I have a simple valuation model that has picked some massive winners for me and, just as important, kept me away from some horrid losers.

Years ago (when I had hair and a god-like body), when I knew even less about the market than I do today, I started wondering about price to sales ratio as a tool to find value stocks. P/S ratios are all over the place for different stocks, so I parked that idea and went back to guessing on valuation of stocks.

Then one day, I opened my weekly Valueline package and found an interesting inclusion. It was a typical Valueline page with 10 years of history (for those of you who are familiar with Valueline, these data pages are real eye charts.) This one was different, however -- it was actually a compilation of 800 operating companies with all the data weighted as though these 800 companies were one mega company.

It occurred to me that I was looking at a sizable piece of the U.S. economy represented as a single company. With that in hand, I could easily determine the price to sales ratio as it applied to the 800 company segment of the economy. It turned out that the P/S for that mega company was .8.

Then I looked at the net margin for this mega company. It was very close to 5%. I then looked for a multi-year growth rate for the mega company. It was also very close to 5%.

So, the epiphany was that the "average" company earned about 5% net profit, and grew at about a 5% annual rate, and had a price to sales ratio of .8.

The individual companies that made up the "mega company" had price to sales ratios in a wide range of .05 to as high as 10. Obviously, most of these companies varied from the average. What caused the extreme range of P/S ratios? The two biggest factors that altered the P/S ratio were net profit percentage and annual growth rate percentage.

I first discovered that holding the growth rate percentage constant and grouping companies by net profit margins for each company produce a linear curve for share price based on net profit percentage. If a company earned 10% net earnings, its stock price tended to be twice that of a company that earned 5% net profit margin. That linear relationship held for companies that earned 1% net profit on sales through companies that earned up to 25% net profit on sales. So, for example, a company that earned 5% per year had a stock price that was .8 times its annual sales per share, a company that earned 10% per year had a stock price that was 1.6 times its annual sales per share.

The stock price premium or discount attributable to variations in growth rate was not such a straight line relationship. Zero or negative growth really hit the stock price, as it should. On the other end, extremely high growth rates are unsustainable, so the multiplier attributable to growth rate is not linear, and it flattens out at about a 2 multiplier.

Since the Valueline data sheet had 10 years of data, I was amazed to find that the values didn't change much over the years. A company with 5% earnings and 5% growth sold for 80% of revenue -- plus or minus about 10% -- in good times or not so good times.

So here is the table and formula that I constructed to determine a fair value for virtually any stock.

A particular net profit margin percentage produces multiplier 1, a particular sales growth percentage produces multiplier 2.

Multiplier 1 times multiplier 2 will produce an expected price to sales ratio for the company being examined.

To determine a "fair" share price for a company being examined the calculation is multiplier 1 times multiplier 2 times annual sales divided by the number of shares outstanding.

Net Profit Margin% | Multiplier 1 | Sales growth rate% | Multiplier 2 |

1 | .16 | -5 | .25 |

2 | .32 | 0 | .5 |

3 | .48 | +5 | 1 |

4 | .64 | +10 | 1.22 |

5 | .8 | +15 | 1.39 |

10 | 1.6 | +20 | 1.53 |

15 | 2.4 | +25 | 1.7 |

20 | 3.2 | +30 | 1.8 |

25 | 4 | +35 | 2.19 |

Example:

ABC, Inc. earns 10% net after taxes, is growing at 15% per year, has sales of $1 billion per year, and 50 million shares outstanding.

From the table, a 10% earnings rate gives a 1.6 multiplier, a 15% growth rate give a multiplier of 1.39. So 1.6 times 1.39 = 2.22. 2.22 times $1billion in sales = $2.22 billion. $2.22 billion divided by 50 million shares = stock price of $44.40. If the actual market price of ABC, Inc. is $30 per share, it is likely to increase toward the $49 calculated price, if the actual market price of ABC, Inc. is $60 per share, it will be at risk to decline toward the calculated $49 per share.

Current P/S ratio, shares outstanding, sales growth, and trailing twelve month sales are all available on Yahoo finance under "Key Statistics."

Now, this isn't the whole answer, but it is a remarkably good quick screening device that will pick some candidates for further due diligence. It works for underpriced stocks (longs), as well as overpriced stocks (shorts). For example, I bought puts on First Solar (FSLR) and Netflix (NFLX) when their stock price was many multiples of the calculated fair value and made hundreds of percents on those puts. My short list now includes Arm Holdings (ARMH) and Salesforce.com (CRM).

The other thing that this formula is good for is some "what if" analysis. If you know that sales growth or net margin on one of your favorite stocks is about to take a jump (or a dump), you can at least make an estimate of the effect on the stock price.

There, what do you think?

I'm going to pick some widely covered stocks and run the numbers:

Apple (AAPL): NM = 26.67%, GR = 27.2% sales = $156 billion, shares outstanding = 939 million. From the table, multiplier 1 is 4.26, multiplier 2 is 1.75. So, the calculated share price should be 4.26 times 1.75 times $156 billion divided by 939 million shares = $1238/share. Now, can Apple maintain those margins and growth rates? My opinion is no. If net margin went to 10% as smartphones commoditize, Apple would be worth $465/share. Lowered net margin would also reduce sales and sales growth rate, further reducing the calculated stock price. I would stay away from Apple -- too many moving parts.

Microsoft (MSFT): NM = 21.71%, GR = -7.9%, sales = $72.4 billion, shares outstanding = 8.42 billion. From the table, multiplier 1 = 3.47, multiplier 2 = .5. So the calculated price is 3.47 times .5 times $72.4 billion divided by 8.42 billion shares = $14.91/share. With the release of Window 8, Microsoft will return to positive growth, which would double the calculated price to about $28, currently where it is selling. No compelling reason to own Softie.

Cisco (CSCO): NM= 17.46%, GR = 4.4%, sales = $46 billion, shares outstanding = 5.29 billion. From the table, multiplier 1 = 2.79, multiplier 2 is .95. So, the calculated price is 2.79 times .95 times $46 billion divided by 5.29 billion shares = $23.07/share. At a current price of $17.29, CSCO appears to have a 33% upside potential with no obvious show stoppers. I might buy the stock and sell the 60 day out 19 calls.

ARM Holdings: NM = 27.5%, GR = 20.3%, sales = $885 million, shares outstanding = 459 million. From the table, multiplier 1 = 4.4, multiplier 2 = 1.54. So, the calculated price is 4.4 times 1.54 times $885 million divided by 459 million shares = $13/share. The market price is actually $31.68. By this model, the stock is over priced by 140%. I am thinking of buying 25 puts for Jan 2014.

A little favorite:

GIII Apparel Group (GIII): NM =3.82%, GR=9.4%, sales $1.29 billion, shares outstanding = 20 million. From the table, multiplier 1 = .61, multiplier 2 = 1.2. So the calculated price is .61 times 1.2 times $1.29 billion divided by 20 million shares = $47.25/share. With a current price of $35.69, GIII appears to have a 32% upside potential. A little secret about this company is that when a sale to a retailer doesn't sell through, the manufacturer usually gets stuck with a return, which has to be liquidated at pennies on the dollar. GIII has an outlet division that takes the returns and sells them for near retail. Buy and sit on it for a year. Stop loss at $30.

Comments()