Please help with this cost basis nightmare!

[guitarguy]

Hi all - I've been discussing this topic as part of a separate thread, but I thought this issue might do better on its own by eliminating all of the other surrounding issues that are now figured out / unrelated.

Big nod to user cas who has been an amazing help thus far for me!

Short background: My grandparents started a taxable stock account for my mother when she was a teenager (from her memory) and it has for the most part remained untouched for decades. I can only find Yahoo historical data back to 1/1/1970 (she would've been 19) so I'm guessing it was not long after then that the account was opened. Until Q2 2012, it had been completely untouched and was setup as a DRIP plan to reinvest all of the dividends into buying more of the stock, which is a single DTE Energy Co. holding (DTE). In 2012 her tax accountant/advisor helped her setup the account to receive the dividends as opposed to reinvest them. Now she is retired and looking to me for help setting up her portfolio, and diversifying this DTE holding which is currently valued at $150k and is making up nearly 1/3 of her total retirement savings is a big part of that equation. With very little to go by in terms of record keeping, I'm having a heck of a tough time figuring out the cost basis and I'd like to get some idea whether this method I'm using to estimate it makes sense.

Other highlights:

• She has records via her online account of the dividend amounts dating back to 2006, so starting here via the spreadsheet below
• She has copies of her tax returns dating back to 2002, so if this methodology makes sense I could probably continue it back through 2002 based on her statements
• No idea what to do beyond that, hoping to figure out a way to estimate, or will be stuck using $0 as the cost basis

Goal: to sell as much DTE stock as possible in Q4 2017, and Q1 2018 while keeping her LTCG income below the critical number to avoid "the hump" https://www.bogleheads.org/wiki/Social_ ... calculator in her marginal tax rate for tax years 2017 and 2018. I've learned how to do all these calculations, now all I need to do is figure out how much DTE to sell!

Here is a look at what I came up with as a way to estimate the cost basis on the shares that I can find records for. Basically I researched the median share price for the day that her dividend payment was made, and calculated based on that share price and the dividend amount, the number of shares that were purchased. That gave me a total number of shares purchased from 2006-2012, and the total amount it cost to purchase those shares. Then I subtracted the total cost from the total she'd receive today for selling those shares, giving me a value for her net LTCG income from the transaction.

Note: Data in blue was taken from her online account (covered shares). Data in black text is my calculation / estimate.



Questions:

1. First of all does this methodology of estimating the cost basis make sense?
2. What should I do when the records run out prior to 2006? Is there a mathematical way to extrapolate, or another means of making a reasonable estimate?
3. I'm having trouble understanding why when I go to the "sell" section of her online account, I don't see various lots. Everything seems to be lumped together as of (what I assume to be) the date that the funds moved out of the DRIP plan and she set herself up to receive the dividend checks instead. See the image below. Since this is the case, when she goes to sell, how am I supposed to determine which shares are being sold? Do we just get to pick?

Greatly appreciate any help on this!!

[cas]

She has copies of her tax returns dating back to 2002, so if this methodology makes sense I could probably continue it back through 2002 based on her statements

<skip>
What should I do when the records run out prior to 2006? Is there a mathematical way to extrapolate, or another means of making a reasonable estimate?

The tax records can almost certainly get you back prior to 2006 back through 2002, to some degree.

The 1099-DIV from DTE would be best. (Actually, if you are really, really lucky your mother threw the 4 quarterly statements from the DTE DRIP program into the tax file. Then you are golden and can fill out the spreadsheet with certain knowledge in all columns.) If the 1099-DIV isn't in the file, then use the amount that is reported on the dividends line on the front of the 1040. You know that the only taxable dividends your mother was receiving were from DTE, so that number on the 1040 should be the same as what was on the 1099-DIV.

So ... the 1099-DIV/1040 dividend line will give you certain knowledge of the total dividends distributed in that year. You know that all those dividends were reinvested, so that is your total cost basis for the year, even if you don't know the exact amount of each quarterly dividend. For the date, I'd just put "various 2004" (or whatever) and assume you'll sell the whole year's worth in one chunk.

