Decile Performance

One of the most compelling arguments in TLBTBTM was the table where JG took all the stocks eligible for MFI and ranked them and then tracked their returns by decile. The table clearly showed a straight stair step function with each decile having a lower return than the previous decile.

Guess what? I can do the same thing now. So I have taken about 2900 stocks as of September 20th and ranked them by earnings yield and return on capital. So the million dollar question is whether we will see in a year that stair step function?

Here are the results for the 1st month plus. Note I am making a simplifying assumptions that dividends in each decile are roughly the same. I can not move this to Google Docs as they can only handle like 250 stock prices in a worksheet.

Row Labels	Average of Change
1	11.0%
2	9.1%
3	12.1%
4	10.0%
5	9.8%
6	10.3%
7	11.0%
8	13.1%
9	11.4%
10	12.2%
Grand Total	11.0%

Hmm, looks pretty random to me so far. I will do my best to let everyone know about how this works out. It will be hard to keep perfect as you might expect, with almost 3000 stocks and the potential for stock splits and companies getting bought etc. Already I found two 1:3 splits, and I just caught those because the stocks were up 300%.