In my Notes on Warren Buffett’s 2011 Letter to Shareholders, I mentioned I’d give you my thoughts on the intrinsic value of **Berkshire Hathaway (NYSE:BRK.A)(NYSE:BRK.B)**.

Well, here they are:

For me, the three key variables in the Berkshire Hathaway intrinsic value equation are:**1. ****Investments per share****2. ****Non-insurance earnings per share****3. ****Expected growth**

This is pretty similar to what my intrinsic value formula would be for any company with two exceptions:**1. ****Berkshire Hathaway is a financial/non-financial hybrid****2. ****Berkshire Hathaway doesn’t pay a dividend****Dividends, Growth, and Value**

Obviously, a company that earns $1 a share and pays an 80 cent dividend is worth more than a company that earns $1 a share and pays no dividend – assuming both have similar long-term growth prospects.

That last part is key. I don’t have some dividend fetish. But I do recognize that a company that is growing while paying a dividend is a much better investment than a company that is growing while retaining all of its earnings.

There’s nothing wrong with retaining all of your earnings. I have nothing against Berkshire doing exactly that. But it does mean that if I expect Berkshire to grow 10% a year while paying no dividend and I expect Microsoft to grow 10% a year while paying a dividend – then they should not trade at the same P/E ratio. The company that pays a dividend deserves the higher P/E ratio, because it is growing just as fast while providing present day cash. And cash today is worth more than cash tomorrow. Now, of course, this doesn’t mean a company growing 20% a year while retaining all of its earnings is worth less than a company growing 10% a year and paying a dividend – in fact, the faster growing company is probably worth more.

I bring up the dividend issue because I’ll be discussing Berkshire as a big cap growth company in a moment. That means comparing it to other companies in that category. And a lot of big cap companies pay dividends. Berkshire doesn’t. So it’s best to compare Berkshire’s growth in intrinsic value to the growth of companies who also retain all of their earnings.**How to Value Berkshire: Book Value or Earnings Per Share?**

The right answer is neither. People always say: *“Berkshire is trading at a price-to-book ratio of 1.2 or 1.3. Historically, it’s traded at this ratio. That means you should buy, sell, hold”* – whatever. That kind of analysis isn’t ideal. It works in very, very broad strokes.

So, for instance, we know that Berkshire Hathaway is worth more than book value. Warren Buffett has said this many times himself. And I don’t think he’s wrong. Berkshire is definitely worth more than its book value (including intangibles). The quality of some of its business is quite high. This includes some very big companies. So, you can be confident that if you get the chance to buy Berkshire below its book value – which is currently $95,453 – you’re getting a bargain.

But that doesn’t really tell us much about Berkshire’s intrinsic value per share. So what does?

Basically, there are two ways to calculate intrinsic value per share. You can use an earnings (or sales) based approach to value a company. Or you can use a book value (or ROE) based approach to value a company.

In general, I prefer using an earnings/sales based approach to value non-financial companies. And I prefer using a book value/ROE based approach to value financial companies.**Valuing Financial Companies**

There’s also a catch with any ROE approach for financial companies. This sounds confusing. But it makes sense if you think about it. Over long periods of time, a financial company – a bank, insurer, or investment holding company – has an earning power equivalent to the **higher** of its return on equity or its growth in book value per share.

Why is that?

Why can’t we just use the company’s return on equity?

Here’s the deal. Sometimes there is a persistent difference between a company’s reported earnings and its economic earnings. This can happen to any business. For non-financial companies the area of special concern is the ratio of free cash flow to reported earnings. Some companies – like **IBM (IBM)** – consistently have higher free cash flow than earnings per share. While other companies – like **Norfolk Southern (NSC)** – consistently have lower free cash flow than earnings per share.

Warren Buffett talked a bit about this in his letter. He mentioned that Burlington Northern will have free cash flow that is less than its earnings for many, many years to come because it spends more money replacing its old assets than it expensed in depreciation while those assets were in use. The reason for this is inflation – the assets cost more to replace at the end of their life than they did when the company first bought them.

That’s an important topic. But not for today. Today, we are focusing on Berkshire Hathaway. And here the mismatch that matters is not the mismatch between earnings and free cash flow – it is the mismatch between net income and comprehensive income.