But then it gets vague ... anybody have an ideas what to use as a reasonable share price in this whole-year situation?

You can get the quarterly dividend distribution date from yahoo, as well as what the $ dividend/share was. You can get the midpoint share price for the days of the 4 quarterly dividend distributions. But then my mind is going fuzzy on what to do with that info to get to a reasonable estimated # of shares bought during that year.

[cas]

I can only find Yahoo historical data back to 1/1/1970 (she would've been 19) so I'm guessing it was not long after then that the account was opened.

BTW, I think the 1/1/1970 limit might be a Yahoo thing for all stocks. DTE may well have been alive and kicking and able to be purchased before then. (Yahoo just doesn't have or is choosing not to show date from before 1970.) But I'm not sure about that. I'm not sure it will end up mattering anyway.

Also- important data point: as far as I can tell, DTE has never had a corporate action (split, reverse split, spin-off, merger) that would affect cost basis in all the decades your mother owned it. These types of actions would cause "adjustments" to have to be made to cost basis and/or share amounts, so it is good they didn't happen.

[guitarguy]

She has copies of her tax returns dating back to 2002, so if this methodology makes sense I could probably continue it back through 2002 based on her statements

<skip>
What should I do when the records run out prior to 2006? Is there a mathematical way to extrapolate, or another means of making a reasonable estimate?

The tax records can almost certainly get you back prior to 2006 back through 2002, to some degree.

The 1099-DIV from DTE would be best. (Actually, if you are really, really lucky your mother threw the 4 quarterly statements from the DTE DRIP program into the tax file. Then you are golden and can fill out the spreadsheet with certain knowledge in all columns.) If the 1099-DIV isn't in the file, then use the amount that is reported on the dividends line on the front of the 1040. You know that the only taxable dividends your mother was receiving were from DTE, so that number on the 1040 should be the same as what was on the 1099-DIV.

So ... the 1099-DIV/1040 dividend line will give you certain knowledge of the total dividends distributed in that year. You know that all those dividends were reinvested, so that is your total cost basis for the year, even if you don't know the exact amount of each quarterly dividend. For the date, I'd just put "various 2004" (or whatever) and assume you'll sell the whole year's worth in one chunk.

But then it gets vague ... anybody have an ideas what to use as a reasonable share price in this whole-year situation?

You can get the quarterly dividend distribution date from yahoo, as well as what the $ dividend/share was. You can get the midpoint share price for the days of the 4 quarterly dividend distributions. But then my mind is going fuzzy on what to do with that info to get to a reasonable estimated # of shares bought during that year.

This is what's fuzzy for me...

Can I just say for example:

• Selling shares from this big lumped tax lot with a single transaction in Q4 2017
• Choosing to assume that all shares obtained through DRIP from Jan 2007 through Jan 2012 are the ones sold in this transaction
• This yields a total of 308 shares sold with an estimated cost basis of $13,253.38 for a total LTCG income gain of $22,120.42
• Print out a copy of the spreadsheet with documentation of how we came up with these numbers, and include with 2017 tax return

...?

Is it that easy?

I mean if she were audited and she showed the IRS how these figures were calculated based on the information available I would think anyone would agree we made a reasonable effort to get it correct with the information at hand...

[cas]

This is what's fuzzy for me...

Can I just say for example:

• Selling shares from this big lumped tax lot with a single transaction in Q4 2017
• Choosing to assume that all shares obtained through DRIP from Jan 2007 through Jan 2012 are the ones sold in this transaction
• This yields a total of 308 shares sold with an estimated cost basis of $13,253.38 for a total LTCG income gain of $22,120.42
• Print out a copy of the spreadsheet with documentation of how we came up with these numbers, and include with 2017 tax return

...?

Is it that easy?

In theory, yes. (You don't even have to print out the spreadsheet and include it with the return. You just have to produce the spreadsheet and supporting documentation (e.g. screenshot of the screen that told you the quarterly dividend amounts 2006 and after, 1099-DIV for 2002-2006) if the IRS audits you.)

This would require you to use the "Specific ID" method of tax lot identification, which is what the screenshot you show seems to be getting at with "By Transaction Date". (FIFO is "First in First out" which would sell the earliest shares (1970) first.)

