The Science of Hitting

# What Is Intrinsic Value?

July 26, 2014 | About:

What is intrinsic value?

Heres the definition people toss around: It is the discounted value of the cash that can be taken out of a business during its remaining life.

So heres the next question: how many readers have sat down and actually ran that calculation? And if you have, did you walk away feeling confident in your numbers?

My bet is zero on the first question and even if Im wrong, then a resounding no to the second one (if youre being honest with yourself).

At most, I would bet a few people have done the following: they estimated cash flows for the next three to five years, then came up with some estimate for terminal value using the Gordon Growth Model or a market / industry multiple. Ill use a simple example of the second approach:

Lets assume a company will earn \$100 each of the next five years, with a 100% payout ratio; at that point, the company is sold for 18X earnings a terminal value of \$1900 (a year five outflow of \$1800 plus our \$100 dividend). At a 10% discount rate, intrinsic value is just shy of \$1500.

Stop to think about what that means in the simplest terms: the investor who pays ~\$1500 today expects to receive \$100 each of the next four years, and then \$1900 in the fifth year. Rather than thinking in discount rates, I think its more intuitive to think of this as a forward rate of return: I must put up \$1500 today to receive \$2300 five years from now an annual return of ~10% (well not exactly: the math assumes reinvestment of the cash flows as they are received at the IRR, which changes things a bit; lets skip that discussion for now and stick to the point at hand).

I think this is a critical point to grasp: discount rates can be viewed as nothing more than required rates of return. When we start thinking about opportunity costs, this is a more intuitive way of grasping the concept: the return on risk free alternatives creates a benchmark for us to build our return requirements; the assumption of risk must come with a more than commensurate increase in expected return (assuming youre risk averse, as most individuals are).

This quote from Seth Klarman (Trades, Portfolio)s Margin of Safety captures this point succinctly:

A discount rate is, in effect, the rate of interest that would make an investor indifferent between present and future dollars... The appropriate discount rate for a particular investment depends not only on an investor's preference for present over future consumption but also on his or her own risk profile, on the perceived risk of the investment under consideration, and on the returns available from alternative investments.

For me, that eliminates the need for academic answers (like CAPM) and creates a rough, but workable, concept: if I think a stalwart like Berkshire Hathaway (BRK.B) is priced for 10-12% annualized returns for many years into the future, how much more should I require to dip into something like Weight Watchers (WTW)? Academia would tell you the answer lies in the historic movement of their stock prices as compared to the index; I think thats nonsense.

Keynes knew what he was talking about: It is better to be roughly right than precisely wrong.

That same idea works for setting the alternative - in my example, Berkshire Hathaway; is 10-12% the right number? While this fluctuates as interest rates move, Im of the opinion that these levels are in the right ballpark; like a circle of competence, the most important thing is a good feel for where the edges lie. At 5% per annum, no thanks; at 20% per annum, back up the truck.

The lack of specificity is a huge no-no for some. My personal opinion is that this is a microcosm of valuation as a whole: Im willing to trade off the false sense of confidence that comes with certainty (like when someone states intrinsic value is \$47.18 per share) for a mental model that is intuitive and more importantly, cognizant of its shortcomings.

So that helps us with half of the puzzle now how about those pesky cash flows? What can we take out of the business? Some people, in search of precision, model results for the coming five years (then gloss over the fact that the majority of their value estimate lies beyond year five); if you think you can do that successfully, best of luck. This task is difficult not only in its inherent uncertainty, but also due to the impact of behavioral biases: once you take a glance at the stock price, you can toy with the numbers to say pretty much anything you want look no further than our analyst friends on Wall Street, who adjust their models and price targets to derive fair value estimates that are, amazingly, almost always within spitting distance of the current stock price.

Clearly, Im a bit skeptical of those who claim the ability to predict whether operating margins will increase or decrease by twenty-five basis points a few years down the road; if were not going to take that approach, what other options do we have?

Again, Keynes words shouldnt be forgotten: It is better to be roughly right than precisely wrong.

Rather than focusing on the numbers for any given year, lets think about the broader concept: what drives intrinsic value? Some math will help us find the answer.

Lets start with a simple case: a company with book value of \$100 per share and a 20% return on equity that pays out 100% of earnings every year. Without any reinvestment, were looking at a perpetuity with \$20 paid to investors a year; with a requirement for 10% per annum in returns, that means intrinsic value is equal to \$200 per share or 2X book. If we adjust our ROE to 30%, we can pay 3X book and still meet our return requirement (10% per annum, assuming perpetuity). The first point is a relatively simple one: a higher ROE business is worth a higher multiple of book value.

