Mike,
Regarding EPS rank....
I don't know how either of the sources calculate their information, but O'neil does mention what factors go into the EPS rank calculation (not the formula). He cites EPS growth on both a long and short term basis (3-year and quarterly), acceleration in EPS and something called earninings stability. Stability seems like possibly a line fit vs. standard error of the EPS numbers. All these factors go into the EPS rank calculation.
From IBD's website:
"Earnings Per Share (EPS) Rating
Exclusive rating found in Investor's Business Daily's SmartSelect® Corporate Ratings. Stocks are rated on a 1 to 99 scale (with 99 being best) comparing a company's earnings per share growth on both a current and annual basis with all other publicly traded companies in the William O'Neil + Co database. Stocks with EPS Ratings of 80 or above have outperformed 80% of all publicly traded companies in earnings. The EPS Rating combines each company's most recent two quarters of earnings-per-share growth with its three- to five-year annual growth rate."
You have to ask yourself what they do for EPS numbers that went negative and then went positive. For instance if you had -0.10 EPS one quarter and then the next quarter you had 0.01 EPS you would theoretically have a gain of (0.01-(-0.01))/(-0.01)*100% = -200%, which is negative even though clearly earnings improved. If you use the absolute value of the denominator then you'd get +200%, but earnings have clearly not tripled 3*(-0.01)=-0.03!
You also have to consider what form of earnings each source uses for their data. GAP Accounting, diluted earnings, etc. IBD uses something different than most other sources - I forget what its called.
Because I have no idea what their formulas are I developed my own ranking system using Excel's PERCENTRANK() function and it does tend to turn up many of the same stocks as IBD and/or Stocktables.com. I've also found similar inconsistencies to what you've found between my own ranking and that of IBD or Stocktables. I was using Reuters Powerscreener to get all my fundamental data.
What it boiled down to for me was that if you use a consistent methodology on what you've researched or have read is "good" and also come up with some way around the zero crossing issue that I mentioned above any of the rankings are probably just about as good as another.
My normalized ranking research has showed that sales growth seems to be a better pointer to future returns and profit margins and ROE are also important.
The goal for the Jack Hershey method is to cull the huge universe of stocks down in to a managable set of "quality" stocks. There are so many ways to define a "quality stock" depending upon who you are. If you look simply for a record of high earnings growth and acceleration +25% or +50% in Q/Q and Y/Y eps growth I think you'll get the right level of quality from an EPS growth perspective and you won't have to agonize over what ranking is better.
- ace
Regarding EPS rank....
I don't know how either of the sources calculate their information, but O'neil does mention what factors go into the EPS rank calculation (not the formula). He cites EPS growth on both a long and short term basis (3-year and quarterly), acceleration in EPS and something called earninings stability. Stability seems like possibly a line fit vs. standard error of the EPS numbers. All these factors go into the EPS rank calculation.
From IBD's website:
"Earnings Per Share (EPS) Rating
Exclusive rating found in Investor's Business Daily's SmartSelect® Corporate Ratings. Stocks are rated on a 1 to 99 scale (with 99 being best) comparing a company's earnings per share growth on both a current and annual basis with all other publicly traded companies in the William O'Neil + Co database. Stocks with EPS Ratings of 80 or above have outperformed 80% of all publicly traded companies in earnings. The EPS Rating combines each company's most recent two quarters of earnings-per-share growth with its three- to five-year annual growth rate."
You have to ask yourself what they do for EPS numbers that went negative and then went positive. For instance if you had -0.10 EPS one quarter and then the next quarter you had 0.01 EPS you would theoretically have a gain of (0.01-(-0.01))/(-0.01)*100% = -200%, which is negative even though clearly earnings improved. If you use the absolute value of the denominator then you'd get +200%, but earnings have clearly not tripled 3*(-0.01)=-0.03!
You also have to consider what form of earnings each source uses for their data. GAP Accounting, diluted earnings, etc. IBD uses something different than most other sources - I forget what its called.
Because I have no idea what their formulas are I developed my own ranking system using Excel's PERCENTRANK() function and it does tend to turn up many of the same stocks as IBD and/or Stocktables.com. I've also found similar inconsistencies to what you've found between my own ranking and that of IBD or Stocktables. I was using Reuters Powerscreener to get all my fundamental data.
What it boiled down to for me was that if you use a consistent methodology on what you've researched or have read is "good" and also come up with some way around the zero crossing issue that I mentioned above any of the rankings are probably just about as good as another.
My normalized ranking research has showed that sales growth seems to be a better pointer to future returns and profit margins and ROE are also important.
The goal for the Jack Hershey method is to cull the huge universe of stocks down in to a managable set of "quality" stocks. There are so many ways to define a "quality stock" depending upon who you are. If you look simply for a record of high earnings growth and acceleration +25% or +50% in Q/Q and Y/Y eps growth I think you'll get the right level of quality from an EPS growth perspective and you won't have to agonize over what ranking is better.
- ace