But then, like you, I'm confused that that screen isn't even showing the covered share lots broken out as separate lots for you to select. (I kind of get that they might show all non-covered shares as one big lump, then put the burden on you to keep records of which lots you actually sold. But the inability to get at the covered shares specifically confuses me.)

But I don't do much selling of shares. I'm curious what people who have more experience selling stocks (or ETFs or mutual fund lots) make of that screenshot.

[guitarguy]

But then, like you, I'm confused that that screen isn't even showing the covered share lots broken out as separate lots for you to select. (I kind of get that they might show all non-covered shares as one big lump, then put the burden on you to keep records of which lots you actually sold. But the inability to get at the covered shares specifically confuses me.)

But I don't do much selling of shares. I'm curious what people who have more experience selling stocks (or ETFs or mutual fund lots) make of that screenshot.

Me too!

My guess is that somehow the shares were all lumped at the time when everything was removed from the DRIP and converted to traditional shares (terminology is probably wrong...but hopefully the idea is there...)

[Wagnerjb]

I agree with your strategy of first selling the shares where you can fairly easily determine the cost basis. And I agree with "cas" that it is very important to indicate to the broker that you want to use "specific identification" of the tax lots when selling. I agree that the screen shot seems to indicate that you can use specific ID, but it is important enough that I would call the broker to be 100% sure.

You probably want to have good documentation of your correspondence with the broker in case of a tax audit (which is exceedingly unlikely anyway). The broker WILL NOT report the cost basis of any of these pre-2012 share purchases to the IRS, so they don't have any better records than you do. You are responsible with determining the cost basis and justifying it in the unlikely case of an audit.

I am not exactly sure what tax "hump" you are trying to avoid, but once you have determined the maximum capital gains that your grandmother can generate (without hitting this "hump"), the selling should be easy.

Best wishes.

[cas]

Here is a look at what I came up with as a way to estimate the cost basis on the shares that I can find records for. Basically I researched the median share price for the day that her dividend payment was made, and calculated based on that share price and the dividend amount, the number of shares that were purchased.

Here's some good support that using the median share price for the day (or, more precisely, the midway point between the high and low shares price for the day, which is what I think you are doing) is the correct approach. (Absent any other information, some people might advocate using the low share price for the day.)

From the prospectus for the DTE Dividend Reinvestment Program, p. 7:

How is the price of the shares determined?

If the Plan Administrator purchases your shares from DTE Energy (“original issue”), the price per share will be the average of the high and low prices on the New York Stock Exchange Composite Tape for DTE Energy common stock on the pricing date. No brokerage commission will be charged. The pricing date for original issue shares purchased in connection with dividend reinvestment and any shares related to cash investments will be the 15th day of the month. If the 15th day of the month is not a business day, the pricing date will be the next business day.

If the Plan Administrator purchases your shares in the public markets or in privately negotiated transactions,
the purchase price per share will be the average price of all shares purchased, including brokerage commissions.

Source: https://www2.dteenergy.com/wps/wcm/conn ... 570a7d5f83

(I had glanced briefly through the prospectus back when I was trying to figure out what "The Plan" referred to on your screen shots, and my brain suddenly coughed up a random memory of the prospectus describing how DRIP purchase price was determined.)

[UncleBen]

http://www.morningstar.com/stocks/xnys/dte/quote.html
Click on maximum and hover over the line. This will get you prices back to 6/1/72.

I would think that DTE shareholder services could provide year end statements for each year that the stock was owned. (I have mine for ATT and baby bells and they show every purchase and dividend reinvestment).

The bad news is that if your cost basis differs from the information provided to the IRS it will call attention to the sale. I find this fairly common when I sell older stocks but am confident enough to use my numbers and have not had any issues. The good news is that the IRS is pretty accepting of your number if you can show that you've made a good faith attempt to estimate the capital gain.

[guitarguy]

I agree with your strategy of first selling the shares where you can fairly easily determine the cost basis. And I agree with "cas" that it is very important to indicate to the broker that you want to use "specific identification" of the tax lots when selling. I agree that the screen shot seems to indicate that you can use specific ID, but it is important enough that I would call the broker to be 100% sure.

