How to get 10 year cumulative return on TSP account

Have a question about your personal investments? No matter how simple or complex, you can ask it here.
Post Reply
Topic Author
Jimsad
Posts: 245
Joined: Mon Mar 20, 2017 5:54 pm

How to get 10 year cumulative return on TSP account

Post by Jimsad »

Hi .
On the TSP account , only one year performance return is reported .
We wanted to find out the 10 year cumulative return just like the way vanguard reports on their accounts .
When we called , the customer service said they do not report that . Is there a way to get this ?
User avatar
HueyLD
Posts: 7668
Joined: Mon Jan 14, 2008 10:30 am

Re: How to get 10 year cumulative return on TSP account

Post by HueyLD »

https://www.tsp.gov/PDF/formspubs/oc05-16w.pdf

This TSP publication shows you how to calculate multi-year cumulative return.
rascott
Posts: 2293
Joined: Wed Apr 15, 2015 10:53 am

Re: How to get 10 year cumulative return on TSP account

Post by rascott »

Rather ridiculous they don't just report this, rather than giving people a math problem to calculate. Seems writing the paper would have taken as much time as just doing it.
Topic Author
Jimsad
Posts: 245
Joined: Mon Mar 20, 2017 5:54 pm

Re: How to get 10 year cumulative return on TSP account

Post by Jimsad »

HueyLD wrote: Fri Sep 13, 2019 9:39 pm https://www.tsp.gov/PDF/formspubs/oc05-16w.pdf

This TSP publication shows you how to calculate multi-year cumulative return.
Thanks for your reply .
Looks complicated And I am not good at Math 🙁
But will give it a try with my high school kids help !
User avatar
vineviz
Posts: 7520
Joined: Tue May 15, 2018 1:55 pm

Re: How to get 10 year cumulative return on TSP account

Post by vineviz »

rascott wrote: Fri Sep 13, 2019 11:58 pm Rather ridiculous they don't just report this, rather than giving people a math problem to calculate. Seems writing the paper would have taken as much time as just doing it.
Perhaps, but it’s valuable for investors to know how to compound a return.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
HueyLD
Posts: 7668
Joined: Mon Jan 14, 2008 10:30 am

Re: How to get 10 year cumulative return on TSP account

Post by HueyLD »

Jimsad wrote: Sat Sep 14, 2019 6:15 am
HueyLD wrote: Fri Sep 13, 2019 9:39 pm https://www.tsp.gov/PDF/formspubs/oc05-16w.pdf

This TSP publication shows you how to calculate multi-year cumulative return.
Thanks for your reply .
Looks complicated And I am not good at Math 🙁
But will give it a try with my high school kids help !
You don't need to be good at math. You just need to set up a very simple Excel spreadsheet and let the spreadsheet do the math for you.

I posted that TSP file as a public service because it is a valuable knowledge for any investor and it is very simple to set up in Excel. I am so pleased that the TSP wants to show us how to fish instead of just feeding us fish.
rascott
Posts: 2293
Joined: Wed Apr 15, 2015 10:53 am

Re: How to get 10 year cumulative return on TSP account

Post by rascott »

vineviz wrote: Sat Sep 14, 2019 6:16 am
rascott wrote: Fri Sep 13, 2019 11:58 pm Rather ridiculous they don't just report this, rather than giving people a math problem to calculate. Seems writing the paper would have taken as much time as just doing it.
Perhaps, but it’s valuable for investors to know how to compound a return.

Yes it is, and BHs likely enjoy doing it. However, I imagine the number of Fed employees who have ever done so is sub 1%.
Topic Author
Jimsad
Posts: 245
Joined: Mon Mar 20, 2017 5:54 pm

Re: How to get 10 year cumulative return on TSP account

Post by Jimsad »

HueyLD wrote: Sat Sep 14, 2019 6:31 am
Jimsad wrote: Sat Sep 14, 2019 6:15 am
HueyLD wrote: Fri Sep 13, 2019 9:39 pm https://www.tsp.gov/PDF/formspubs/oc05-16w.pdf

This TSP publication shows you how to calculate multi-year cumulative return.
Thanks for your reply .
Looks complicated And I am not good at Math 🙁
But will give it a try with my high school kids help !
You don't need to be good at math. You just need to set up a very simple Excel spreadsheet and let the spreadsheet do the math for you.

