Why do value investors obsess over finding intrinsic values instead of discount rates (expected returns)?

Question

From what I read about value investing, value investors seek to buy stocks that sell below their "intrinsic value". To that end, they use stock valuation models to find the all-important "intrinsic value". Some of the simpler stock valuation models I read about are: dividend discount model (DDM), discounted cash flow model (DCF), and residual income valuation (RIV). These models produce an output which is the "intrinsic value". However, these models always require a "discount rate" as one of the inputs. This "discount rate" requirement is the source of my confusion.

As far as I am aware, the "discount rate" is merely a speculative guess, and to make matters worse, a slight error in the discount rate will have a large effect on the calculated intrinsic value. The capital asset pricing model (CAPM) isn't going to be helpful in finding the discount rate, because the "market risk premium" is uncertain (and also because most value investors seem to think that the CAPM is nonsense). "Garbage in, garbage out" looks like a really big risk here.

Given the unacceptably large uncertainty of the discount rate which causes an unacceptably large uncertainty in the intrinsic value, why do value investors obsess over finding the intrinsic value?

If it were me, I'd rearrange all these valuation formulas to find the discount rate instead of the intrinsic value. I'd run these models in reverse. I'd substitute the current stock price for the "intrinsic value" to see what "discount rate" (i.e. expected return) comes out of these valuation models. If the expected return is higher than what I would normally accept, then the stock is undervalued. Otherwise, the stock is overvalued. In this scheme, there is no need to think about the "intrinsic value". And yet when I look around, value investors seem to obsess over "intrinsic values"… so I think I might be wrong. Is my proposed method wrong in any way?

Answer 1

You seem to be approaching the problem the same way as you are theorizing a 'value investor' is. Both approaches require some amount of reliance on financial models as they currently exist.

'They' try to find the intrinsic value of a stock, by estimating, say, future dividend cash flows, and applying a discount rate. They then compare the 'intrinsic value' to the stock price, and decide whether to buy or sell.

You are trying to find the discount rate by estimating future dividend cash flows, and working backwards from the current market price. If the discount rate implied is higher than what you think is appropriate [meaning you think the market is assessing the risk of the company higher than what you think it is, because ultimately a discount rate's purpose is to devalue future cash flows that are worth less due to the interest-free rate + the risk of this specific item being valued], it implies that you have found an undervalued stock, which you could purchase. ... but how do you decide the discount rate is appropriate? Isn't that exactly what a value investor has done?

As you indicate in your question, you have just rearranged the formula, not introduced a new variable. If someone assumes a discount rate of 10%, values some cash flows, and determines a company is worth $1B, and each of its 1,000,000 shares is worth $1k and thus a good deal compared to the $950 market price, how is that different from you calculating that the market valuation of $950 implies that the same cash flows are being discounted at 10.5%, which is a good value compared to your assumed fair discount rate of 10%?

It is not that your method is incorrect or anything - it is just another way of looking at the same problem. Intuitively, one method may make more sense to someone than another method, but be aware that you haven't resolved the fundemental issue you identified, which is 'how would I know that 10% is a good discount rate to use for comparison?'.

Answer 2

You are correct that different discount rates will result in drastically different valuations. But there isn't one objectively true discount rate, but this is a personal preference.

Some kind of discounting is always necessary because otherwise the value (= all future earnings) could be infinite. If it is not represented explicitly, other assumptions will have a similar effect. For example, we could alternatively only look at earnings over a fixed duration, e.g. 30 years after which we expect to sell the shares. But this would lead to suboptimal decisions since expected earnings in year 30 are equally important as those in year 1, but expected earnings in year 31 wouldn't matter at all, despite having a huge influence on the value of the stock in year 30. Using some discount rate avoids difficulties of a fixed horizon.

Aside from mathematical considerations, there are also financial reasons why a discounting future earnings could be sensible:

• inflation
• uncertainty, e.g. the company might go bankrupt
• opportunity cost (money now is more valuable than money in 10 years because you could use that money in the meanwhile)

Especially how you weigh opportunity costs is a fairly personal decision. For most people, the correct benchmark isn't “I could get 6% returns in an index fund” but “I could pay down this loan in 10 instead of 12 years” or “I could afford to go on a roadtrip with my dog before she dies”.

You can alternatively express these effects as part of the expected earnings, but it ends up being mathematically equivalent to discounting.

While there is no true discount rate, you can effectively calculate the discount rate assumed by the market as a whole: take some valuation formula, plug in the current stock price, and solve backwards for the rate. Assuming that everyone agrees on other factors such as fundamentals, value investing is then the exercise of finding assets where the market's discount rate differs from your personal discount rate. Closely related: if you don't discount enough, you're going to be the greater fool in the greater fool theory.

Answer 3

Is my proposed method wrong in any way?

I feel the best answer is, yes, in fact as a straightforward matter many traders essentially completely agree with the thrust of the question:

1. as mentioned, every parameter involved is: a guess

2. concepts like "intrinsic value" can only be put in quotes ^†

Again, yes, the fact is there are drastically differing opinions on the nature of markets, and yes the viewpoint expressed in the thrust of the question, is a completely normal viewpoint held by many.

Any such "market epistemology question" such as the one asked here can only be answered by either

• "yes, FWIW 50% of people agree with that and 50% don't",

or, if it's a totally whacky idea (astrology or such) then

• "nobody thinks that, it's whacky".

In the case of "FWIW 50% of people agree with that and 50% don't" (such as the question at hand) each side violently, vigorously, pointedly, endlessly, relentlessly disagrees with the other side.

---

† regarding "intrinsic value" specifically, I've always felt it just remarkably goes against the reality that there's a bid and an ask, and that's literally all there is in markets.