You probably want to have good documentation of your correspondence with the broker in case of a tax audit (which is exceedingly unlikely anyway). The broker WILL NOT report the cost basis of any of these pre-2012 share purchases to the IRS, so they don't have any better records than you do. You are responsible with determining the cost basis and justifying it in the unlikely case of an audit.

I am not exactly sure what tax "hump" you are trying to avoid, but once you have determined the maximum capital gains that your grand mother can generate (without hitting this "hump"), the selling should be easy.

Best wishes.

I will make sure to contact the broker and make sure that selecting specific ID when selling will generate the sale in that manner, and document the communication.

Re: the tax hump...this was incredibly interesting to learn about. To put it somewhat correct and simplified, there is a huge marginal tax rate introduced (up to 40-50%) as capital gains income exceeds the top of the 15% bracket because each additional dollar generates $1.85 in additional taxable income: $1.00 being the capital gains now taxed at 15%, and then based on combined income, $0.85 of additional social security benefit income being taxed at ordinary tax rates. If interested there is a lot of info in the wiki which I linked to in my OP.

[guitarguy]

Here is a look at what I came up with as a way to estimate the cost basis on the shares that I can find records for. Basically I researched the median share price for the day that her dividend payment was made, and calculated based on that share price and the dividend amount, the number of shares that were purchased.

Here's some good support that using the median share price for the day (or, more precisely, the midway point between the high and low shares price for the day, which is what I think you are doing) is the correct approach. (Absent any other information, some people might advocate using the low share price for the day.)

From the prospectus for the DTE Dividend Reinvestment Program, p. 7:

How is the price of the shares determined?

If the Plan Administrator purchases your shares from DTE Energy (“original issue”), the price per share will be the average of the high and low prices on the New York Stock Exchange Composite Tape for DTE Energy common stock on the pricing date. No brokerage commission will be charged. The pricing date for original issue shares purchased in connection with dividend reinvestment and any shares related to cash investments will be the 15th day of the month. If the 15th day of the month is not a business day, the pricing date will be the next business day.

If the Plan Administrator purchases your shares in the public markets or in privately negotiated transactions,
the purchase price per share will be the average price of all shares purchased, including brokerage commissions.

Source: https://www2.dteenergy.com/wps/wcm/conn ... 570a7d5f83

(I had glanced briefly through the prospectus back when I was trying to figure out what "The Plan" referred to on your screen shots, and my brain suddenly coughed up a random memory of the prospectus describing how DRIP purchase price was determined.)

This is good news!

Now, maybe I will try to contact them and see how I can get a hold of the amount of dividends that where paid out to her annually prior to 2002 when the records go away. If I can find that, I could split into 4 chunks, estimate based on all of the previous dividends being paid on the 15th of the month (quarterly), find the median (yes - I actually did mean the mean) share price on that day, and I can estimate that way.

[cas]

Now, maybe I will try to contact them and see how I can get a hold of the amount of dividends that where paid out to her annually prior to 2002 when the records go away. If I can find that, I could split into 4 chunks, estimate based on all of the previous dividends being paid on the 15th of the month (quarterly), find the median (yes - I actually did mean the mean) share price on that day, and I can estimate that way.

Along those lines, I was thinking that a way to proceed for 2002-2005 (using the annual dividend amount from the tax return/1099-DIV) is to make the simplifying assumption that the 4 quarterly dividend payments for the year were all for the same amount.

This isn't *exactly* correct, due to the compounding effect of the DRIP purchases, but you also aren't dealing with vast sums here ... it is pretty close to to correct. And it would allow you to use the 2006-2010 spreadsheet technique to work back to 2002.

(I also had a thought on how to deal with the pre-2002 cost basis estimate. I'll type that after I submit this.)

[BL]

I believe the "covered" date is Jan 2011 for stocks, not 2012 as for mutual funds.

