I read an interesting post: on "Where Is Ze Moola":
DCF can lead to large mistakes
I 100% agree with his conclusions. If small changes in parameters lead to big differences in the outcome then one should take the results with a (large!) grain of salt. I think that predicting the future 10 years out is anyhow madness, whichever model is used. I think may be the only possible useful application for DCF is when one wants to compare two similar companies with each other.
As an "angel" investor I receive regularly Business Plan where profits are projected in the area of USD 50 million 5 years out, and we did not invest in them .... apparently we think the projections are "slightly" unrealistic. We did invest in some of them, but even if they "only" reach 1/10th of their forecast in Year 5 we would already be very, very happy. To be frank, I don't even pay attention to them at all, I just look how realistic the forecasts are for the first 2 years until the company is cash flow positive (which is already a major feat).
One of my Corporate Governance recommendations is to do away with the "independent" reports, since they are not independent at all: "whose bread I eat, his song I sing". I challenge the authorities (Securities Commission and Bursa Malaysia) to prove me wrong, to give the statistics how many times the "independent" reports did not follow the Majority Shareholders. I can only remember one single case out of many dozens. That alone already proves how unbelievable biased these reports are, and thus how useless they are (actually, they are worse, they are doing real damage to the Minority Investors).
In brochures where the DCF model is used the underlying assumptions are never revealed. Thus, the minority shareholders can never check or challenge the outcome. And since the reports are so biased, one can safely assume that the DCF valuations are also very much biased.
Which leads me to the following recommendations:
- Do away with all "independent" reports: they are very biased
- Do away with DCF valuations in all circulars: small changes lead to big differences making them very unreliable
- If (unfortunately) DCF models are still used, at least provide the underlying assumptions so that it can be checked how reasonable they are