Someone who reads my articles asked me this question:

That’s a great question. And a very hard one to answer. Let’s start with the easy part: do I use a DCF?

No. I never use a

Why not?

A DCF is a pretty complicated and subjective approach that draws your attention away from the variables that matter. I’ve said before that the things to focus on – the variables to use in your analysis – are numbers that are:

Constant means the numbers you use should have some amount of durability. Don’t use last year’s sales growth rate. Can you use the 10-year sales growth rate?

Probably.

You wouldn’t want to do that for a homebuilder or a company that is in the business of selling some commodity that has seen tremendous price growth over the last 10 years. But you could certainly incorporate long-term growth in a market, technology, population, etc. It’ll be easier to do this for some companies than for other companies. Focus on the companies where you know what matters and where the things that matter can be measured.

Any attempt at valuation involves assumptions on your part. One good rule: don’t project faster growth over the next 10 years than the company achieved over the last 10 years. It’s not enough to add a big margin of safety to the end of your appraisal process. You should incorporate conservatism into every level of your analysis. So, always assume growth is at least a little bit slower as the company ages. The bigger a company gets, the more likely it becomes that its rate of growth will be lower in the future than it was in the past.

Consequential means you only want to focus on numbers that matter. Don’t worry about variables that don’t move the needle. Focus on things like free cash flow, earnings, dividends, and book value. And on the long-term trend in sales, EBITDA, and book value. For companies where book value doesn’t matter – don’t use it.

When valuing a company based on its ability to generate cash earnings – focus on free cash flow, earnings, EBITDA, and sales. When valuing a company based on its ability to grow its asset value – focus on book value, earnings, revenue, and leverage.

The top line matters at (almost) all companies. So always pay attention to sales and sales growth.

Usually, these are the things that matter. But sometimes other things matter more. Obviously, there are sometimes hidden assets, court cases, promising new drugs, etc. that don’t appear in a company’s financial statements at all. If you’re looking at a company where one of those things overshadows the usual financial metrics – then toss the usual financial metrics out. Focus on what matters. Use common sense. Always keep the company’s value to a private owner – a 100% buyer – at the front of your mind.

Calculable just means that the things you focus on need to be things you can measure and play around with. Numbers you can “crunch”. Numbers you can put in simple formulas.

For example, an operating margin is calculable when you pair it up with a price-to-sales ratio and a tax rate. There’s usually not a lot of uncertainty about those figures, so margins can be very helpful in any intrinsic value estimate – to the extent you think the margins are stable.

What’s not calculable?

If a company grew 35% a year over the last 10 years – there’s not much you can do with that information. Sure, it matters. And sure you can adjust that rate down in the future. But isn’t that just being arbitrary. Are you really using the growth rate at all if you do that? Couldn’t you just as easily pluck a rate from thin air?

It’s very hard to quantify something like 35% revenue growth. I always bring this point up when people want to know my view of

Intrinsic value estimate are tough. In fact, they’re tougher than what’s usually need to make oodles of money in specific stocks.

The best bargains are often obvious. A net-net is obvious. A great insurance company selling below tangible book value is obvious. A wide moat, growing, consistent company selling at 11 times earnings is obvious.

Buying stocks like those can make you rich. And buying them doesn’t require making an actual intrinsic value estimate. It just requires recognizing that the stock’s price is lower than some conservatively calculated value.

Coming up with a truly

That’s a lot to ask.

There are some formulas out there. You can use a DCF. GuruFocus has a perfectly good DCF calculator you can use. After you type in a stock’s ticker symbol,

That’s the biggest problem with most intrinsic value calculations. You have to make some assumptions you’re comfortable with

So, I want to stress that there’s a difference between the rationale for buying a stock – your proof that it’s cheap – and an intrinsic value estimate. Sometimes, an intrinsic value estimate might show a stock is very cheap. And yet you shouldn’t buy that stock.

That’s because the reliability of your assumptions is key.