An investment is considered covered if it is: Shares of corporate stock acquired on or after January 1, 2011. Shares of stock in mutual funds and stock acquired in connection with a dividend reinvestment plan are generally not covered unless acquired after January 1, 2012

from
https://ttlc.intuit.com/questions/21014 ... pital-loss

[guitarguy]

I believe the "covered" date is Jan 2011 for stocks, not 2012 as for mutual funds.

An investment is considered covered if it is: Shares of corporate stock acquired on or after January 1, 2011. Shares of stock in mutual funds and stock acquired in connection with a dividend reinvestment plan are generally not covered unless acquired after January 1, 2012

from
https://ttlc.intuit.com/questions/21014 ... pital-loss

Yes, this is true.

I have data for covered shares that were accumulated via the dividend reinvestments starting Jan 2011. These are line items (in blue) in my spreadsheet calculation.

[cas]

No idea what to do beyond that [pre-2002 when availablity of tax return records ceases], hoping to figure out a way to estimate, or will be stuck using $0 as the cost basis

I think you can do better than $0.

At the very worst, you can determine the very lowest price paid for DTE stock between 1/1/1970 and 12/31/2001 (via Yahoo records, downloading data into excel, and some sorting work (See Note 1 below)). Then you can group all the pre-2002 shares of DTE into one lot with a "various 1970-2001" acquisition date (See Note 2 below) and assume that every single share in that lot was bought at that very lowest price. (In reality, most -quite possibly all- shares will have been bought for higher than that price, but the IRS can't argue that any were bought lower than that price.)

If I go the Yahoo chart for DTE ( https://finance.yahoo.com/quote/DTE/chart?p=DTE ) and restrict it to show 01/01/1970 - 12/31/2001, a rough guesstimate is that that low occurred at the end of 1974 (there was a big market crash around then) and is somewhere in the vicinity of $8 - $10/share. Not great, but better than $0/share.


But ... you can probably do a bit better.

Instead of using the absolute lowest price for any day, you can use the midway between the low and the high daily price (let's call that "average" daily price), since the DTE prospectus for the DRIP tells you that is closer to how the DTE DRIP was priced. You'd have to create a new column for "average" in the spreadsheet, then sort on that rather than on "low."

And, beyond that, you know that the DTE would have been purchased on only 4 days out of the year (the 4 days when the quarterly dividends were issued.) If you use that same historical price chart from Yahoo, but on the "Show" pull down menu select "Dividends only" it will show you what those days were. (Nominally DTE says (in that prospectus again) the dividend dates are Jan 15, Apr 15, Jul 15, Oct 15 ... adjusted when that day falls on a weekend or holiday. Some of the dates shown in Yahoo seem a bit more off than just adjusting for a weekend or holiday ... not sure what is up with that ... whether Yahoo has iffy data or whether it is correct.)

In any case ... if you induced Excel to actually sort (on the "Average" column) and show all the full lines low-to-high "Average", you wouldn't have to use the very first line (which would be the lowest "average" price EVER in that time frame.) You could scan down the dates until you found the first date in the list that was an actual dividend distribution date, and then use THAT "average" price/share as your price/share for every share in your big "Various 1970-2001" lot.

Technical details:

1 You might know better tools, but my thought was that you can use the Yahoo historical price table ( https://finance.yahoo.com/quote/DTE/history?p=DTE ) to download all the 1970 - 2001 daily price information into Excel. Or, if Excel on a PC is going to be unhappy about that amount of data, you can eyeball the price chart, get a pretty good idea when the lowest low must have been, and restrict on a smaller number of years (e.g. 1973 - 1976?). I'm only a novice at spreadsheets, but I'm guessing that Excel can be induced to then sort on the "Low" (or "Average") column from low to high? (I know that data manipulation languages like R can also handle this type of thing, if you happen to have one of those in your engineering toolbox and think Excel isn't a good fit.)

2. I'm lumping all the pre-2001 shares into "various 1970-2001" based on this from the Instructions for Form 8949 of the IRS 1040, which is the form used to report the sale to the IRS at tax time:

Stock acquired on various dates.

If you sold a block of stock (or similar property) that you acquired through several different purchases, you may report the sale on one row and enter “VARIOUS” in column (b). However, you still must report the short-term gain or (loss) on the sale on Part I and the long-term gain or (loss) on Part II.

