Cost Basis Wrong Even After All Shares Removed Then Re-Added

chipwhite935
chipwhite935 Member ✭✭
edited April 2022 in Investing (Windows)
A stock I own did a corporate spinoff. I removed all 400 shares, which were in 2 lots of 100 and 300 each. Then I re-added the 2 new lots with the new correct cost basis. But now, the total cost basis calculated by Quicken is off by 0.20 even though the basis of both lots are shown correctly in the report (See screenshot). What it shows is:
10315.14 + 3894.90 = 14210.24 (should be 14210.04!)

I've run "validate file" and "rebuild investing lots," but that did not fix it. Obviously, I could probably fudge it, but it makes me wonder what else may be wrong that I haven't found yet.

Answers

  • Sherlock
    Sherlock Member ✭✭✭✭
    If you haven't already, I suggest you check Show closed lots: left click on the small gear icon at the upper right of the Holdings region on the Account Overview window.
  • chipwhite935
    chipwhite935 Member ✭✭
    OK. What I see there greyed out are the 2 times I've removed and re-added my 400 shares (See screenshot). I don't notice anything that appears incorrect, though.
  • Sherlock
    Sherlock Member ✭✭✭✭
    OK. What I see there greyed out are the 2 times I've removed and re-added my 400 shares (See screenshot). I don't notice anything that appears incorrect, though.
    The closed lots appear to have been sold for nothing.  Is that correct?
     
    Regarding the total cost basis issue, I suggest you determine when the issue occurred by resetting the As of: date at the upper right of the Holdings region on the Account Overview window.
  • chipwhite935
    chipwhite935 Member ✭✭
    The closed lots are not sales. I used "shares-out" and then "shares-in" in December 2021 just to update my cost basis.
    If I go back a year, my total basis is correct, but the bases of the two lots don't add up to the total. (15527.81 != 11303.68 + 4224.34) I can't explain it.

    This stock issues a return-of-capital several times a year, so perhaps it has something to do with how Q splits the R.O.C. between the two lots?

    In 2021, I received several instances of R.O.C. and there was a spinoff mid-year also affecting the basis. That's why I chose to remove, then re-add my lots with new bases per my financial statement.
  • q_lurker
    q_lurker SuperUser ✭✭✭✭✭
    Yes, RtrnCap will affect basis.  Quicken should be applying it in the normal case as a per-share adjustment.  So it you got a $100 RtrnCap on your 400 shares in the two lots of 300 and 100 shares each, the basis of the 300-share lot would reduce by 300/400 or $75 and the basis of the 100-share lot would drop by 100/400 or $25. 

    Now to the spinoff -- Quicken is screwing up there because they are now using a RtrnCapX transaction as part of the spinoff action.  In that case, the basis adjustment is (should be) different.  First one determines the fair market value of the parent and the spinoff after the spinoff.  If the spinoff valuation is 30% of the total, the spinoff gets 30% of the basis from each lot.  Rounding your numbers, that would be 30% of $10,000 from the 300 share lot ($3000, or $10/share) and 30% of $4,000 from the 100 share lot ($1200 or $12/share).  But Quicken would apply it as 30% of the $14,000 total basis = $4200 (right total) at $10.50/share to both lots (wrong amount to each lot).  But that does not explain the $0.20 error on your total.  

    Your choice to Remove and Add shares should have bypassed that spinoff problem.  

    Have you traded in and out of that security before?

    If you set the Portfolio View As of date to 11/4/13, does it show the correct basis for the 300-share lot?  What about on 6/14/15 = day before the next lot was acquired?  6/15/15?  Can you narrow down to when the $20 variation does first develop?  
  • chipwhite935
    chipwhite935 Member ✭✭
    I've never traded out of this security. I've just used RtnCap and sometimes Shares-out/in. I have to manually enter the RtnCaps because my broker doesn't supply them until I get a statement after the close of the year.

    I narrowed down an early (perhaps not the first) addition discrepancy between 7/13/2017 and 7/14/2017 after a RtnCap of 18.34 on 7/14:

    7/13/2017: 11891.09 + 4420.15 = 16311.24
    7/14/2017: 11877.35 + 4415.57 = 16292.90 (Total CB OK, but lots are off by 0.02)

    Probably some kind of internal rounding thing.
  • q_lurker
    q_lurker SuperUser ✭✭✭✭✭
    What I see in those numbers --
    18.34 / 4 = 4.585 which needs to be rounded some direction
    18.34 * 3 / 4 = 13.755 needs rounding also

    4420.15 - 4415.57 = 4.58;  apparently rounded down
    11891.09 - 11877.35 = 13.74; apparently rounded down too much.  Hmmm.

    But the total is right.  Quicken seems to recompute things multiple times rather than once and done.  It may be recomputing the cost basis with each presentation (original cost less each total RtrnCap ...) and then doing the same on the per lot numbers rather than by lot and then adding.

    So my next question, is the portfolio view value basis per lot the only computation?  -- If you enter a sale on 7/14/17 for both lots, what would the capital gains report show?  That could be a wholly independent calculation that might use the 4415.57 / 11877.35 pair, or a 4415.56 / 11877.34 pair or some other variation.  It is not uncommon for me to find a penny or two difference between my brokerage value and Quicken values. 

    Which takes me to the next thought:  If Quicken is building up a $0.02 / quarter discrepancy on the cost basis in the portfolio view, over 10 or so quarters, that could be your $0.20 variation.  I'd normally expect those rounding type variations to balance over time, but maybe there is something in your pattern and ratios that goes against that expectation.          
  • chipwhite935
    chipwhite935 Member ✭✭
    I agree that there may be a buildup of rounding errors over time, but that Q, under the hood, is keeping two separate tallies rather than adding lots to get the total. If I follow 2017 to the end of the year, the total cost basis is still correct, but the total of the lots as shown now diverges by 0.07.

    But that doesn't completely explain the original error I'm having where the total cost is wrong after the last thing I did was remove all 400 shares, and then re-add both lots specifying the known costs of each.
This discussion has been closed.