Like I said, there are many formulas out there. I talk about Ben Graham a lot. So, I’ll show you one of Graham’s formulas:

“V” means intrinsic value. And “g” means growth rate.

Graham intended for this formula to be a way of showing what price a stock would trade at given a level of expected growth (and thus also what level of growth was expected of a stock trading at a certain price). He didn’t intend his formula to be used as the sole criterion for buying stocks. And Graham didn’t propose the formula as some sort of universal intrinsic value equation.

But it’s a decent starting point. It’s a formula you can look at for any stock you’re interested in.

Let’s look at Warren Buffett ’s recent purchase of

Here, I’m using IBM’s 10-year sales growth rate of 7% as its expected future growth rate. You can argue with this. But let’s see how it works with that backwards looking number.

Well, 8.5 + 14 is telling us that 22.5 is the right P/E multiple for IBM according to Ben Graham. Does that sound right?

Actually, it sounds a little high to modern ears. But, there’s a dividend issue here. The P/E multiple you’re willing to pay for a stock should depend on its future growth in per share earnings

If a company had a high dividend payout ratio and was growing at 7% a year – a P/E ratio of 22.5 sounds fine to me. On the other hand, if a company was retaining all of its earnings and still only growing its earnings per share by 7% a year – I’m not sure a P/E of 22.5 would make sense.

Later in his life, Graham developed a more complicated formula that incorporated interest rates into the equation. Personally, I find the way he did it kind of clunky and not necessarily much of an improvement. Graham recognized a legitimate concern – the risk that changes in the level of interest rates would cause changes in the level of P/E ratios – but he didn’t do a really good job of addressing it.

If you just Google

There’s a good reason for this. The toughest part of developing an intrinsic value formula is making it

These are the things you can measure. But they don’t allow you to make an apples to apples comparison.

The future cash flows – which a DCF uses – are what matters. Future cash flows are what allows you to make an apples to apples comparison. It’s a universal approach. But it involves a lot of guesswork. You can’t measure future cash flows. You can only project them.

The things you can measure: earnings, dividends, past growth rates, and current interest rates also matter. Personally, I think those are the things to focus on. Because those are things you can measure.

Moving on to a DCF does more to improve your appraisal in theory than in practice. A DCF leaves the toughest part to you. It tells you to project future cash flows.

The hard part of an intrinsic value estimate is getting from the things you know: earnings, dividends, past growth rates, interest rates, etc. to the future you don’t know.

My own view is that it’s better to start on firm practical ground – actual observable data in the present day – than to try to get on the best theoretical ground by using a DCF.

I’m interested in what works in practice. And I don’t think a DCF does.

To me, a DCF isn’t a very useful tool beyond just reminding you of the important principle that cash today is worth more than cash tomorrow.

In most cases, the actual number crunching of a DCF does not leave an investor with a clearer understanding of what a stock is worth.

That’s because the danger in every DCF lies in the assumptions you are making.

*Geoff,**How do you estimate intrinsic value…do you do a DCF or some other method?**John*That’s a great question. And a very hard one to answer. Let’s start with the easy part: do I use a DCF?

No. I never use a

__iscounted__**D**__ash__**C**__low analysis (DCF).__**F**Why not?

A DCF is a pretty complicated and subjective approach that draws your attention away from the variables that matter. I’ve said before that the things to focus on – the variables to use in your analysis – are numbers that are:

**1.****Constant****2.****Consequential****3.****Calculable**Constant means the numbers you use should have some amount of durability. Don’t use last year’s sales growth rate. Can you use the 10-year sales growth rate?

Probably.

You wouldn’t want to do that for a homebuilder or a company that is in the business of selling some commodity that has seen tremendous price growth over the last 10 years. But you could certainly incorporate long-term growth in a market, technology, population, etc. It’ll be easier to do this for some companies than for other companies. Focus on the companies where you know what matters and where the things that matter can be measured.

