Basis my limited affair with DCF, the limitation that I more often than not find is related to being precise with the inputs -rather with all combination of inputs. With multiple variables to be assumed/modeled, the end outcome may culminate into an altogether different shape with every decision being selected/omitted. In short, it becomes an effort to get ‘Objective’ output with the help of ‘subjective’ inputs.
Deepak (@deevee) - Think you have just touched upon an excellent tool/perspective to refine the entire puzzle called DCF based valuation!!!
What I understand/assume is that you are talking about plotting ‘valuation points’ generated out of different possible input/assumption combination to generated Gaussian curve and from thereon deriving a DCF fair value range - say PlusMinus 1 Std Deviation (or some other pre-defined confidence interval).
As said, this looks to be an interesting perspective backed with proven, well-accepted logic. Sort of finding a method in the madness (for lack of a better expression). Will love if you can share more details around this approach. When and how this approach can be applied? Assumptions if any? supporting back testing etc etc.
You can consider utilizing any of the existing threads (link, link2) around DCF or can initate a new one for more engaged discussion.
Thanks in advance,
Tarun