| Re: Discounting anti-lulz
First, for full disclosure, I don't use DCF. In my opinion, DCF is the biggest piece of **** and it doesn't work in any way. Second, most people project, on average, out 5 years. The least being 3 and the most being 10. Once you reach the end of the period you slap on a perpetual growth rate (so in a 5 year model this would be 6+ years) using the CF from year 6. Then you PV that perpetual flow. The thing is though is that more often than not most of the final value will be perpetual growth. (I am just talking about a simple 1-stage, as you know things can be more complex with 2 or 3 stager)
The formula for the discounted perpetual growth cash flow (that you then add to discounted cash flow of Y1-Y5) I have in the spreadsheet I designed is:
(CF Y6*(1+PGR))/(WACC-PGR)
where PGR = perpetual growth rate
Now for why DCF doesn't work. First, the calculations are complex. Forecasting cash flow, net debt and working capital requirements isn't happening. Second, the value you get never sounds right and always comes out high. Third, it doesn't make any sense for most decent businesses to value the company on an income metric after capex. For example, a company may make $100 and invest it all in more equipment to earn even higher returns. In DCF, the company wouldn't be worth anything. Fourth, WACC is stupid. It is impossibly difficult and time-consuming to work out with all that beta business which doesn't even represent a good metric of risk. It makes more sense just to look at debt, look at interest expense, look at cash flows and work out what the financial position is like. Or in other words, you should realize there is a big difference between cost of debt which is quite clear and cost of equity is entirely normative. I use a model based on separating the operating and financial parts of the business and it is far more effective. I can send you over a spreadsheet if your really interested in this stuff but the point is DCF isn't great (its useful to understand why though) as it has an intellectual and theoretical logic but almost no practical logic (i.e it would work if you perfect foresight into infinity).
EDIT: just to clear up what you said earlier as well, the only time you would involve the actual market price of the equity would be in comparing it to the output of the model.
Last edited by cr6196; Sep 9, 2011 at 9:52am.
|