Any attempt at valuation involves assumptions on your part. One good rule: don’t project faster growth over the next 10 years than the company achieved over the last 10 years. It’s not enough to add a big margin of safety to the end of your appraisal process. You should incorporate conservatism into every level of your analysis. So, always assume growth is at least a little bit slower as the company ages. The bigger a company gets, the more likely it becomes that its rate of growth will be lower in the future than it was in the past.

Consequential means you only want to focus on numbers that matter. Don’t worry about variables that don’t move the needle. Focus on things like free cash flow, earnings, dividends, and book value. And on the long-term trend in sales, EBITDA, and book value. For companies where book value doesn’t matter – don’t use it.

When valuing a company based on its ability to generate cash earnings – focus on free cash flow, earnings, EBITDA, and sales. When valuing a company based on its ability to grow its asset value – focus on book value, earnings, revenue, and leverage.

The top line matters at (almost) all companies. So always pay attention to sales and sales growth.

Usually, these are the things that matter. But sometimes other things matter more. Obviously, there are sometimes hidden assets, court cases, promising new drugs, etc. that don’t appear in a company’s financial statements at all. If you’re looking at a company where one of those things overshadows the usual financial metrics – then toss the usual financial metrics out. Focus on what matters. Use common sense. Always keep the company’s value to a private owner – a 100% buyer – at the front of your mind.

Calculable just means that the things you focus on need to be things you can measure and play around with. Numbers you can “crunch”. Numbers you can put in simple formulas.

For example, an operating margin is calculable when you pair it up with a price-to-sales ratio and a tax rate. There’s usually not a lot of uncertainty about those figures, so margins can be very helpful in any intrinsic value estimate – to the extent you think the margins are stable.

What’s not calculable?

If a company grew 35% a year over the last 10 years – there’s not much you can do with that information. Sure, it matters. And sure you can adjust that rate down in the future. But isn’t that just being arbitrary. Are you really using the growth rate at all if you do that? Couldn’t you just as easily pluck a rate from thin air?

It’s very hard to quantify something like 35% revenue growth. I always bring this point up when people want to know my view of

**Apple (AAPL, Financial)**. My view of Apple is that I can see a lot of its qualities for what they are – but I have a really hard time putting what I know about Apple into numbers. I don’t want to say Apple’s value is incalculable. A lot of people obviously do calculate the company’s value. I just find it very hard to value because I know what matters at Apple – I just don’t feel comfortable quantifying the things that matter and plugging them into some formula.Intrinsic value estimate are tough. In fact, they’re tougher than what’s usually need to make oodles of money in specific stocks.

The best bargains are often obvious. A net-net is obvious. A great insurance company selling below tangible book value is obvious. A wide moat, growing, consistent company selling at 11 times earnings is obvious.

Buying stocks like those can make you rich. And buying them doesn’t require making an actual intrinsic value estimate. It just requires recognizing that the stock’s price is lower than some conservatively calculated value.

Coming up with a truly

**honest**intrinsic value estimate is very hard. A truly honest intrinsic value estimate shouldn’t be overly conservative but it still shouldn’t cause you to buy stocks that will lose you loads of money.That’s a lot to ask.

There are some formulas out there. You can use a DCF. GuruFocus has a perfectly good DCF calculator you can use. After you type in a stock’s ticker symbol,

*“DCF”*is one of the tabs you can click on. Go ahead. Play around with it. It’s a good tool. And a good way to get a feel for how a DCF works and exactly what assumptions are important. It’s also good at showing you how much a few changes to your assumptions can radically change the intrinsic value estimate you get.That’s the biggest problem with most intrinsic value calculations. You have to make some assumptions you’re comfortable with

**and**some assumptions you’re a lot less comfortable with. When you buy a stock – you want it to be because of assumptions you are totally comfortable with.So, I want to stress that there’s a difference between the rationale for buying a stock – your proof that it’s cheap – and an intrinsic value estimate. Sometimes, an intrinsic value estimate might show a stock is very cheap. And yet you shouldn’t buy that stock.