I posted that TSP file as a public service because it is a valuable knowledge for any investor and it is very simple to set up in Excel. I am so pleased that the TSP wants to show us how to fish instead of just feeding us fish.
Not sure I fully agree with this .
So if somebody does not know how to do spread sheet , should not be investing ?
Topic Author
Jimsad
Posts: 245
Joined: Mon Mar 20, 2017 5:54 pm

Re: How to get 10 year cumulative return on TSP account

Post by Jimsad »

If vanguard does it , why not TSP?
They should make it as easy as possible for investors to look at their returns .
Not everybody has the time (non retired folks) and not everybody enjoys spread sheets and math calculations
dbr
Posts: 33638
Joined: Sun Mar 04, 2007 9:50 am

Re: How to get 10 year cumulative return on TSP account

Post by dbr »

Seriously, as much as it is a good thing for people to have some math knowledge of return and compounding, issuing a textbook instead of just running the data is absurd.*

*I first learned how to compute internal rate of return for a portfolio from an excellent essay in the notes to my 401k plan. But the difference is that they computed the number and then provided a reference to how they do it rather than just sending a person to the reference.

**They also make the standard mistake of calling the result the geometric average of the annual returns when it is actually the compound annual growth rate (CAGR). See here https://en.wikipedia.org/wiki/Geometric_mean and here: https://en.wikipedia.org/wiki/Compound_ ... rowth_rate
jsprag
Posts: 201
Joined: Sun Nov 26, 2017 11:25 am

Re: How to get 10 year cumulative return on TSP account

Post by jsprag »

Jimsad wrote: Fri Sep 13, 2019 5:46 pm Hi .
On the TSP account , only one year performance return is reported .
We wanted to find out the 10 year cumulative return just like the way vanguard reports on their accounts .
When we called , the customer service said they do not report that . Is there a way to get this ?
My annual statements have 1, 3 and 5 year performance figures for the individual funds (G, F, C, S, I), all the lifecycle funds, and my personal time-weighted return over those same periods. They may not do 10 year, but they're calculating more than just annual.

If you're looking for 10 year performance on individual funds (not your personal account) then those are here.
User avatar
vineviz
Posts: 7520
Joined: Tue May 15, 2018 1:55 pm

Re: How to get 10 year cumulative return on TSP account

Post by vineviz »

dbr wrote: Sat Sep 14, 2019 9:32 am **They also make the standard mistake of calling the result the geometric average of the annual returns when it is actually the compound annual growth rate (CAGR). See here https://en.wikipedia.org/wiki/Geometric_mean and here: https://en.wikipedia.org/wiki/Compound_ ... rowth_rate
In this use case, geometric mean and CAGR are identical and interchangeable.

More generally, CAGR is equivalent to geometric mean in the absence of internal cash flows. In other words, CAGR is the time-weighted geometric mean.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
dbr
Posts: 33638
Joined: Sun Mar 04, 2007 9:50 am

Re: How to get 10 year cumulative return on TSP account

Post by dbr »

vineviz wrote: Sat Sep 14, 2019 1:45 pm
dbr wrote: Sat Sep 14, 2019 9:32 am **They also make the standard mistake of calling the result the geometric average of the annual returns when it is actually the compound annual growth rate (CAGR). See here https://en.wikipedia.org/wiki/Geometric_mean and here: https://en.wikipedia.org/wiki/Compound_ ... rowth_rate
In this use case, geometric mean and CAGR are identical and interchangeable.

