How is Yahoo data adjusted for dividends?
Author: gregg
Creation Date: 4/13/2009 9:51 AM

#### gregg

#1
Hi,
Is there a guide online on how Yahoo data is adjusted for dividends? On Ex-div date is price just adjust down by the dividend amount for all data before the ex-div date?

Thanks,
Greg

#### Cone

#2
The User Guide, Data > Data Manager > Create New DataSet > Data Providers > Yahoo!

#3
Thank You!

#### ss161

#4
i found the following explanation on this site. it's what i do in code to fidelity data now:

this is the direct link to CRSP:

QUOTE:

Yahoo adjusts the close (factor) based on the standards laid down by the Center For Research In Security Prices: http://www.crsp.com/

The guidelines can be found by searching the site using "dividend adjustment", as the criteria, or by going to the Data Description Guide (an extract from the document is attached to the end of this post).

For additional discussion on adjustment and an explanation of how AmiQuote calculates the OHL values refer to:

YahooAmiBrokerGroup message: http://finance.groups.yahoo.com/group/amibroker/message/93631

AmiBroker Knowledge Base article: http://www.amibroker.com/kb/2007/08/04/amiquote-and-free-data-from-yahoo/

Example 1 - Agilent Technology (A) dividend:

In this example, the dividend on the 1st November, 2006 , is the first split event in recent history i.e. on 2/21/2008, the date of this post, (the Close and the Adj Close are the same up until the first ex dividend day).

For the dates preceding the split, the raw prices are adjusted by Amiquote during the download process (the adjustments, for 10/31/2006, are marked in yellow).

Note: the volume (marked in green) is not adjusted.

The adjustment factor, prior to the split date, is a constant = = AdjClose/Close.

The AdjFactor, at the dividend date = = (Close - dividend)/Close = = (35.60 - 2.057)/35.60 = = 33.54/35.60 = = 0.94

example 2 - Caterpillar Inc (CAT) 2:1 split:

In this case the split is deep in the history and prior adjustments have been made for other split events i.e. the AdjClose and the Close are different prior to the split of 7/14/2005 (refer to the attached CRSP extract that contains information on compound adjustment factors).

Note: As before the Volume, marked in green, is not adjusted. The adjustment factor prior to the split, marked in blue, is halved on the split date.

#### Cone

#5
That's one way to do it too, but it's not the way the Wealth-Lab's Y! Provider does it, and, it's not as accurate because Y!'s AdjClose loses precision after every cash distribuition. We calculate the adjustment factor using the equation 1 - dividend/previous close. Like splits, it is applied to both prices and volume (inversely) so that the product (Price * Volume) does not change after adjusting the data.

#### ss161

#6
i'm probably going to embarrass myself with an incorrect display of math, but, aren't they the same thing:

that is, doesn't ( Close - dividend )/Close == 1 - dividend / close

#### Cone

#7
That's right, but since your pseudo variable was AdjClose, I thought you were using Yahoo!'s value for AdjClose instead of calculating it for yourself. That's the one that loses precision, but frankly, for all practical purposes I'm sure it's quite good enough.