Price history after a stock split

Parador

I’m about 60% through a 2007 to Premier 2019 Mac transfer of a number of accounts. This particular issue is a brokerage account and with a security with a 10:1 stock split that happened several months ago. The account is not linked to an online account. Everything ever entered was done by hand or as part of a security price update over many years. After the data was automatically converted (open the 2007 file with 2019) the plot of account value looked like a saw tooth. Prior to the stock split quicken updated the historical prices to be the post split equivalent value (i.e. 1/10th of what was there before) for every possible date it had previously had a stock price except for those that had been entered manually. For the manually entered transactions the calculated value (price x number of shares) was correct. For all other pre-split dates the value computed and displayed was off by a factor of 10. I did a stock price rebuild and allowed it to change any price entry it wanted to. The saw-tooth look is gone but now there is a step change in displayed value at the date of the stick split that isn’t really what happened. Anyone else seen this sort of behavior? Quicken should know how to provide historical value not just historical price. If it is expected to match what is displayed in an online brokerage account history or be able to compare to a statement it needs to know the difference between current day historical prices that may have been altered by a split and the prices actually paid throughout an investments history.

Parador

Thanks for the ideas! It is a Fidelity fund although held in a non-Fidelity account. The before and after are shown below. The stock split occurs at the steep step function at the right side of both plots. The “teeth” in the  chart are caused by the difference between a manually entered pre-split price (dividend reinvests) that I hadn’t yet allowed Quicken to update and the post split historic that it apparently uses for the dates along the way. What Quicken appears to be doing is just using the pre-split quantity and the post-split historical pricing and so is off by 10X for all dates prior to the split. I had discovered the “Placeholders” opportunity when I first tried to get the total share count correct and had already deleted them.

Parador

And this will illustrate the stock prices that Quicken is using for a portion of the before plot.

Parador

The after is only good post split. In the prices list you'll see the stock price rattling back and force by 10X depending on date. That appears to be just the difference between price data that Quicken just downloaded and data that was put in manually over the years along with transactions. Any computation of gains will be wrong if it computes based on the data presented by the graph or stock price list. To fix this I'd have to go into Quicken's prices list and manually change every split adjusted stock price for every date it has up until the split date back to the pre-split amount to make the whole data-set correct. Basically multiply everything that has been divided by 10, by 10, up to the date of the split. That's a lot of manual edits and would probably instantly turn me into a customer for another platform after being a loyal Quicken user since we printed checks on daisy wheels!

Effectively what the graph is trying to tell me is:  I bought 100 shares for $10 a share so I have a cost basis of $1000. Post the 10:1 split I now have 1000 shares at $1 and a cost basis of $100. Clearly not the case. They either have to have the price per share both pre- and post- split correct or they need to adjust for the split internally to the program. I'm not sure how it is coded but maybe your earlier comment about a problem with the quote server may be still in effect? I have some more accounts to convert and they have some stock splits as well. So far I can't tell if this is a bug in the program or a flaw in their data. Is there anyway to verify the quote server data?

james Franke

Parador said:

The after is only good post split. In the prices list you'll see the stock price rattling back and force by 10X depending on date. That appears to be just the difference between price data that Quicken just downloaded and data that was put in manually over the years along with transactions. Any computation of gains will be wrong if it computes based on the data presented by the graph or stock price list. To fix this I'd have to go into Quicken's prices list and manually change every split adjusted stock price for every date it has up until the split date back to the pre-split amount to make the whole data-set correct. Basically multiply everything that has been divided by 10, by 10, up to the date of the split. That's a lot of manual edits and would probably instantly turn me into a customer for another platform after being a loyal Quicken user since we printed checks on daisy wheels!

Effectively what the graph is trying to tell me is:  I bought 100 shares for $10 a share so I have a cost basis of $1000. Post the 10:1 split I now have 1000 shares at $1 and a cost basis of $100. Clearly not the case. They either have to have the price per share both pre- and post- split correct or they need to adjust for the split internally to the program. I'm not sure how it is coded but maybe your earlier comment about a problem with the quote server may be still in effect? I have some more accounts to convert and they have some stock splits as well. So far I can't tell if this is a bug in the program or a flaw in their data. Is there anyway to verify the quote server data?

The rebuild changes all the prices before the split (say $111) to 1/10 the value ($11). On the portfolio view, you then see a huge increase in value - but really the value of the graph has decreased before the split. Go to Motley, Yahoo whatever for price history and the original values before the split have been changed to 1/10 the original. So, if you manually change values in your price history the auto update of prices changes them right back to the wrong values.  I guess there were never stock splits before the software was created?

Parador

Just so it is clear. The rebuild of the price history is how I got from the Before to the After chart. I allowed it to change the manually entered prices during the rebuild. If all stock splits in Quicken now work this way than the program has serious programming logic problems and is useless for investment tracking. Can anyone assure me that it does work correctly for all other (or at least most, some) stock split situations?

EPK29

Same problem (FMAGX); just converted from Q2016 for Mac to Q2019 (didn't want to, but the new owners of Quicken left me no choice). The same thing appears to have happened with the other securities as to which I've had stock splits.

Quicken for Windows (don't ask: running VMWare was worse than Q2016, although the Quicken product was much better) always got splits right: it went back and adjusted prices and volumes from first purchase. Always got the cost basis right, too. I simply cannot believe that the database architecture on the Mac (which is Quicken's excuse for everything they do under Windows that they don't do on Mac) is SO different that they can't do this simple, basic operation.

My strong guess, based on the comments above and the way my transaction register looks, is that there is no workaround that preserves cost basis (which is what really matters). When Quicken fixes the bug, I know the answer will be to delete the stock split transaction, then rebuild the price history, then reinsert the stock split transaction. But right now, it's a bug.