Minggu, 02 Oktober 2016

[smf_addin] Digest Number 3808

5 Messages

Digest #3808
1a
Re: Getting Yahoo Options Strikes After Hours by "Randy Harmelink" rharmelink
2b
Re: Yahoo adjusted close prices by "Randy Harmelink" rharmelink
2d
Re: Yahoo adjusted close prices by "Randy Harmelink" rharmelink

Messages

Sat Oct 1, 2016 12:17 pm (PDT) . Posted by:

"Randy Harmelink" rharmelink

I get results similar to the previous for UVE on 2017-05-19:

UVE UVE UVE
2017-05-19 2017-05-19 2017-05-19
C C C
Y G OX

$12.50
$15.00
$17.50
$20.00
$22.50 $22.50
$25.00 $25.00
$30.00 $30.00
$35.00

Number of puts and strikes shouldn't matter, since the function returns one
or the other. Although OptionsXPress returns the same strike prices for
both puts and calls. Yahoo only returns one strike price. But that's all
they have on their web page right now, so that's to be expected.

I doubt it will have more when the market is open.

I've had a limit price on an option at my broker for a call that doesn't
exist on Yahoo. I placed that limit order several weeks ago. I doubt it
will ever show up on Yahoo. :(

Make sure your Internet Options are set to always get a fresh web page when
one is requested, then try again by running the smfForceRecalculation macro.

On Sat, Oct 1, 2016 at 1:45 AM, Jonathan jbr863@
​...wrote:

>
> Odd. The specific stock I was testing is "UVE" @ "5/19/2017". I still
> get nothing for any of the sources. Is the problem related to the uneven
> number of put and call strikes available?
>
>

Sat Oct 1, 2016 12:43 pm (PDT) . Posted by:

peter.oliver

Thanks Randy. The difference between the add-in result from the CSV download and what the web displays in this specific case was huge: 30.78 to 28.13 => 8.6% less.

Sat Oct 1, 2016 1:54 pm (PDT) . Posted by:

"Randy Harmelink" rharmelink

Right. My point was that it wasn't even accurate after ONE adjustment.

It could be their process echoes that after EVERY adjustment, making it
worse and worse over time.

Yahoo used to have a problem with the CSV file because the data was in two
decimal places. That meant after decades of adjustments, the adjusted
closing price might be just pennies. It's impossible to have adjusted
closing prices with that level of accuracy. It's just not possible to
adjust a $0.03 adjusted closing price for a $0.15 dividend on a $10 stock...

On Sat, Oct 1, 2016 at 12:43 PM, peter.oliver@
​...wrote:

>
> Thanks Randy. The difference between the add-in result from the CSV
> download and what the web displays in this specific case was huge: 30.78 to
> 28.13 => 8.6% less.
>

Sat Oct 1, 2016 5:05 pm (PDT) . Posted by:

peter.oliver

Randy, the amount of dividends paid by ABBV from 02 Jan, 2013 to 02 Oct, 2016 is $6.99/share in 15 payments. If the close price on 02 Jan, 2013 is $35.12 then the correct adjusted close price is $28.13 like the web page is displaying.

I checked half-dozen of other dividend payments and the adjustments are correct on the web page. I assume that they are subtracting the amount of paid dividends backwards since the end date to the closed prices one by one according to the definition of the adjustment since there was no split.

You calculated it obtaining a ratio of closed price less dividend by closed price on the first day after dividend payment and multiplying it to all history, and it should be repeated for each dividend. It seems that the difference gets bigger and bigger for longer periods of time. I'm also using it for stocks bought in 1999 and the difference is huge.

Sat Oct 1, 2016 6:31 pm (PDT) . Posted by:

"Randy Harmelink" rharmelink

It MUST be done using ratios.

So you've identified the difference. Yahoo must be getting the DISPLAYED
data and the CSV data from different vendors, who do the adjustments
differently. That's a shame. They didn't do they due diligence on data
quality of the two vendors.

If they just subtract the dividends, eventually the adjusted closing price
would become negative. For example, look at this garbage for IBM:

http://finance.yahoo.com/quote/IBM/history?period1=-252349200&period2=-228848400&interval=1d&filter=history&frequency=1d

So which of the following sets of data would you trust:

Date Open High Low Close Adj Close* Volume
1-Oct-62 -52.4 -52.4 -52.57 340.37 -52.57 1,848,000
28-Sep-62 -52.44 -52.36 -52.44 353.5 -52.4 1,416,000
27-Sep-62 -52.37 -52.31 -52.46 350 -52.44 2,880,000
26-Sep-62 -52.19 -52.18 -52.37 355.5 -52.37 2,051,200
25-Sep-62 -52.28 -52.19 -52.32 369.25 -52.19 2,440,000

Date Open High Low Close Volume Adj Close
10/1/1962 353.499734 353.499734 340.250266 340.374756 1848000 1.357997
9/28/1962 350.250006 356.499767 350.250006 353.499734 1416000 1.410362
9/27/1962 355.499983 359.750247 348.50024 350.000238 2880000 1.3964
9/26/1962 369.249737 369.749987 355.499983 355.499983 2051200 1.418342
9/25/1962 362.000263 369.249737 359.000266 369.249737 2440000 1.4732

On Sat, Oct 1, 2016 at 5:05 PM, peter.oliver@
​...wrote:

>
> Randy, the amount of dividends paid by ABBV from 02 Jan, 2013 to 02 Oct,
> 2016 is $6.99/share in 15 payments. If the close price on 02 Jan, 2013 is
> $35.12 then the correct adjusted close price is $28.13 like the web page
> is displaying.
>
> I checked half-dozen of other dividend payments and the adjustments are
> correct on the web page. I assume that they are subtracting the amount of
> paid dividends backwards since the end date to the closed prices one by one
> according to the definition of the adjustment since there was no split.
>
> You calculated it obtaining a ratio of closed price less dividend by
> closed price on the first day after dividend payment and multiplying it to
> all history, and it should be repeated for each dividend. It seems that the
> difference gets bigger and bigger for longer periods of time. I'm also
> using it for stocks bought in 1999 and the difference is huge.
>
For the Add-in, Documentation, Templates, Tips and FAQs, visit http://ogres-crypt.com/SMF

Tidak ada komentar:

Posting Komentar