If you go to GuruFocus’s 10-Year summary page for Berkshire Hathaway, you’ll see a row called *“Return on Equity”*. If you average the ten figures for 2001 through 2010 you’ll see that Berkshire Hathaway’s 10-year average return on equity is 7.6%. However, if you take the book value per share from one year – say 2002 – and subtract the book value from the year before and then divide the difference by the earlier year’s book value you’ll see that those return on equity numbers don’t match Berkshire’s change in book value. Sometimes, Berkshire’s change in book value can be 10% while its reported return on equity is a measly 1%.

Which number matters more?

For Berkshire Hathaway, it’s the change in book value. That’s because Berkshire’s reported earnings – other than earnings from its non-insurance companies – doesn’t accurately reflect Berkshire’s economic earnings. Berkshire marks its investments to market. So, Berkshire’s book value fluctuates with changes in the price of: IBM, Coca-Cola, Wells Fargo, etc. But Berkshire’s reported earnings do not reflect these changes.

So we want to focus on the roughly 10% growth in Berkshire’s book value over the last 10 years. That gives us a rough approximation of two things. One, it obviously gives us a rough approximation of Berkshire Hathaway’s expected future growth.

But, secondly, it gives us an idea of what Berkshire’s economic return on equity is. Basically, if Berkshire’s book value is $100,000 this year and $110,000 next year we can – very roughly – say that Berkshire’s economic earnings were about $10,000.

This is something that gets overlooked a lot. You have to value Berkshire either by putting its parts into two buckets: **1. ****Investments per share****2. ****Earnings per share**

And valuing each of those buckets. Or, you have to translate Berkshire’s investment assets into a stream of earnings or its non-insurance businesses’ earnings into assets. What you can’t do is compare apples and oranges. You can’t measure both parts of the company using the price-to-book ratio or both parts of the company using the price-to-earnings ratio.

Using that very rough approximation of Berkshire’s economic earning power – which I pegged around $10,000 a share – we can say that Berkshire at $120,000 is valued a lot like other companies trading for about 12 times earnings.

That’s one way of doing it. But it’s imprecise. And it obscures the two sources of value. **Intrinsic Value: The Warren Buffett Way**

I prefer actually breaking Berkshire Hathaway into two parts – the same way Warren Buffett does in this year’s letter to shareholders:*“Though Berkshire’s intrinsic value cannot be precisely calculated, two of its three key pillars can be measured. Charlie and I rely heavily on these measurements when we make our own estimates of Berkshire’s value. The first component of value is our investments: stocks, bonds and cash equivalents….Berkshire’s second component of value is earnings that come from sources other than investments and insurance underwriting. These earnings are delivered by our 68 non-insurance companies…In Berkshire’s early years, we focused on the investment side. During the past two decades, however, we’ve increasingly emphasized the development of earnings from non-insurance businesses, a practice that will continue…There is a third, more subjective, element to an intrinsic value calculation that can be either positive or negative: the efficacy with which retained earnings will be deployed in the future. We, as well as many other businesses, are likely to retain earnings over the next decade that will equal, or even exceed, the capital we presently employ. Some companies will turn these retained dollars into fifty-cent pieces, others into two-dollar bills.”*

So Warren Buffett believes there are 3 key parts to calculating Berkshire’s intrinsic value per share:**1. ****Investment Per Share****2. ****Earnings Per Share****3. ****Capital Allocation**

I don’t want this to become Geoff’s subjective thoughts on Warren Buffett, Berkshire Hathaway, etc. I want to keep it about the calculation rather than my opinions. So, I’m going to take part #3 – capital allocation – and just say that Warren Buffett will grow Berkshire Hathaway’s intrinsic value per share by 10% for the foreseeable future.

As it turns out, this is a reasonable estimate. Berkshire has delivered book value growth around 10% over the last decade. Stock market valuations and earnings levels were pretty high 10 years ago. We can argue about that. But I don’t think anyone would accuse me of using peak earnings and peak P/E multiples on Berkshire’s businesses and stocks by using today’s numbers. If Berkshire was able to grow book value 10% a year over the last 10 years – which weren’t especially great times to grow a holding company – then Berkshire has a shot at doing the same over the next 10 years.

