The number one lesson of stock valuation: you are always wrong. To put it statistical terms, when you estimate the current intrinsic value of a firm’s equity – that is the present value of future cash flows to the equity holders – your error term is necessarily large. So to deal with that, we do plenty of sensitivity analysis and scenario analysis and see what kind of value range we can obtain under a variety of circumstances.

One of the exercises I do that I have found to be particularly useful is to reverse engineer the stock valuation to obtain the market-anticipated growth rate. I do this because that’s a number that is much closer to a real value with which we can argue. If someone puts a valuation on Facebook that says Facebook is over/under/correctly valued, that’s one thing. You have to really dig into to the analytics to understand that. But, if I say to you “the market is expecting Facebook to grow by 50% over the next two years” that’s something real you can get your head around immediately.

I use a DCF model to back out growth rates. The DCF looks like so:

Market Price = D(1)/(1+r) + D(2)/(1+r)^2 + … + P(N)/(1+r)^N

where D(t) is the cash flow to equity holders at the time period t, P(N) is the terminal value of all cash flows after time N, and r is the required return. P(N) is:

P(N) = D(N+1)/(r-g)

where N is the last period we directly calculate the cash flow to equity holders (maybe 3 or 5 years away), and then we expect those cash flows to grow at a constant rate g forever after that. The g should be similar to the overall economy’s growth rate.

The main place where growth shows up is going from D(1) to D(2) to D(3) and so on. Normally, we calculate growth rates and then see what kind of valuation we end up with. My way is to start with the market price and find the growth rates that make that market price work. Then I argue with the growth rates rather than the valuation.

Some of you option fanatics out there might realize that these are basically the same principles underlying the implied volatility measure. One takes the option prices as given, plugs in all the option parameters, and finds the volatility that makes that option price work. Then, you argue with the volatility, not the option price.

Once I figure out how I can do this, I plan to post some of my modeling exercises here. So if you’re interested, stay tuned!