Message Font: Serif | Sans-Serif
 
No. of Recommendations: 0
that AmSurg is a screaming buy. I use statistical models that account for seasonality of sales to project future revenues and earnings. Warranted Market Value is based on demonstrated growth rate and AAA bond yields. The formula was developed by Ben Graham, Warren Buffett's finance professor.
For AMSGA, I project the next report will be
Rev - $34.4M
Net - $1.5M
WMV - $21.75

My 12 month target is $27.

If you're statistically literate, the revenue linear regression has an R^2 of .99, a very tight fit.

IMHO
JAGAD5
Print the post Back To Top
No. of Recommendations: 0
Jagad5,

Thanks for your analysis, but a little over this novice's head. AMSGA has had quite a run up in past few weeks. Still a "screaming buy?" Also, in the spirit of TMF, clarification of your analysis would be greatly appreciated for educational purposes.

Thanks
Print the post Back To Top
No. of Recommendations: 0
I use the past 12 months of sales and eps data to estimate what the future performance might be. My analysis accounts for seasonal behavior. AmSurg just had their "slowest" quarter of the year, which ahs historically produced about 24% of their revenue. The spring quarter produces ~26%. In this case, it's not such a big deal but the model works pretty well on toy companies and other's with big swings so I trust it for many cases.

Once I seasonally adjust the past data, I use linear regression to forecast what the next few quarters sales will be. I use the eps data to measure the net profit margins to get net income and then earnings per share.

I use a formula developed by Ben Graham to estimate a fair PE for the stock and that's where I get the fair price. The formula is

PE = 4.4*(8.5+2*G)/Yaaa
G is the growth, 20% growth is entered as 20.0
Yaaa is the yield on AAA rated bonds, 8.125% is entered as 8.125. This 4.4/Yaaa adjusts for the changes in interest rates. The yield on AAA bonds was 4.4% when Graham built the model.

Before I buy, I check to see how well the model fits. This is measured using a statistical function called the correlation coefficient. It can be interpreted as the % of the variation that is accounted for by the model. In the AmSurg case, I get R^2 of .99 meaning that 99% of the variation in the past sales is being treated by the model. In laymen's terms, the model fits like a glove for this stock.

I have been using this model for several years and it has a good track record of finding undervalued stocks. Sometimes it finds them way too soon but so far, it has rewarded me very well. Right now, it is highlighting a number of automotive supplier firms as undervalued. They may stay that way for awhile but if they continue to grow at 20+% per year, they won't continue to trade at PE's of 6. Look at TWR and ARM.

Questions, comments and derogatory remarks are always invited. Thanks.

IMHO
JAGAD5
Print the post Back To Top
No. of Recommendations: 1
I'm inpressed with your formula. Tell me, what do you think of MTH,CANI,and SYMM? Mike TT
Print the post Back To Top