Side comment: If you have time for some light reading, I'd recommend perusing Form 8949 ( https://www.irs.gov/pub/irs-pdf/f8949.pdf ) and the instructions for Form 8949 ( https://www.irs.gov/pub/irs-pdf/i8949.pdf ). I guess technically the accountant will be filing out the Form 8949, based on the 1099-B and information you give him/her, but knowing what is on the form may make things more clear in general,will help you look at the 1099-B and make sure it is correct, and will help you know what information you need to give the accountant.

I'd also recommending skimming the "Stocks and Bonds" section (starts p. 41) of Chapter 4 ("Sales and Trades of Investment Property") of Publication 550 "Investment Income and Expenses, including Capital Gain and Loss" ( https://www.irs.gov/pub/irs-pdf/p550.pdf )

You said you are an engineer ... I'm sure you have plenty of experience with dull technical documents. The IRS stuff isn't scintillating, but you'll be able to handle it.

3. Note: I'm using 12/31/2001 in my example as an endpoint, but you can decide whether you want to try to get pre-2002 data from WF. In that case, the date range in my examples would change. It might come down to them saying they'll send you old statements for that $15/year fee, in which case you'd have to decide at what point the $15/year ceases becoming worth it relative to tax you would avoid paying if you knew the true correct cost basis for that pre-2002 year. (At some point, due to the compounding effect, there aren't going to be all that many $ in dividends per year ... and much of it you will be paying 0% tax on anyway.)

Apologies ... I know all that was pretty dense, and I described it pretty fast. You've been so quick on the uptake on everything, that I figured you'd pick up on what I was getting at without much help. But if you have no idea what I was getting at at some point ... ask.

[Wagnerjb]

Re: the tax hump...this was incredibly interesting to learn about. To put it somewhat correct and simplified, there is a huge marginal tax rate introduced (up to 40-50%) as capital gains income exceeds the top of the 15% bracket because each additional dollar generates $1.85 in additional taxable income: $1.00 being the capital gains now taxed at 15%, and then based on combined income, $0.85 of additional social security benefit income being taxed at ordinary tax rates. If interested there is a lot of info in the wiki which I linked to in my OP.

No problem. I am a CPA and there are an awful lot of tax "humps" or '"cliffs" or "phase out zones" that people try to avoid. You did a nice job there of explaining exactly which one you are trying to avoid.

Thanks.

[guitarguy]

No idea what to do beyond that [pre-2002 when availablity of tax return records ceases], hoping to figure out a way to estimate, or will be stuck using $0 as the cost basis

I think you can do better than $0.

At the very worst, you can determine the very lowest price paid for DTE stock between 1/1/1970 and 12/31/2001 (via Yahoo records, downloading data into excel, and some sorting work (See Note 1 below)). Then you can group all the pre-2002 shares of DTE into one lot with a "various 1970-2001" acquisition date (See Note 2 below) and assume that every single share in that lot was bought at that very lowest price. (In reality, most -quite possibly all- shares will have been bought for higher than that price, but the IRS can't argue that any were bought lower than that price.)

If I go the Yahoo chart for DTE ( https://finance.yahoo.com/quote/DTE/chart?p=DTE ) and restrict it to show 01/01/1970 - 12/31/2001, a rough guesstimate is that that low occurred at the end of 1974 (there was a big market crash around then) and is somewhere in the vicinity of $8 - $10/share. Not great, but better than $0/share.


But ... you can probably do a bit better.

Instead of using the absolute lowest price for any day, you can use the midway between the low and the high daily price (let's call that "average" daily price), since the DTE prospectus for the DRIP tells you that is closer to how the DTE DRIP was priced. You'd have to create a new column for "average" in the spreadsheet, then sort on that rather than on "low."

And, beyond that, you know that the DTE would have been purchased on only 4 days out of the year (the 4 days when the quarterly dividends were issued.) If you use that same historical price chart from Yahoo, but on the "Show" pull down menu select "Dividends only" it will show you what those days were. (Nominally DTE says (in that prospectus again) the dividend dates are Jan 15, Apr 15, Jul 15, Oct 15 ... adjusted when that day falls on a weekend or holiday. Some of the dates shown in Yahoo seem a bit more off than just adjusting for a weekend or holiday ... not sure what is up with that ... whether Yahoo has iffy data or whether it is correct.)

