All equity valuation models, except perhaps the book value approach, involve making a certain amount of assumptions or estimates. And, thus, it may look like it is harsh to single out and criticize the discounted cash flow (DCF) approach to equity valuation. However, there is a lot of merit in the criticisms and this blog delves deeper into the same.
Key DCF assumptions cannot be validated even post-facto
It is extremely difficult to identify the appropriate tenure for the high-growth period
DCF requires forecasts into much longer periods and the farther the forecast period, the lower the accuracy
Let us look into each of them in detail:
Key DCF assumptions are not observable and hence they cannot be validated
All models make certain assumptions such as sales growth rates and margins. Analysts can validate the accuracy of these models once the company publishes the actuals. If there is significant variation, the variance can be studied and the assumptions for subsequent periods can be recalibrated.
However, DCF models require the following additional assumptions: (i) Cost of equity, (ii) Terminal growth rate, and (iii) Weights for capital components (for WACC). None of these assumptions are directly observable and hence one would never know - even years after the valuation - whether the assumptions used were correct or otherwise.
For instance, if we have taken a terminal growth rate of 5%, how do we ever validate that the industry will grow at 5% till eternity? Similarly, if one were to take the cost of equity, or market risk premium at x%, unless there is an investor survey very close to the time of valuation, it is not possible to identify whether such rates were ever true.
DCF requires identification of high growth period and that is extremely difficult
Often, DCF models are built as a multi-stage model with the first part being the high growth period that is explicitly forecasted and the second part being the exit value or perpetual value.
While it is reasonable for us to agree that every business will eventually mature, it is next to impossible to know when that would be. For instance, several technology companies continue to grow at double-digit growth rates even decades after their listing. But DCF expects analysts to take a call on this high growth period when we are valuing the stock.
DCF requires longer-term forecasts that are likely to be way less accurate
Typical relative valuation uses financial numbers pertaining to the next 12 - 24 months of periods. On the other hand, DCF requires that we forecast the data for every year of the high growth period.
The problem is that, by nature, the farther the forecast tenure, the less accurate they are. For instance, while a company may have some idea about the direction of upcoming festive season sales, they may have no clue about the sales 5 years down the line.
Relying on assumptions way farther into the forecast horizon further reduces the reliability of DCF models.
So, does that mean that DCF models are not to be used at all? We would neither say that. DCF can be used to back-calculate the business performance that is required to justify the valuation to apply a sanity check on whether the current valuation appears reasonable or otherwise. But more on this in a different blog.
Subscribe to our website and follow us on YouTube and LinkedIn to access knowledge-sharing content frequently posted by us.
Comentarios