Killoran suggested looking at Google instead and I've run a comparison between Yahoo, Google and ADVFN which produces some strange (and frustrating issues).
Reason for posting is two-fold
1. To seek any group think thoughts on what the issues may be and what is actually the "best" source of historical price data.
2. As a cautionary note for anyone using historic price data assuming it to be "correct".
As I noted on of the values I saw in the Yahoo data was the use of "Adjusted" closing prices, these adjust for stock splits/ consolidations and for dividends paid, effectively giving a Total Shareholder Return [TSR] measure.
Actually I'm less interested in the dividend element as it implicitly assumes dividend reinvestment, which may not be appropriate, but factoring out consolidations/splits is certainly useful.
So cutting to the chase here is the comparison between the three data sources for BARC going back over time.
(apologies for the formatting, I couldn't seem to make [ table ] work )
Date Y-Actual Y-ADJ Advfn Goog Y/Ratio Note
23-Jun-17 197.55 197.55 197.55 197.55 1.0000
09-Jun-17 206.60 206.60 206.60 204.78 1.0000 1
02-Mar-17 229.50 229.50 229.50 229.50 1.0000
01-Mar-17 232.90 230.90 232.90 232.90 0.9914 2
12-Aug-16 163.40 162.00 163.40 163.40 0.9914 3
11-Aug-16 163.25 161.85 163.25 163.25 0.9914
10-Aug-16 163.70 161.30 163.70 163.70 0.9854 4
18-Sep-13 277.20 252.92 277.20 277.20 0.9124
17-Sep-13 276.20 252.00 299.00 299.00 0.9124 5
29-Apr-02 566.26 329.54 613.00 613.00 0.5820
25-Apr-02 36063.00 5246.88 2440.00 610.00 0.1455 6
So what is happening?
For recent data the prices match as you would expect.
1. (9-Jun-17) there are a occasional small mismatches on given days.
2. (1,2-Mar-17) You can see the impact of the 2p dividend on the Yahoo adjusted price. The closing price on the day before XD day is adjusted down by the value of the dividend.
3. This is carried back in time as a ratio.
4. (10,11-Aug-16) - Adjustment for another dividend (1p)
Actual closing price 10-Aug of 163.7p based on the 0.9914 ratio would give an adjusted closing price of 162.3p, this is reduced by the 1p dividend to 161.3p
All looks good up this point!
5. (17-18 Sep-13) Both Google and ADVFN show a fall in the ACTUAL price from 299.00 to 277.2, Yahoo shows a 1p increase.
So here is what I *think* happened :
> The actual price did fall by 21.8 reflecting the dilution.
> My view is that you should treat the RI like a dividend as the default position is that your rights are sold and you receive cash.
> Yahoo made that adjustment, but made it to the ACTUAL price rather than just to the the ADJUSTED price.
On the plus side the fact that Yahoo have made this adjustment does mean you reflect the value in the TSR view, but the fact they made the adjustment in the "wrong" place is bloomin confusing. As rights issues are not generally recorded in accessible dividend tables, it would be difficult to automatically capture this using the BARC/ADVFN numbers.
(The 8% difference this creates between Yahoo and Goog/ADVFN does persist back in time).
6. (25,29-Apr-2002) There was a 1->4 stock split.
Google have adjusted the historical prices before that date to give a like for like comparison over time.
ADVFN have not so they report the actual closing price before that date.
Yahoo - What the heck have they done?????
Looking at the change in the ratio between the closing price and the adjusted closing price they have attempted to apply the stock split; but what they have done to the actual closing price before that date I really cannot fathom.
Unless there is something I have missed the Yahoo prices before that date are goobledygook.
What do I conclude?
Bear in mind this is based on sample of 1, so it is an act of faith at this stage to draw conclusions.
The "best" data source would appear to be the Google data (although it is the one that seems to have the most small single day anomalies).
The benefit of Google over ADVFN is that Google DO (appear on this one sample) to adjust prices for splits/consolidations (ADVFN data would present an unexplained 75% loss).
Unfortunately the Google data (nor ADVFN) do not allow you to automatically identify the benefit of the rights issue (they present an unexplained 7% loss) and unlike dividends , you cannot easily look this up in a table.
Assuming that Yahoo always make this adjustment to the actual as opposed to adjusted price - I suppose it might be possible to look for changes in the ratio Yahoo Actual vs Google actual?
These would probably have to be manually assessed as they would also throw up the change caused by Y's bizarre treatment of the stock split.
It's frustrating that there doesn't seem to be one consistently 'correct' source of information.
I'd certainly be grateful for any thoughts on this or any feedback from anyone that has tackled this kind of issue before!
Also if anyone has any contemporaneously collected price history for any stock(s) going back a long period this would be valuable to further test this. (Especially if there are Rights Issues, Consolidations, Splits, Special Dividends involved).
Regards,
Gromley