The second concept is a bit trickier: as we start moving away from dividends and towards reinvestment, intrinsic value can jump off the charts. Lets assume a thirty year window for reinvesting 50% of the corporations earnings, at which point it matches the prior example (100% payout): the ability to reinvest over those 30 years back into the core business at a 20% return on equity brings intrinsic value from 2X book in the base case to nearly 5X book; thats a going-in multiple of 10X earnings versus 25X earnings.

So clearly were concerned with two things: the attractiveness of the underlying business (in this case, as captured by ROE), as well as the ability to find attractive reinvestment opportunities (retention ratio); as a long term investor, I believe those two factors are of critical importance.

Now, we get into a more difficult conversation: how do we assess the future opportunities to reinvest and depending on our assessments, how does that impact intrinsic value?

While I plan on doing that, its not going to happen today this article has taken long enough to write on its own! As always, I hope others will share their opinions particularly if they think Ive made an error.

### About the author:

The Science of Hitting
I desire to own high-quality businesses for the long-term. In the words of Charlie Munger, my preferred approach is "patience followed by pretty aggressive conduct." I run a concentrated portfolio, with the top five positions accounting for the majority of its value. In the eyes of a businessman, I believe this is sufficient diversification.

 Currently 4.84/512345 Rating: 4.8/5 (19 votes) Voters:

Batbeer2 - 6 years ago

Hi SoH,

Thanks for the article.

>> how many readers have sat down and actually ran that calculation?

It probably isn't done often enough but there are many who have calculated the cash return of a business in liquidation. In my view, that's as precise as it gets. The model you give is relevant for the more complex case of a business as a going concern.

I think it is fair to say that it is much easier to estimate the intrinsic value of a business in terminal decline. The residual value is 0, so that is one factor that drops out of the equation.

The Science of Hitting - 6 years ago    Report SPAM

Batbeer - Good points; as it relates to that, I personally don't have any experience in those situations. Have you made investments in companies that ultimately ended in liquidation? Did the numbers turn out as expected or were the assets thrown away by managers on the way down? I would be interested in your experience; thanks for the comment!

The Science of Hitting - 6 years ago    Report SPAM

Fast Weekly - glad you enjoy the articles, and hope to see some more of your comments in the future. Just to clarify some of your comments, you said you use 2% over the 10-year or a floor of 5%; do you assume it's static over the entire forecast period? If not, does your measure of intrinsic value adjust for the full rate swing as rates change? Said differently, if the rate goes from 5% to 6%, you believe the company is worth ~15% less? And do you keep the discount rate at 10-year plus two regardless of the underlying security? Interested in understanding your thought process - thanks for the comment!

Batbeer2 - 6 years ago

Hi Science of Hitting,

Just as there is a difference between buying low P/FCF stocks and buying stocks with a high dividend yield, there is also a difference between estimating the liquidation value of a stock and buying liquidation plays.

1) Here's one that I talked about that was trading at a discount to my estimate of liquidation value (Bristow): http://www.gurufocus.com/news/52678

It wasn't liquidated.

2) Here's another (Steinway) that was trading at a discount to liquidation value. That one was eventually liquidated. I haven't crunched the numbers but I think investors eventually recieved about 130% of what I initially estimated. http://www.gurufocus.com/forum/read.php?2,42192,222884#msg-222884

3) More recently I talked about Guiness Peat Group. It has been liquidating its assets and has shareholder-friendly management. http://www.gurufocus.com/news/238272/guinness-peat--value-contest

Even Vodafone, a company I wrote about a couple of years ago, can be analysed from the viewpoint of a slow liquidation. Rightly or wrongly, I have always been more confident when using the assets as a basis to estimate the sum of all future returns.

The main point I'm trying to make is that instead of using cashflow or earnings as a basis for estimating future cash returns, you can use the (fair market value of the) assets on the balance sheet. In my view, the result is less imprecise.

The Science of Hitting - 6 years ago    Report SPAM

Batbeer: Awesome - thank you for providing those links; I'm going to spend more time reading about those investments, but even a quick glance at the events around Steinway in 2012 / 2013 looks pretty interesting :) Thanks again!

Batbeer2 - 6 years ago

If you like to read about stuff like that, AHC is not unlike Steinway.

Jtdaniel - 6 years ago

Hi Science,

Excellent topic and article. My starting point is to use an 8% discount rate for all companies. If the actual interest rate on treasuries exceeded 8%, I would use the treasury rate as my discount rate. For Berkshire (B), I would divide the EPS by 8% (\$8.30 / 0.08) = \$103.75. The 8% can be considered an estimated, initial year rate of return. Now this is rather conservative for a great business like Berkshire, so if I was willing to accept a 7% initial year rate of return I could pay \$118.57 (\$8.30 / 0.07). Using the earnings yield (E/P), I could also calculate a 7% initial year rate of return at the current per-share price (\$8.30 / \$127.55 = 7).

