MF cost basis conundrum

[GrayHare]

OK fellow tax computation fiends, I'm getting conflicting opinions about the following. For determining my cost basis of a mutual fund, if during previous years I have sold specific lots and identified them for tax purposes, can I during later tax years sell bunches of shares (non-specific lots) and use average cost basis (average of all shares other than the previously sold/identified ones)?

Some people say no because that's switching cost basis methods, something not allowed. Others say it's permitted because the math works, that is, I could assemble of bunch of specific shares that would average to the average cost basis of them all; these people say that once I do use average basis I must stick with it since the math does not support switching the other direction (from average to specific lot identification).

I am asking because after several decades of buy and hold I lack compete info on some MFs set to reinvest: for some of the reinvestments the old statements show the dollar amount of the reinvestment but omit the number of shares then purchased.

 

[braumeister]

My understanding has always been that the cost basis is determined when you buy it, so you can't change it later.

 

[LOL!]

It seems that you don't have a number for average basis anyways, so why use it.

If one sells shares without using average basis, one doesn't need to specifically identify the lots sold if one doesn't want to. In that case the IRS says that you used FIFO (same as with average basis). Also one can put the Date Acquired as "Various" if one bunches a bunch of lots together for selling on the same date.

So you might be able to figure out the number of shares and the basis of those shares, but not the exact purchase dates and prices by a little forensic work with your statements and with the internet. For instance, you probably have recent records and can "remove" those shares from the calculations.

In the old days, one actually had to put "Avg Cost" or something like that on their Schedule D to specify that they used Average basis and were stuck using it from then on.

 

[LOL!]

My understanding has always been that the cost basis is determined when you buy it, so you can't change it later.

This seems true on the face of it, but if one uses Average Basis, then the average price_per_share_paid does change with subsequent buys and sales. That's why I never use Average Cost (or Average Basis) because I could never make the calculation. With Specific Identification, the basis is what you paid for a share and that should not change (but should take into account splits, reverse splits, conversions, etc.).

 

[GrayHare]

It seems that you don't have a number for average basis anyways, so why use it.

No, I can calc the average of the non-specific-ID shares since the statements show total number of shares and total spent.

 

[LOL!]

But do you have to back out the previous shares sold via specific id from that calculation?

 

[MichaelB]

OK fellow tax computation fiends, I'm getting conflicting opinions about the following. For determining my cost basis of a mutual fund, if during previous years I have sold specific lots and identified them for tax purposes, can I during later tax years sell bunches of shares (non-specific lots) and use average cost basis (average of all shares other than the previously sold/identified ones)?

Some people say no because that's switching cost basis methods, something not allowed. Others say it's permitted because the math works, that is, I could assemble of bunch of specific shares that would average to the average cost basis of them all; these people say that once I do use average basis I must stick with it since the math does not support switching the other direction (from average to specific lot identification).

I am asking because after several decades of buy and hold I lack compete info on some MFs set to reinvest: for some of the reinvestments the old statements show the dollar amount of the reinvestment but omit the number of shares then purchased.

The law changed in 2012. Previous to that, you chose a method and then stuck with it. Since then you are free to change methods. You do need to keep careful records to show your calculations are correct.

Not sure how that resolves your problem. You still need to have accurate cost info on all your shares, both sold and still holding.

 

[kaneohe]

..................................

I am asking because after several decades of buy and hold I lack compete info on some MFs set to reinvest: for some of the reinvestments the old statements show the dollar amount of the reinvestment but omit the number of shares then purchased.

don't know what kind info you have and how complete.....but if all you're missing is the # of shares in the reinvestment, can't you look at the total shares after the reinvestment (but before before the next reinvestment )
and compare with the number before.

 

[GrayHare]

But do you have to back out the previous shares sold via specific id from that calculation?

Yes, and I have the data for that because the cost basis of those specific ID shares shows on the Sched D of years ago when they sold.

 

[GrayHare]

don't know what kind info you have and how complete.....but if all you're missing is the # of shares in the reinvestment, can't you look at the total shares after the reinvestment (but before before the next reinvestment )
and compare with the number before.

If it were just one month of missing reinvest that could work, but the number of reinvested shares are missing from the broker statements for a couple years.

 

[GrayHare]

The law changed in 2012. Previous to that, you chose a method and then stuck with it. Since then you are free to change methods. You do need to keep careful records to show your calculations are correct.

Not sure how that resolves your problem. You still need to have accurate cost info on all your shares, both sold and still holding.

If I can now use average cost basis the problem goes away since I know the current number of shares, and the total dollars of buys+reinvests. Dividing those two values (after subtracting shares sold years ago) tells me the average basis per share.

 

[pb4uski]

Doesn't the fund provide you with basis information? VG does for me but I turnover things enough that I don't have any really old purchase lots.

 

[GrayHare]

Doesn't the fund provide you with basis information? VG does for me but I turnover things enough that I don't have any really old purchase lots.

Only recent stuff. The MF in question dates from the Reagan admin. Part of the joy of buy and hold.

 

[kaneohe]

My memory (believe at own risk) says that in the olden days you could switch from specific shares to average at any time but could not change back w/o IRS blessing.
This suggests a similar thing for non-covered shares today.
Changing Cost Basis Methods - Fairmark.com Fairmark.com

 

[GrayHare]

My memory (believe at own risk) says that in the olden days you could switch from specific shares to average at any time but could not change back w/o IRS blessing.
This suggests a similar thing for non-covered shares today.
Changing Cost Basis Methods - Fairmark.com Fairmark.com

That link is great news, thanks!