That’s because the reliability of your assumptions is key.

Like I said, there are many formulas out there. I talk about Ben Graham a lot. So, I’ll show you one of Graham’s formulas:

**V = EPS * (8.5 +2g)**“V” means intrinsic value. And “g” means growth rate.

Graham intended for this formula to be a way of showing what price a stock would trade at given a level of expected growth (and thus also what level of growth was expected of a stock trading at a certain price). He didn’t intend his formula to be used as the sole criterion for buying stocks. And Graham didn’t propose the formula as some sort of universal intrinsic value equation.

But it’s a decent starting point. It’s a formula you can look at for any stock you’re interested in.

Let’s look at Warren Buffett ’s recent purchase of

**IBM (IBM, Financial)**using Ben Graham’s formula.**V = $13.25 * (8.5 + 2(7))**Here, I’m using IBM’s 10-year sales growth rate of 7% as its expected future growth rate. You can argue with this. But let’s see how it works with that backwards looking number.

**V = $13.25 * (8.5 + 14)**Well, 8.5 + 14 is telling us that 22.5 is the right P/E multiple for IBM according to Ben Graham. Does that sound right?

Actually, it sounds a little high to modern ears. But, there’s a dividend issue here. The P/E multiple you’re willing to pay for a stock should depend on its future growth in per share earnings

**and**its future dividend payments.If a company had a high dividend payout ratio and was growing at 7% a year – a P/E ratio of 22.5 sounds fine to me. On the other hand, if a company was retaining all of its earnings and still only growing its earnings per share by 7% a year – I’m not sure a P/E of 22.5 would make sense.

Later in his life, Graham developed a more complicated formula that incorporated interest rates into the equation. Personally, I find the way he did it kind of clunky and not necessarily much of an improvement. Graham recognized a legitimate concern – the risk that changes in the level of interest rates would cause changes in the level of P/E ratios – but he didn’t do a really good job of addressing it.

If you just Google

*“intrinsic value formula”*you will be overwhelmed by the number of results. Most of them are attributed to one investor or another – often Ben Graham or Warren Buffett – but none of them are as practical as you’d expect.There’s a good reason for this. The toughest part of developing an intrinsic value formula is making it

*“one size fits all”*. Different companies get their value from different places. Some companies have amazing brands, some own priceless real estate, some have big investment portfolios, others have lots of cash flow which they plow back into the business, others pay all their cash out in dividends, still others eschew dividends entirely and focus on share buybacks, while yet other companies are focused on fabulous sales growth today that will hopefully turn into amazing earnings down the road.These are the things you can measure. But they don’t allow you to make an apples to apples comparison.

The future cash flows – which a DCF uses – are what matters. Future cash flows are what allows you to make an apples to apples comparison. It’s a universal approach. But it involves a lot of guesswork. You can’t measure future cash flows. You can only project them.

The things you can measure: earnings, dividends, past growth rates, and current interest rates also matter. Personally, I think those are the things to focus on. Because those are things you can measure.

Moving on to a DCF does more to improve your appraisal in theory than in practice. A DCF leaves the toughest part to you. It tells you to project future cash flows.

The hard part of an intrinsic value estimate is getting from the things you know: earnings, dividends, past growth rates, interest rates, etc. to the future you don’t know.

My own view is that it’s better to start on firm practical ground – actual observable data in the present day – than to try to get on the best theoretical ground by using a DCF.

I’m interested in what works in practice. And I don’t think a DCF does.

To me, a DCF isn’t a very useful tool beyond just reminding you of the important principle that cash today is worth more than cash tomorrow.

In most cases, the actual number crunching of a DCF does not leave an investor with a clearer understanding of what a stock is worth.

That’s because the danger in every DCF lies in the assumptions you are making.

**Ask Geoff a Question About How to Estimate a Stock’s Intrinsic Value****Check out the Ben Graham Net-Net Newsletter**
## Please Login to leave a comment