In any case ... if you induced Excel to actually sort (on the "Average" column) and show all the full lines low-to-high "Average", you wouldn't have to use the very first line (which would be the lowest "average" price EVER in that time frame.) You could scan down the dates until you found the first date in the list that was an actual dividend distribution date, and then use THAT "average" price/share as your price/share for every share in your big "Various 1970-2001" lot.

Technical details:

1 You might know better tools, but my thought was that you can use the Yahoo historical price table ( https://finance.yahoo.com/quote/DTE/history?p=DTE ) to download all the 1970 - 2001 daily price information into Excel. Or, if Excel on a PC is going to be unhappy about that amount of data, you can eyeball the price chart, get a pretty good idea when the lowest low must have been, and restrict on a smaller number of years (e.g. 1973 - 1976?). I'm only a novice at spreadsheets, but I'm guessing that Excel can be induced to then sort on the "Low" (or "Average") column from low to high? (I know that data manipulation languages like R can also handle this type of thing, if you happen to have one of those in your engineering toolbox and think Excel isn't a good fit.)

2. I'm lumping all the pre-2001 shares into "various 1970-2001" based on this from the Instructions for Form 8949 of the IRS 1040, which is the form used to report the sale to the IRS at tax time:

Stock acquired on various dates.

If you sold a block of stock (or similar property) that you acquired through several different purchases, you may report the sale on one row and enter “VARIOUS” in column (b). However, you still must report the short-term gain or (loss) on the sale on Part I and the long-term gain or (loss) on Part II.

Side comment: If you have time for some light reading, I'd recommend perusing Form 8949 ( https://www.irs.gov/pub/irs-pdf/f8949.pdf ) and the instructions for Form 8949 ( https://www.irs.gov/pub/irs-pdf/i8949.pdf ). I guess technically the accountant will be filing out the Form 8949, based on the 1099-B and information you give him/her, but knowing what is on the form may make things more clear in general,will help you look at the 1099-B and make sure it is correct, and will help you know what information you need to give the accountant.

I'd also recommending skimming the "Stocks and Bonds" section (starts p. 41) of Chapter 4 ("Sales and Trades of Investment Property") of Publication 550 "Investment Income and Expenses, including Capital Gain and Loss" ( https://www.irs.gov/pub/irs-pdf/p550.pdf )

You said you are an engineer ... I'm sure you have plenty of experience with dull technical documents. The IRS stuff isn't scintillating, but you'll be able to handle it.

3. Note: I'm using 12/31/2001 in my example as an endpoint, but you can decide whether you want to try to get pre-2002 data from WF. In that case, the date range in my examples would change. It might come down to them saying they'll send you old statements for that $15/year fee, in which case you'd have to decide at what point the $15/year ceases becoming worth it relative to tax you would avoid paying if you knew the true correct cost basis for that pre-2002 year. (At some point, due to the compounding effect, there aren't going to be all that many $ in dividends per year ... and much of it you will be paying 0% tax on anyway.)

Apologies ... I know all that was pretty dense, and I described it pretty fast. You've been so quick on the uptake on everything, that I figured you'd pick up on what I was getting at without much help. But if you have no idea what I was getting at at some point ... ask.

This makes sense. I could run the numbers this way, essentially using a purchase price for the entire remaining lump sump of shares I don't have paperwork for that is the minimum of the average share price of Jan/Apr/Jul/Oct 15 of each year from 1970-2001.

Or, I could run the numbers by possibly dividing up the total number of shares by the number of distribution dates over those decades, and just guess that approx an equal number of shares were purchased each time, then find the mean share price for that day, and that would be that. Because this is occurring over so many decades...it obviously wouldn't be correct but it wouldn't be totally unreasonable in that we're making an attempt given the information we have. I'd have to also make an assumption at the initial investment. I could see how that compares to using the minimum price as in option #1 at least.