Berkshire's long-term growth rates for EPS, FCF and pre-tax income are all in the 9 - 10% range. I see no reason for a significant decline, especially with \$48+ billion in cash available to great capital allocators. I am confident that Berkshire can grow earnings at 7% for the next 20 years, even if Buffett and Munger soon step aside. I can also get a margin of safety by assuming a low earnings multiple (P/E of 12) at the end of the 20-year period. If current EPS of \$8.30 increases annually by 7% for the next 20 years, 2034 EPS will be \$32.12. \$32.12 X 12 (P/E) = \$385.44. If I buy shares at the prevailing price of \$127.55, I can reasonably project a 20-year compounding rate of return of at least 5.69%.

In order to do better than 5.69% , I need: (1) to pay a lower price per share, (2) Berkshire to grow EPS by more than 7%, or (3) the market to assign a P/E higher than 12 in 2034. Of course, I only have control over variable #1. As 5.69% is not what I am looking for, the prudent decision is to wait for a lower share price.

Hpeterscheck - 6 years ago

Really good article and comments.

Slightly separate question... What order of filtering do you tend to use? By that I mean do you start with, say, "do I think this company will be around in 20-40 year" or do you start more with "do I understand this business" or maybe "I noticed that something about this company is interesting (bad/good newsm shows up on a screen, etc).

I've be een taking different approaches over time and I find myself gravitating pretty strongly to #1, but it has odd consequences. For example, while TJ Maxx looks really good in many ways, it's hard for me to confidently let it pass the "20-40 year" test. That's not to say that I don't think it'll be around... Rather... The confidence I have in my ability to know that is highly variable compared to, say, walmart for which it is quite a bit higher. But even Walmart is lower than, say McDonalds which is lower than Coca Cola.

Its odd because I'm finding it hard to quantify that measure and yet it's the one that provides me with the most differentiation when it comes to very long term investments. Of course price is still critical, but I can wait :).

i think this ties in directly with a DCF analysis which as you guys say is more important at years 5-10 than 1-5 and thus as the time horizon goes further out my answer to the 20-40 year question increases in priority.

As a result I'm building up a list of companies I think have a very high 20-40 year success probability and using that as an initial filter. No idea if that's smart or not and it's always a process in transition.

does this make sense? Still pretty new at investing in a more methodical way and trying to learn from people who have been doing it much longer and much more successfully.

The Science of Hitting - 6 years ago    Report SPAM

Jtdaniel - Interesting approach; so using your math on Berkshire, what's your required rate of return before you would pull the trigger? Are you saying if it got to 8% then you would be a buyer? Thanks for the comment - look forward to your response.

The Science of Hitting - 6 years ago    Report SPAM

Petershk - Thanks for the kind words! I think what you're talking about was addressed by Alice Schroeder in an interview a few years back; when discussing a potential investment that Warren had turned down in the 1950s, here is what she had to say about his analysis:

“He said no because he went through the first step in his investing process, and this is where I think what he does that is very automatic [to him] but isn’t very well understood: he acted like a horse handicapper. And the first step in Warren’s investing process is always to say, 'What are the odds that this business could be subject to any kind of catastrophe risk that could make it just fail?' – and if there is any chance that any significant amount of his capital could be subject to catastrophe risk, he just stops thinking. No. And he won’t go there.”

I think in order to have confidence in the out years, as you highlight, this must be top of mind; especially when you're valuing going concerns with an expectation for outsized ROE's (trading at a multiples of book), you are likely to see a serious impairment of capital if the moat is diminished / eliminated. As you note, that doesn't lend itself to quantification; alas, that's definitely an early part of my analysis.

Thanks for the comment!

Jtdaniel - 6 years ago

Hi Science,

Yes, I would buy Berkshire at an 8% discount rate -- \$103.75. That would be pretty consistent with Berkshire's buy-back policy. In fact, I used this approach in September 2003 and still have those shares. The split-adjusted per-share cost was \$50.80 (\$2,540 / 50:1 split). My 11-year, annual compounded rate of return is 8.73%. A similar calculation can be run using growth in equity (using historical ROE) and projecting P/B down the road. For Berkshire, it confirms that 8% is a reasonable figure.

The Science of Hitting - 6 years ago    Report SPAM

Jtdaniel - Perfect, thanks for the clarification!

