www.imdb.com/title/tt0058824/
I used to always watch Lost in Space and would hope Dr. Smith would get eaten taken or left behind.

Thank you, nasa, for the link - I'm gonna take some time & watch the ones they have posted there for `full viewing.' These would be fun for my church library - course, I'd have to critique them first!

 

[nasa1974]

Good Job Tom.


Some people just don't get it and I'm not trying to bust your chops but If you where ever a computer programmer you know the deal from everyone using the program and especially from the CFO or the President. It's either right or wrong.
.

Poolman,

I really understand you where not trying "bust your chops" hence all the smilely faces. I do understand the underlying problems that can happen. I am not a programmer but just the issues I had playing around with my Excel spreadsheet, I do feel the pain so to speak.

On another note: the "Lost in Space" bumbling Dr. Smith is character actor Jonathan Harris. Here is an interesting link: http://www.imdb.com/title/tt0058824/
I used to always watch Lost in Space and would hope Dr. Smith would get eaten taken or left behind.

 

[tsptalk]

Some people just don't get it and I'm not trying to bust your chops but If you where ever a computer programmer you know the deal from everyone using the program and especially from the CFO or the President. It's either right or wrong.

Tom, It looks like you corrected the program and I commend you for that especially if you've never programmed before. It sorta got dropped in your lap. Do or Die and your doing a good job. 1 change can fix the program or screw up the works.

I actually do have a strong programming background, although I didn't use PHP until we started the premium service stuff, and the autotracker.

I don't think I could have taken this on without having that background. I did work a lot with Oracle database / SQL programming, so the MySQL stuff was easy to pick up. I actually enjoy messing with the database the most, rather than programming in PHP. Ocean's program is pretty complex. He thought of everything. He's the real programmer.

 

[poolman]

..that was the Dr /Mr Smith - civilian or some kind of a spy that I just hated cuz he ALWAYS was the cause of problems....deliberately.
-but as I said, we watched it whether the kids wanted to or not !!! :embarrest:

(-course, I could be remembering the name wrong! )and of you mean the actor's name - someone else will have to help out there...I got enuff problems with the characters' names!!!

No, I am not looking for real names. This is what I was looking for. "Dr /Mr Smith"

I couldn't stand him, lol. I hear what your saying having to watch the show though.

 

[grandma]

LOL, Wasn't that kid Will Robertson ? :cheesy:

I actually loved the show Lost In Space. So Cool
Who was that DimWit that always caused problems ?

..that was the Dr /Mr Smith - civilian or some kind of a spy that I just hated cuz he ALWAYS was the cause of problems....deliberately.
-but as I said, we watched it whether the kids wanted to or not !!! :embarrest:

(-course, I could be remembering the name wrong! )and of you mean the actor's name - someone else will have to help out there...I got enuff problems with the characters' names!!!

 

[poolman]

...kinda like starting out in space 0.01 degree off at base, I guess -
Thanx to both of you...
I'm not a Star Trekker & have no interest in being `lost in space.' ..even tho my kids grew up with that TV show shoved at them, cuz I loved to hate Mr Smith & that weird little kid - !!!!!

LOL, Wasn't that kid Will Robertson ? :cheesy:

I actually loved the show Lost In Space. So Cool

Who was that DimWit that always caused problems ?

 

[grandma]

...kinda like starting out in space 0.01 degree off at base, I guess -
Thanx to both of you...
I'm not a Star Trekker & have no interest in being `lost in space.' ..even tho my kids grew up with that TV show shoved at them, cuz I loved to hate Mr Smith & that weird little kid - !!!!!

 

[poolman]

Good Job Tom.

Folk's I used to be a programmer. Did the Crap for 14 years. I'm really not trying to make a Big deal out of this. In fact I could care less from the standpoint of People's returns being out of whack by .01

The Program either works accurately or it doesn't. The thing is the problem can grow exponentially throughout the year.

If this problem where happening with the top 10 individuals it would be more of a High alert than where the problem exists now.

Some people just don't get it and I'm not trying to bust your chops but If you where ever a computer programmer you know the deal from everyone using the program and especially from the CFO or the President. It's either right or wrong.