Either way, since she'll basically be paying zero tax on the majority of this anyway, I think any of these approaches as long as they're somewhat logical that we made an attempt, I would think should be just fine.

[guitarguy]

Re: the tax hump...this was incredibly interesting to learn about. To put it somewhat correct and simplified, there is a huge marginal tax rate introduced (up to 40-50%) as capital gains income exceeds the top of the 15% bracket because each additional dollar generates $1.85 in additional taxable income: $1.00 being the capital gains now taxed at 15%, and then based on combined income, $0.85 of additional social security benefit income being taxed at ordinary tax rates. If interested there is a lot of info in the wiki which I linked to in my OP.

No problem. I am a CPA and there are an awful lot of tax "humps" or '"cliffs" or "phase out zones" that people try to avoid. You did a nice job there of explaining exactly which one you are trying to avoid.

Thanks.

Thanks!

I read a lot, and got lots of help from fellow BHs, to understand that concept and how to do those calculations!

[cas]

Either way, since she'll basically be paying zero tax on the majority of this anyway, I think any of these approaches as long as they're somewhat logical that we made an attempt, I would think should be just fine.

The biggest side effect I see of a really low-ball estimate on the cost basis is that you might have to take a year or two more to sell off all the DTE and still stay within/near the 0% tax rate on the capital gains. (Low cost basis/share = higher capital gains per share = smaller # of shares you can sell per year and still stay within your additional-income-via-capital-gains budget -> potentially additional year(s) to sell of the DTE.) Probably still shouldn't be a problem, although there is that age 70 RMD-starting deadline where tax planning gets trickier if you are still selling DTE.

[guitarguy]

Either way, since she'll basically be paying zero tax on the majority of this anyway, I think any of these approaches as long as they're somewhat logical that we made an attempt, I would think should be just fine.

The biggest side effect I see of a really low-ball estimate on the cost basis is that you might have to take a year or two more to sell off all the DTE and still stay within/near the 0% tax rate on the capital gains. (Low cost basis/share = higher capital gains per share = smaller # of shares you can sell per year and still stay within your additional-income-via-capital-gains budget -> potentially additional year(s) to sell of the DTE.) Probably still shouldn't be a problem, although there is that age 70 RMD-starting deadline where tax planning gets trickier if you are still selling DTE.

Yeah...that's the only side effect really is it'll take longer to sell when taking taxes into account.

Once the RMDs from her IRA start, to the tune of about $12k per year based on a quick online calculator I found, it will just further reduce the amount of LTCGs she can take before she gets into tax hit territory. With the spreadsheet I made though...fortunately it won't be any more difficult a calculation! So it shouldn't be too difficult to plan for that if it comes to it.

[technovelist]

I can only find Yahoo historical data back to 1/1/1970 (she would've been 19) so I'm guessing it was not long after then that the account was opened.

BTW, I think the 1/1/1970 limit might be a Yahoo thing for all stocks. DTE may well have been alive and kicking and able to be purchased before then. (Yahoo just doesn't have or is choosing not to show date from before 1970.)

1/1/1970 is "day zero" for Linux/Unix operating systems. That is probably the reason for the limit on Yahoo stock prices.

[Ace1]

GG
I will add in a +1 to the calculations suggested above.
Given the age of the account, you might also contact the Investor Relations dept of the company.
They might have some original purchase data plus some of the subsequent dividend reinvestment
facts you could use as the starting point.
Good Luck.

[inbox788]

I only scanned some of the responses, but wanted to point out if it hasn't already been discussed, which is the dividends paid out have already been taxed, so be careful not to pay double tax when you add things up. I made a cost basis error, which in effect caused me to pay taxes when I received shares and again as when I sold them. It was a big error, and I was doing things myself with the help of TurboTax, but luckily I discovered the error. It has been the only time I've had to file and amended 1040X.

In your shoes, I would simply take a midpoint or average for the yearly values, which given the growth over decades means it's really a small percentage difference in cost basis value. You'll be paying a fairly significant capital gains tax anyway. Seems like a reasonable course, not a lot of dollars, and not as burdensome as requiring specific numbers or using minimums or worse zero. But INAL nor CPA.