Can we take it one step further? If the stock had free cash flow of $2.80 a share and was going to grow 4% a year forever – would that be enough to know the stock was worth buying?
There are two basic ideas here. One is a value based approach and one is a return based approach. Exactly how you calculate intrinsic value – or any value – is open to debate. And exactly how you calculate a forward return is open to debate. Do you use reported earnings? Or owner earnings? Or free cash flow? Or free cash flow plus growth?
There are a lot of possible questions to answer. I’d like to focus on the most basic one. Should you think in terms of a price and value relationship (like a $40 stock price compared to a $60 intrinsic value) or a forward return relationship (like a 10% free cash flow yield and a 3% growth rate)?
Someone recently emailed me a question that made me look at how I actually approach investment decisions. The answer is that I split the difference. I use both approaches to some extent.
I don’t do intrinsic value calculations. However, I do use value calculations of some sort. What I tend to focus on is conservative “floors” above which I think the actual intrinsic value must be. For example, if I think a stock is clearly worth more than tangible book value – I’ll use tangible book value as a floor below which I think the stock can be purchased and you’ll know you’re getting a good deal.
I use “private owner value” or what the company would fetch in a takeover the same way. If I think a stock like Ark Restaurants (NASDAQ:ARKR) would probably go for about 7 times EBITDA in a negotiated transaction for the entire company (just based on what other publicly traded restaurant stocks have been acquired for in the past five years or so) then I’d feel I was getting a good deal as long as I paid less than five times EBITDA for the stock.
That’s not the same as making an intrinsic value calculation. It’s a lot different from a DCF. When I think of this more theoretical idea of intrinsic value – I tend to think of an investor like Bill Miller.
A true intrinsic value approach – using something like a DCF – is obviously the theoretically correct way to value a business. If you knew a stock had a free cash flow yield of 5% this year and it could grow free cash flow by 10% a year every year for the next decade or two – you could come up with an actual discounted value of what the company’s stock is worth today.
I can’t argue with that concept. But I also can’t say I’ve found it very useful in practice. I’ve never felt comfortable with that approach.
The biggest reason why I’ve never embraced this approach is that I think it can mislead you about your margin of safety. Let’s say a stock is trading at 15 times earnings and earnings and free cash flow are roughly equivalent at this company. Furthermore, this is a wonderful business that can grow sales, earnings, etc., without adding capital to the business. We estimate that for a long period into the future – we’ll use 20 years in this example – the company will grow its top line and bottom line by 10% a year. And all this while the company will be completely capable of paying out everything it earns in dividends, or buying back stock, or making acquisitions to fuel further growth.
It’s clear we have a 16% to 17% return potential (1/15 = 6.67%; 6.67% + 10% = 16.67%) in this stock for the next 20 years. It sounds like a good stock to buy. And it’s going to sound like a good stock to buy regardless of whether we do an intrinsic value calculation, a forward return calculation, etc.
But I think the forward return calculation will be a bit more honest. Using that approach, we would see the practical truth. This is not a cheap stock. It’s trading for a “normal” 15 times free cash flow. It is, however, a fast grower – 10% a year is about twice as fast as most big companies can grow – and it doesn’t need reinvestment to grow. Those are the key points. And they are easy to see – separately – in a forward return calculation. In this approach, we think we’re paying a fair price for a great business.
A true intrinsic value approach gives us an even clearer idea of what price we should pay for the stock. But I think it loses some subtlety. Let’s take a look at why.
Imagine we think – as I pretty much do – that the S&P 500, bonds, etc., are not going to return more than 7% a year over these next 20 years. If we’re being honest – and not just slapping a big, arbitrary risk premium on our discounted cash flow calculation – we should really use that 7% as our discount rate. Here’s what happens if we discount a 10% grower trading at 15 times earnings using a 7% discount rate.
We can easily end up with an intrinsic value that puts a stock earning about $1 a share today – and trading for $15 a share – around $72 a share. That happens because the present value of the company’s earnings in 2033 would be a lot higher – $1.72 a share – than the $1 it is earning today. In other words, the company would be having more and more valuable earnings every year because it is growing earnings by 10% a year versus say 7% for the best alternative.
Is this a problem?
In a theoretical sense, it’s really not. A company trading at 15 times earnings today that really will grow 10% a year – without needing to reinvest those earnings – really is worth a huge premium over the S&P 500. That’s because the S&P 500 will grow a lot slower (like 6% a year) over the next 20 years while always needing to reinvest some (like one-third) of its earnings just to keep that slower pace.
So, when the S&P 500 looks like its trading at 15 times earnings, it’s really trading at closer to 25 times what it can pay out this year and still grow at something like 6% a year.