Offsetting this is Berkshire’s immense size. As a company gets bigger, growth gets harder. This is true at Berkshire. But I think – if we mean this to be an honest rather than conservative intrinsic value calculation – that a 10% growth rate at Berkshire is reasonable from this point forward.

Now, we come to two indisputable facts. Berkshire’s investments per share and earnings per share – as laid out by Warren Buffett himself – are:**1. ****Investments Per Share: $98,366****2. ****(Pre-Tax) Non-Insurance Earnings Per Share: $6,990**

To simulate what Berkshire’s non-insurance businesses would produce in after-tax earnings on their own, we’ll multiple $6,990 by 0.65. This assumes a 35% corporate tax rate. The result is $4,544 a share.

So, now we can think of Berkshire as consisting of $4,544 in after-tax earnings from its wholly owned businesses. And a pile of investments that amount to $98,366.

If we assumed Berkshire could get a return of 10% a year on its investments – we could translate everything into earnings. One-tenth of $98,366 is $9,837 a share in capital appreciation, interest, etc. from the investment side. Together, those two streams of earnings – capital appreciation/interest/etc. and earnings from Berkshire’s businesses – would total $14,381 in economic earnings per share.

This would mean buying Berkshire stock today – now trading at $120,000 – is equivalent to buying a stock trading around 8 times earnings.

I think that is too aggressive. For a bunch of reasons. One, there are some taxes associated with Berkshire’s investment portfolio. Although much of the equity portion can be deferred indefinitely – it’s still not tax free. Unlike float, taxes tend to actually get paid out at some point. Berkshire can’t stop companies it owns from being acquired. Real taxes will be paid on many of those investments sometime in the future. Also, I think 10% is slightly on the aggressive side of reasonable when it comes to Berkshire’s portfolio. I think Berkshire will grow intrinsic value about 10% a year in the future – but it’ll do that through a combination of earnings growth and investment growth. And the earnings growth could be faster.

So, I prefer simply taking Berkshire’s $98,366 in investments at their stated book value and then capitalizing Berkshire’s $4,544 in estimated after-tax earnings from its non-insurance businesses.

What is the right multiple? For me – and this is just me – a company that grows 10% a year while retaining all of its earnings is worth 13 times earnings. I won’t bore you with exactly how I got that number. (Basically, I took what I thought a no-growth business is worth and then I looked at the difference between what Berkshire can grow its value at and what a AAA bond would return each year). If you think 12 times earnings is the right number for a company growing 10% and paying no dividend – that’s fine. If you think 15 times earnings is the right number – that’s fine too. If you think the right multiple is 10 times earnings or 20 times earnings – I will argue with you there.

I don’t think a company that pays no dividend is worth 20 times earnings if it is only growing at 10% a year. And I definitely don’t believe a company that is growing 10% a year is worth only 10 times earnings. In my book, a no-growth company is still worth around 8 times earnings – so you’d only be adding a couple points to the P/E ratio for 10 whole percentage points of growth. That seems wrong.

The right range is somewhere in the neighborhood of 15 times earnings. I think it’s 13 times earnings. But if you say 12 times or 16 times – I won’t insist I know better.

What matters more in deciding whether or not to buy Berkshire is how likely 10% growth is versus 5% growth or 15% growth. For example, if Berkshire only grows its intrinsic value by 5% a year over the next 10 years – it’s definitely not worth 13 times earnings today.

So, let’s go ahead with my idea of what the market should pay for a company growing its value by 10% a year while paying no dividend. I said 13 times earnings was an honest multiple for a company like that.

So, let’s apply it. Berkshire has $4,544 in non-insurance earnings. And $4,544 times 13 equals $59,072 a share. That’s a good estimate of what Berkshire’s non-insurance businesses are worth. Add this to the $98,366 in investments and you get $157,438.

Okay. Is that Berkshire’s intrinsic value?

Well, there is another way to look at Berkshire. Let’s say that Berkshire is worth at least its book value per share. That’s $95,453 a share. Next, let’s say that Berkshire Hathaway will grow its book value at 10% a year. In this way, we can use Berkshire’s annual increase in its per share intrinsic value as the company’s earnings.