Tom, It looks like you corrected the program and I commend you for that especially if you've never programmed before. It sorta got dropped in your lap. Do or Die and your doing a good job. 1 change can fix the program or screw up the works.

 

[nasa1974]

Tom,

Thanks for all your hard work.

Are you kidding me! We have some folks complaining that they are .01% off. What are you afraid of, being at 145 instead of 144 on the tracker? Give me a break. :suspicious::blink::laugh::nuts:

OK! OK! I have stopped laughing. It's the principal and we are just keeping Tom aware of things before we have a major software issue.

Have a great evening everyone.

 

[tsptalk]

OK, a new approach. It seems to be working, so please let me know if you see anything "buggy".

 

[tsptalk]

Somehow the program is seeing (for instance) BamaAF's YTD Return as being 0.21% which is correct but is pulling from some Field somewhere 0.20%

So I think the problem is further along in the process than what you have given above. If you want to continue to find the problem then look to see where this 0.20% is being pulled from.

I went ahead with a test to display the return to 4 decimal places and it actually showed the correct number +0.2050) for those who now show 0.20%, so I'm not sure why when we drop back to 2 places, it doesn't round up like the others.

I decided to add 0.005% to each return, then truncate to 2 positions before displaying - kind of a manual rounding. The return now shows correctly, but a couple of other strange thinkgs happened with GTO and handballer's returns. Their returns look right, but they don't show up in the precise order they should.

For anyone interested, besides the rounding above, here's what I decided to do. I will now store the return to the 4th decimal place, so when it is selected it might be more precise in the order it is retrieved. We'll see if that works - but we won't see the results until Monday's returns are calculated. That is, unless I continue to play with it for the 1/22 returns date.

 

[grandma]

What?! Are you freak'n kidding me?! One hundredth of a percent! There goes my trip to the Caymans.
Thanks for the AutoTracker Tom and Ocean. Great feature for the message board. [/QUOTE] (emphasis mine..grandma)


My feelings, too ! & the Caymans has never been a goal of mine anyway- unless, of course, you can make my returns (for REAL, of course) to where I could even have a `hidden account' down there !!:laugh:

 

[poolman]

Some of you have noticed that your return is off by 0.01%. It is most noticeable for the accounts that have been 100% in the G fund all year.

It is actually not a big problem, and your account may actually be off by < 0.005% - and will likely seem correct after the next day's processing.

When we calculate the return, we subtract your current return from your return on 12/31/09, then divide by your return on 12/31/09. Balance and return are taken to the 4th decimal place during calculations.

Not everyone starts the year with the same balance since we rolled last year's balances into this year's when the new year started. New members always start with $10,000. Old members start the year with their 2009 ending balance.

So, depending on when you started or how you did in 2009, your starting balance might be different than someone else's. And that difference will show up on a particular day when a fund closes the day on a return near the XX.005% area. Some round up 0.001% and some don't.

Come the following day, it will likely look better. Some days it will be off - but never by more than 0.01%.

It actually looks correct in the database, but for some reason it does not display precisely on the report. If you notice when you sort by YTD return, some of the 0.20% returns are above the 0.21% returns. So, I believe the actual return that we have is correct, but what is displaying is off by .01%. I can't tell why exactly. I'd say rounding, but that doesn't make a lot of sense. I'll keep looking.

Tom,

I have done some calculations with my and BamaAF's account. With the information that you gave above and using the Member returns analysis as of: 2010-01-22. The calculations do work correctly according to my work.

Something is not working correctly for said individuals when running the program Members ranking and current allocation as of: 2010-01-22 (COB) Somehow the program is seeing (for instance) BamaAF's YTD Return as being 0.21% which is correct but is pulling from some Field somewhere 0.20%

So I think the problem is further along in the process than what you have given above. If you want to continue to find the problem then look to see where this 0.20% is being pulled from.

Anyway I am not trying to make a Big Deal out of this unless you want to find the problem.

 

[Show-me]

What?! Are you freak'n kidding me?! One hundredth of a percent! There goes my trip to the Caymans.

Thanks for the AutoTracker Tom and Ocean. Great feature for the message board.