Page 1 of 1

Retroactively reclassified distributions vs estimated tax payments

Posted: Tue Dec 17, 2019 7:37 pm
by boomer_techie
Next year I quite likely will need to make unequal estimated tax payments, followed up by filing 2210-AI to defend my payments. This year it is just a dry run to work out the mechanics. Filling out 2210AI essentially requires a 1099 package through each quarter. Toward that end, I've constructed a spreadsheet that will generate the 1099 and 1099-like data from monthly broker statements and year end numbers published by fund companies. (I suspect I'll be able to fully predict my 1099 before my broker actually produces it.)

One issue without a clear answer is how to handle retroactive reclassification of fund distributions, i.e. when a fund takes from the dividend pile and calls it return of capital, long term gains, or several other possibilities. (In 2018 Vanguard's BND ETF actually took from the LTCG pile and put some back in the dividend pile!) In the spreadsheet, I calculate such reclassifications proportionally across all distribution payments. (For example, if the 1099 form shows a fund with $100 of distributions has $10 of LTCG, then I assume that 10% of each payment is LTCG in whatever months my broker statements show a distribution payment.)

The above generates tax quarter numbers with just monthly broker statements and the end of year 1099. Should I instead download a fund's distribution record once the numbers are changed to reflect reclassifications? For example, currently Vanguard's REIT ETF (VNQ) shows no return of capital for 2019. I assume in January 2020 the numbers will be edited to show the return. Looking at prior years, it seems that Vanguard does the same proportional assignment as my calculations. But BND's LTCG->dividend assignment wasn't proportional. To me it seems that having to access "private" data goes beyond the spirit of form 2210-AI.

Re: Retroactively reclassified distributions vs estimated tax payments

Posted: Tue Dec 17, 2019 11:53 pm
by lstone19
Most such reclassifications will work to your advantage and reduce taxes due. It’s OK for Schedule AI to overstate income for a quarter. So long as you can prove that income for the quarter was not more than what you put on Schedule AI, you’ll be OK. There’s no penalty for overpaying your taxes.

Re: Retroactively reclassified distributions vs estimated tax payments

Posted: Thu Dec 19, 2019 2:50 am
by boomer_techie
lstone19 wrote:
Tue Dec 17, 2019 11:53 pm
Most such reclassifications will work to your advantage and reduce taxes due.
Yes I'm aware that most reclassifications would reduce taxes.
There’s no penalty for overpaying your taxes.
This is mostly an academic exercise. The differences here (for me) are tiny - and likely swallowed up by whatever buffer amount I choose to overpay. However, I'd like to definitively establish procedures I may use for years to come. My first inclination is to use just the documents supplied to me, i.e. to proportionally assign according to monthly statements. It is only recently that I realized funds edit their published distribution histories after the end of the tax year.

I find the BND example interesting: I received a small LTCG distribution in April 2018. And then in December, one nines times as big. Two months later, the 1099 reduced LTCG by an amount curiously exactly the same as the December payment. I calculated taxes assuming two distributions, each 1/10th as much as the original amounts. Months later I was looking at historical Vanguard distributions and noticed that only the April payment is listed. Aha!

Re: Retroactively reclassified distributions vs estimated tax payments

Posted: Thu Dec 19, 2019 8:32 am
by lstone19
boomer_techie wrote:
Thu Dec 19, 2019 2:50 am
This is mostly an academic exercise. The differences here (for me) are tiny - and likely swallowed up by whatever buffer amount I choose to overpay. However, I'd like to definitively establish procedures I may use for years to come. My first inclination is to use just the documents supplied to me, i.e. to proportionally assign according to monthly statements. It is only recently that I realized funds edit their published distribution histories after the end of the tax year.
I use Fidelity and can only speak based on what I see there but they provide a lot of detail on-line regarding dividends and distributions and you can drill down to see the details of each distribution. However IIRC, if say a mutual fund paid dividends quarterly and at the end of the year reports that 60% were qualified dividends, then Fidelity, absent any other information, will apply that proportionately across all distributions for the year.

Re: Retroactively reclassified distributions vs estimated tax payments

Posted: Thu Feb 13, 2020 7:38 am
by boomer_techie
boomer_techie wrote:
Tue Dec 17, 2019 7:37 pm
(I suspect I'll be able to fully predict my 1099 before my broker actually produces it.)
I'm getting very close to accomplishing the above - just waiting on iShares / Blackrock to publish their distribution detail.

However, I've run into a discrepancy! For qualified dividends, I can't calculate the exact same number that my broker shows. The results are withing a few cents of each other, but sometimes one is higher, sometimes the other. A can't come up with rounding or truncation options that work for all values.

In particular, ($0.17965 qual/share * 31 shares) + ($0.21186 qual/share * 31 shares) should calculate to $12.16 according to my broker. FYI unrounded this is (0.17965 * 31) + (0.21186 * 31) = 5.56915 + 6.56766 = 12.13681. Try as I might, I can't get it to round to a value bigger than $12.14.

(Solving this is merely academic - I have insufficient (ordinary income + LTCG + qualified dividends) for the exact amount of qualified dividends to matter.)