Hpeterscheck - 6 years ago

Science,

yes that's exactly it. I liked her book a lot and I hadn't read the Berkshire letters before and now read them + the original partner letters. What I took away from it was that many of buffet's mistakes were because of that... Even Berkshire itself. Even in the 60s it doesn't pass the "will this company be successful in 40 years" test... And probably why Buffett never approved management's turn around plans ;).

i think about my own big mistake 18 months ago with jcp. In some ways it looked cheap and lots of great investors piling in. I violated my "don't buy big debt" rule but worse, I violated the "will it exist in 40 years" filter. The reason is more that now I have to stress between "take loss" or "ride out." Because I don't have confidence in long term survival I don't have confidence in valuation and suddenly I'm a speculator. I would contend no one knows what jcp is worth because you can't see even 2-5 years with any confidence... Even liquidation valuation is suspect because if jcp dumps it's real estate on the market, that will have an impact there, plus you don't know inventory value discount, etc. so the mistake isn't buying jcp or not,,. The mistake is not knowing what I was doing :).

im in tech and this exactly prevents me from buying Microsoft and Apple. They are great companies that have been around a while now.... But someone tell me that when you think 20-40 years from now whether you'll use iOS or android or windows, office or google drive or something else... Xbox or iPad mini. iPhone or galaxy. It's impossible to say... I'll probably be drinking coke, eating Doritos and heading to mcdonalds with my grand kids from time to time.

The Science of Hitting - 6 years ago    Report SPAM

Petershk - Very much agreed with all you've said, particularly as it relates to JCP; thanks again for the comments and hope to hear from you again in the future.

Rdj1234 - 6 years ago    Report SPAM

Hi TSOH

I'm not very math saavy and I was wondering if you could show me the math for the two following examples so I can experiment with them myself.

Let’s start with a simple case: a company with book value of \$100 per share and a 20% return on equity that pays out 100% of earnings every year. Without any reinvestment, we’re looking at a perpetuity with \$20 paid to investors a year; with a requirement for 10% per annum in returns, that means “intrinsic value” is equal to \$200 per share – or 2X book. If we adjust our ROE to 30%, we can pay 3X book and still meet our return requirement (10% per annum, assuming perpetuity). The first point is a relatively simple one: a higher ROE business is worth a higher multiple of book value.

The second concept is a bit trickier: as we start moving away from dividends and towards reinvestment, intrinsic value can jump off the charts. Let’s assume a thirty year window for reinvesting 50% of the corporation’s earnings, at which point it matches the prior example (100% payout): the ability to reinvest over those 30 years back into the core business at a 20% return on equity brings intrinsic value from 2X book in the base case to nearly 5X book; that’s a going-in multiple of 10X earnings versus 25X earnings.

Another thought provoking article as per your usual

Thank You

Roger

The Science of Hitting - 6 years ago    Report SPAM

Roger - Glad you liked the article; send me an email at thescienceofhitting at gmail dot com and I'll send you some material that will help. Thanks for the comment!

AlbertaSunwapta - 6 years ago    Report SPAM

There's always been a lot of talk about intrinsic value but never any rigourous testing and presentation of the concept that I've come across. Can anyone draw a picture of a company's lifetime intrinsic value? When I first encountered the idea of intrinsic value, I wondered what a series of intrinsic values for a company over it's lifetime might look like. Would it form a nice parabolic curve from growth to maturity to death. That thought lasted two seconds.

So, I've raised this somewhere else and someday hope to see results (I'm too lazy to do it myself), but what I have yet to see is a historical analysis of the daily, weekly or monthly intrinsic value of a company out of the past - maybe one that pretty much died a typical death. One could look at a company that was taken out while still healthy but that final return could be more misleading. I'd prefer one where the final assets truly became just "salvage" value or say, if necessary, entering Chapter 11 status.

So you could look at that company's situation and calculate periodic intrinsic value estimates assuming certain future events were unpredictable and calculate the error against the perfect hindsight calculations. You could also plot it against it's own trading prices as well as an index to see what insights general market movements might provide. )(For instance, looking at Polaroid on a historical intrinsic valuaiton basis, when and why was it a good time to buy or sell it, based on intrinsic valuation calculations vs. market prices, and not just it's price rising with the tide over time?)

And one more comment. I've liked the idea of buy and hold but in buying something below intrinsic value, one is seeking to obtain outperformance as the price reaches intrinsic value. So, for the most successful intrinsic value investors that buy and hold indefinitely as long as the company is growing, their outperformance is obtained early and then continued holding yields just an average return. So why buy and hold? Why not, as some managers do, sell when a security reaches intrinsic value. Why not then shift to a market index or ETF and earn a market return, then wait for another underpriced security to appear? Is it hoping for a market pricing error where a sale price far above intrinsic value can be obtained? Or that one was wrong in their growth estimates and that they underestimated the security's prospective growth and so underestimated intrinsic value at purchase. (i.e. Good things come to those who wait.)

Stocdoc - 5 years ago    Report SPAM

Hi,

Could you please send me the math that you were going to send to Roger.

Thanks.

Stocdoc