There is a real and huge difference between that kind of business – a normal business – and a business that can grow 10% a year beyond its reinvestment needs.
Of course, a business that grows 10% a year for 20 years without requiring reinvestment is rare. So, this kind of extremely high valuation – in this case, we’re being told that paying more than 30 times this year’s earnings would actually be getting the stock at a 50% discount to intrinsic value – is rare even using a DCF.
But the worrying part for me is when you combine the theoretically valid idea of a DCF with the usually practical idea of a margin of safety.
I don’t for a second believe you are actually getting a 50% margin of safety when paying 30 times earnings for a 10% grower – even if that 10% grower has no reinvestment requirements.
In that situation, you are betting entirely on growth. There is no fallback position. If you miscalculate growth – assuming a 10% annual growth rate when the reality turns out to be 5% a year – your intrinsic value estimate plunges from $72 a share to $35 a share.
A halving of the growth rate causing a halving of the intrinsic value certainly sounds reasonable. And it is if we just focus on the value of the company apart from the analysis required to understand its future.
But when we look at the future as uncertain – we realize something. The line between an estimate of a future growth rate (for the next 20 years) of 10% and 5% is really a very fine line. In fact, it may be a lot finer than the line between a 0% and 5% growth rate.
Putting aside situations like the last two decades in Japan, most investors over the last century have gotten used to inflation. Often the rate of that inflation is in the 1% to 6% range over the long-term. Population growth in a country like the U.S. – which is higher than in some other developed and even a few developing (like China) countries – is 1% a year. Growth in real output per person is rarely going to be more than let’s say 1% to 2% for very long.
You can easily see how we can imagine a company with stable market share and stable real prices over time to manage nominal growth of anywhere from 2% to 9%. We have to layer a series of high long-term assumptions about real growth per person, growth in the number of people, and inflation to get anywhere near a 9% growth rate. On the low end, we can still imagine a few percent of nominal growth even in pretty stagnant economies as long as we are looking out decades rather than quarters.
The range from say 0% to 6% is probably a lot easier to understand than the range of say 7% through 12%. In my experience, this is even more true than it appears in theory.
The reality is that as a company’s growth prospects improve enough where the industry is likely to experience nominal growth of say 7% to 12% – remember, this means doubling every 6 to 10 years – the danger of competition becomes more serious.
You can look at industries like cars, television and cell phones during the periods where they were capable of such regular upward growth. In all cases, the number of companies that flopped as investments during this period was a lot higher than it is in industries where growth is in the low single digits.
The number that matters is owner earnings growth. And unless you are very, very certain the company you are analyzing has a wide moat the risk of sales growth and owner earnings growth decoupling gets a lot higher as the industry grows faster.
At very high growth rates, this is easy to see. I did some backtesting of 15% and 25% growers. Peter Lynch – who was investing in times of much higher inflation – mentioned that he avoided companies growing too fast. In his time, anything over 25% would clearly be too fast.
Well, the interesting thing when I backtested consistent 15% growers and consistent 25% growers is that the 15% growers performed better over long forward holding periods (like 5 to 10 years) than the 25% growers. And this often wasn’t due to a stock price difference. The 15% growers actually ended up slowly overtaking the 25% growers because they simply kept growing at a steady pace for longer.
Part of this may have to do with return on assets. A lot of fast growers are able to grow quickly because of high returns on assets. High returns on assets – when those returns are then retained to fuel further growth – tend to lead to decelerating future returns on assets. The two rules of maintaining high returns on assets are that 12% is often a lot more sustainable than say 24% and that retaining half your earnings is often a lot more sustainable than retaining all your earnings. There have been some amazing historical exceptions. They tend to have local advantages and spread across the country from town to town.
For these reasons, I think an intrinsic value calculation is often more of an “upside” calculation than a good gauge of safety. Even stocks where it appears you are buying at a 50% or 75% discount to intrinsic value may not really be safe.
A true margin of safety comes either in the form of virtually guaranteed growth (like Warren Buffett’s belief in Coke’s ability to grow daily consumption per capita in existing markets back in the 1980s, or a monopoly’s ability to raise prices) or in some conservative estimate of a value floor.
Measures like net current asset value, tangible book value, the fair market value of assets like real estate and takeover values are better measures of a margin of safety than future cash flow-based approaches.
Future cash flows often tell you what your upside is. But I don’t think they are a good choice for measuring your downside.
When you buy a stock, the first question is the downside. It is always more important to know what your margin of safety is than to know what your best case returns might be.
I prefer a two-part approach. I like to think about annual returns in the future. But I also like to think about a value floor that is already present today.
Ask Geoff a Question