More generally, CAGR is equivalent to geometric mean in the absence of internal cash flows. In other words, CAGR is the time-weighted geometric mean.
No. See this article for the f-mean where f for the geometric mean is log(r) and f for the CAGR is log(1+r): https://en.wikipedia.org/wiki/Generaliz ... zed_f-mean or relationship with logarithms in https://en.wikipedia.org/wiki/Geometric_mean The funciton f for the arithmetic mean is the identity, for the harmonic mean is 1/r, for the quadratic mean is r^2 (used with risk assuming no correlation) and so on.

It is correct that there are the additional concepts of internal rate of return, dollar weighted vs time weighted, other kinds of weighted means, etc.

The point of being picky about this is that if a person were to ask what is CAGR and how do you calculate it and someone were to say it is geometric average of the returns and a person were to look that up, they would get an answer that doesn't work. For example, Excel has a function for the geometric mean of a set of data GEOMEAN(of a set of positive data elements) that does not produce CAGR. There is no function for CAGR in Excel but you can calculate it this way: https://www.excel-easy.com/examples/cagr.html Strangely enough that hint does not even use the GEOMEAN function even though it could.
User avatar
vineviz
Posts: 7520
Joined: Tue May 15, 2018 1:55 pm

Re: How to get 10 year cumulative return on TSP account

Post by vineviz »

dbr wrote: Sun Sep 15, 2019 8:00 am
vineviz wrote: Sat Sep 14, 2019 1:45 pm
dbr wrote: Sat Sep 14, 2019 9:32 am **They also make the standard mistake of calling the result the geometric average of the annual returns when it is actually the compound annual growth rate (CAGR). See here https://en.wikipedia.org/wiki/Geometric_mean and here: https://en.wikipedia.org/wiki/Compound_ ... rowth_rate
In this use case, geometric mean and CAGR are identical and interchangeable.

More generally, CAGR is equivalent to geometric mean in the absence of internal cash flows. In other words, CAGR is the time-weighted geometric mean.
No. See this article for the f-mean where f for the geometric mean is log(r) and f for the CAGR is log(1+r):
For better or worse, in finance, returns are always expressed as decimal multiplier equivalents (i.e. 1+r).

Otherwise, it'd be literally impossible to compute the geometric or harmonic mean of a series like +5%, -5%, 0%. So when financial professionals talk about return series they are always (and they know this explicitly if they've been properly trained) referring to 1.05, .95, 1.0 and not the colloquial convention of +5%, -5%, 0%.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
dbr
Posts: 33638
Joined: Sun Mar 04, 2007 9:50 am

Re: How to get 10 year cumulative return on TSP account

Post by dbr »

vineviz wrote: Sun Sep 15, 2019 8:46 am
For better or worse, in finance, returns are always expressed as decimal multiplier equivalents (i.e. 1+r).

Otherwise, it'd be literally impossible to compute the geometric or harmonic mean of a series like +5%, -5%, 0%. So when financial professionals talk about return series they are always (and they know this explicitly if they've been properly trained) referring to 1.05, .95, 1.0 and not the colloquial convention of +5%, -5%, 0%.
Evidently so. That leaves some questions about what is being quoted when a return is quoted as x%, x being a number between (-100%+delta) and something algebraically larger.

But, yes, inside the fraternity no one is actually confused. They just speak a funny language.

More than that when CAGR is quoted it too is a percent and not a growth multiplier. This is consistent with another definition of general average which is that the general average is the (inverse transformation) [arithmetic average] (transformation) of the data. In the case of CAGR the transformation is a composition of three transformations (read l-r in order, reverse for the standard notation for composition of functions): (divide by 100:add 1:take the log) and the inverse is: (take the antilog:subtract 1:multiply by 100). Article is here: https://academickids.com/encyclopedia/i ... sed_f-mean

Finally another definition of mean is the value that if put in place of the values of a set of data in a calculation gives the same result. IRR, for example has the nature of finding a number that applied to each of a set of cash flows generates the total investment cost even if the individual investments are made at irregular periods and the return in each period varied. In the case of IRR there is no f-mean formula that I have found. If someone knows one it would be interesting to see it. I haven't tried to express the time weighted average return for unequal periods as an f-mean.
Post Reply