Introduction and Yahoo Finance Data

My name is Jason Ruspini and the pseudo-eponymous host has invited me to post here from time to time.

This first post is going to be boring, but in terms of setting foundations, Eddy Elfenbein has noticed an apparent error in Yahoo's stock data adjustment. When going back in time, the widening of the spread between the adjusted and raw close data implies unrealistically high dividends and returns.

Eddy thinks the dividends may not be accounting for stock splits correctly. When you glance at the data however, the dividends seem to be adjusted for splits. Otherwise the numbers would be roughly double before the split on 2/5/08. Looking at some other sources and trying to replicate Yahoo's numbers, the issue became clear: their calculations and dividends are ok, but their unadjusted data for this stock is wrong. Over at Google, the data is right, but it's only adjusted for splits, not dividends, which isn't very useful for return analysis.

I don't know how prevalent this issue is with Yahoo, but if you don't have access to premium data, I have created an example spreadsheet to cobble together correct unadjusted and adjusted data from Yahoo and Google. Populate columns A through F as shown, and Column M will contain the right adjusted prices. It's just an example however, as long series will be unwieldy in the Google Docs spreadsheet. None of this may be important for short histories, but a lot of people use Yahoo data and Eddy's observation had been nagging me. In the next post, I'll describe a rebalancing strategy that will help tame the value destruction problem of leveraged ETFs.

Update: In preparing that next post, I noticed that Yahoo dividend information was incomplete for IYF.