This might sound strange. But it makes sense. If we know a company is worth at least book value and we know book value is increasing by 10% a year – we have a minimum idea of what that company’s economic earnings per share are.

It’s:

(Year 2 Value – Year 1 Value) = Earnings

The increase in value over the span of a year is essentially Berkshire’s annual earnings. In this case, we said we were confident that book value per share was a conservative proxy for intrinsic value. So, intrinsic value is at least $95,453. And we expect book value to grow 10% a year. So 10% of $95,453 is $9,545 a share in *“earnings”*.

You can use this as your P/E ratio. If Berkshire is trading at $120,000 a share that is $120,000 / $9,545 = 12.6 times earnings.

I think that is about right. In fact, I think that Berkshire is likely to add a bit more than $9,545 a share to intrinsic value each year because Berkshire’s intrinsic value is greater than its book value.

If we use the same 12 to 16 times earnings range we mentioned before, we’d come up with an intrinsic value of anywhere between $115,000 a share and $153,000 a share.

Overall, if we used the most aggressive – too aggressive for me – estimate of over $14,000 in normal earnings for Berkshire and applied my 13 times multiple we’d get $182,000 a share. I think the earnings number is a bit too high there. The multiple of 13 may be about right. But you could certainly convince me 15 times earnings is the right multiple for Berkshire.

You can see how most of the difference in estimates of Berkshire Hathaway’s intrinsic value comes from different views of what is a fair price-to-earnings multiple to pay for a company growing as fast as Berkshire Hathaway while retaining all of its earnings.

For me, the range of reasonable intrinsic values for Berkshire falls into an area between about $120,000 a share and $180,000 a share.

My own personal preference – what I would pick if I had to pick just one number – is $160,000.

That is basically the number I got when I applied a P/E of 13 to Berkshire’s estimated after-tax non-insurance earnings and added their capitalized value - $59,000 – to Berkshire’s investments per share of almost $100,000.

Now, some of you may think that adding investments per share – without subtracting Berkshire’s *“float”* – is too aggressive. I’ve never believed that was the case.

Basically, I think Berkshire Hathaway shouldn’t be valued using a liquidation approach. I think Berkshire Hathaway should be valued as a growth company.

Berkshire’s insurance operations sometimes make money. In fact, they earned an underwriting profit in each of the last 9 years. I’ve never assumed that insurance would deliver underwriting profits. Instead, I’ve just assumed that insurance would provide Berkshire with a way of carrying all those investments per share at a cost of 0%. In other words, a breakeven result.

I don’t have any special insight into Berkshire’s insurance operations. Some parts have clear competitive advantages – like GEICO. Others – like Berkshire Hathaway Reinsurance – depend on key people. Warren Buffett believes Berkshire’s insurers as a group will average at least a breakeven underwriting result over the long-term. I’m willing to take his word on that.

Once you make that assumption, a valuation around $160,000 a share seems right to me. That’s simply the value of the combination of Berkshire’s earnings from its non-insurance business and the value of the investments Berkshire owns.

At $120,000 a share, Berkshire Hathaway’s shares aren’t trading at a Ben Graham style two-thirds of intrinsic value. But the stock certainly isn’t overvalued. So, if you believe in Berkshire Hathaway and Warren Buffett and want to invest alongside him – buying Berkshire at $120,000 should work out better than buying the S&P 500.

For those looking to plunk down less than $120,000 – Berkshire’s “B” shares go for $80. That’s exactly equivalent to $120,000 on the “A” shares.

So pick whichever share class suits your budget.**Ask Geoff a Question about Berkshire Hathaway’s Intrinsic Value****Check out the Ben Graham Net-Net Newsletter****Check out the Buffett/Munger Bargains Newsletter**

**Also check out:**

### About the author:

*Geoff Gannon
*

**Geoff Gannon**

Pooky- 4 years ago

From the article:"The per-share book value of both our Class A and Class B stock increased by 4.6% in 2011. Over the last 47 years (that is, since present management took over), book value has grown from $19 to $99,860, a rate of 19.8% compounded annually.*"

So, you can be confident that if you get the chance to buy Berkshire below its book value – which is currently $95,453 – you’re getting a bargain.

Geoff - minor point, but where are you getting your book value figure from?

[i][/i]