Terminal Value Mystery

Many finance textbooks and analyses show that Terminal Value (TV) could occupy or contribute more than 50% of total DCF-based value, and in many cases, it could even find it approximately 80-90% of that estimated value. McKinsey on Valuation (7th Edition, Chapter 14, 2020) displays TV as a percentage of the total value.

I am wondering, how to interpret this? What is the point to do all this hard-ball exercise of projection, if, at the end of the day, most of the value is at the end of that projected period and beyond. The advice is always to extend the projection into more years, but we know there is a danger in doing this. Discount rate probably is not quite impacted in the long run, but cash flow projection, this might as some people call it "voodoo".

Prof. Ivo Welch in his textbook Corporate Finance (4th Edition, 2017) said that in the long run, errors in cash flow projections couldn't be compared with errors in discount rate estimates. In general, terminal value increases in importance with the growth rate of firm cash flows and decreases with the length of the planning period (Titman and Martin (Valuation, 2016)). Many finance authors will recommend that the length of the explicit forecast period will stop when the company business is entering the stable-growth period, and then we could calculate the terminal value under the stable-growth assumption. Before that, the company business is implicitly assumed away to have high growth. However, the question of how long a company business will be able to sustain high growth is perhaps one of the more difficult questions to answer in a valuation (Damodaran on Valuation, 2nd Edition, 2006). McKinsey gave a very general idea about this "entering the terminal period", by saying that choosing an appropriate point of transition depends on the company and how it is changing over time. A company undergoing significant change may require a long, detailed window, whereas a stable, mature company may require very little detail in its forecasts (Valuation, 7th Edition, 2020).

We also do know that most analysts will use very limited methods to assess TV, either (1) Moderate Growth-Gordon Method (or assuming very small growth (see Geoffrey Moore's book Living on the Fault Line using product-life-cycle analysis to tell that the cash flow growth rate somehow will fall under the required rate of return) and also Good to Great by Jim Collins, which it is only below ten percent of the companies being researched that could give a superior return in the long run), or (2) (projected or current) Multiple Method.

?Will that mean it is somehow "not quite right" about using those methods to assess TV? I know we don't have any other tools to do this.

With a bigger chunk of all company value resting on TV, will that mean most of the value to be created in that Terminal Period and beyond? Sounds it does not make business and finance sense. However, at the same time, we also know that value creation has a longer-term view, but this does of course not necessarily mean it will only come up at the end of the Terminal Period and beyond. For example, when a store is opened, the value creation will be there for many years, after going thru its ramp-up period, when visitor traffic starts building up from Day or Year One. Even when it is declining, the company could do many kinds of stuff to keep the flowing of the traffic, by revitalizing the store look and atmosphere, bringing new products, etc.

Having a large TV portion will only mean something, that the company business has future "growth options", not factored into this DCF-based method. For example, the value of Walmart's business will include all those stores that they have not opened yet, in addition to the current stores. Yet, using this logic is also problematic, will that mean we will "punish"?those not-yet-opened stores saying that the investment dollars being spent on that to-be-opened stores will not give a return above the required rate of return (=superior return)? If yes, then it will not add value under capital budgeting analysis. However, we do know that Walmart will keep opening its stores. Expansion is always the main option, if not what else?

?I hope you could follow me?

?Many believe as well that discount rate has a very limited link to Cash Flow being projected, right? The discount rate comes from the investment community, and if using portfolio-based investment, then business risk will not matter so much, except a part of business risk that has a direct contribution to the market risk. More papers now come up that link high equity risk premium with what they call "disaster risk" or "rare event risk". I guess, the book The Black Swan by Nassim Nicholas Taleb, a quant, wrote it beautifully. So far Financial Econometrics is still far from solving the equity risk premium puzzle and even Prof. Sheridan Titman said, we don't even know what is the exact beta, and risk premium further down.

?With the discount rate factoring disaster risk in its calculation, when we are moving further down into the future, the uncertainty is surely getting higher, and how come we could say that we are ok with that 90% of the estimated value and label it as "Terminal Value"?

?Though in a different context, Prof. John Maynard Keynes's words keep buzzing inside my head when I am sitting writing this, ...in the long run, we are all dead.

Some comments via email to me:

Ex-consultant:

I agree with you that TV has "heroic" assumptions and one has to be careful with all of them. I have given my readers and/or firms several options?about what happens beyond the last period. One of them is to assume 0% growth up to several scenarios for different G's. And yet, it's crystal ball guessing.

I have proposed from the most conservative approach that means a perpetuity with G=0% up to an interval of very conservative scenarios with real growth 0% (only inflation that is another?crystal ball guessing to a limited range of real growth). Indeed, it is VERY difficult, even, IMPOSSIBLE to foresee?which innovations, policies, world or local market scenarios could be considered. On one extreme it is also unrealistic that the firm will shut down?in the next N= 5, 7, 10, 12, 15, 20 years or worse, that it will not obtain any economic profit in those future years. I remember a book by [Aswath] Damodaran which showed % of TV on total value and clearly it's crazy to accept that TV be 50% or mor?of actual estimated value of a firm. Usually one can do is to extend the explicit projection period and/or examine % of TV value on total value and adjust Growth and other items in a TV formula in other?to keep TV as a conservative % of total value. How would be that % for you to be satisfied with the calculation? 10%, 20%... 50%. Well, that would be a kind of reverse engineering: set the maximum acceptable % of TV on total value and define the G to get that goal. In some consulting work a few years ago we did that, trying not to have more than 15%-30% of total value attributed to TV. Of course this might be something arbitrary, but you have to show explicitly to the "customer" "buying" the valuation, a kind of scenarios for different assumptions?of G.

I propose this reading : Mejia, F. y Vélez-Pareja, I. (2010). Cost of Capital and Value without Circularity for Constant Growth Perpetuities (https://papers.ssrn.com/abstract=1659446). Give a look to this paper and see if it could be useful for you.

Another comment from international corporate finance and financial modelling trainer and Wiley Finance author on such topic:

1. When investors use the EV/EBITDA they are ignoring the bias in this ratio caused by different asset lives.
2. When using Gordon’s model, they completely ignore the completely obvious fact that higher growth must come along with higher capital expenditures.?It is mind boggling that they can change the terminal growth and not change the capital expenditures in the terminal period.
3. The McKinsey Value = NOPAT x (1-g/ROIC)/(WACC-g) seems to be a good method, but it is very easy to prove that this method does not work when the current ROIC is different from the long-run ROIC.

?I really believe that you can do better by adjusting the value driver formula and explicitly assume gradual growth and ROIC versus cost of capital to the mean.?It is not so difficult.