Is it an insurance company, a holding company, a hedge fund, a conglomerate, a closed-end mutual fund, private equity? The answer depends on whom you ask and when. The annual report says "holding company." The SEC files it under "fire, marine, and causality insurance." I use "conglomerate" in my vernacular, but when it comes to  Berkshire Hathaway (NYSE:BRK-A), I'm just a casual observer.

As most of us know, Berkshire comprises a lot of moving parts. If you count everything -- complete ownership, partial ownership, and passive ownership -- you're looking at least 80 different components covering nearly as many different businesses. Berkshire covers a wide swath of the U.S., and increasingly world, economy.

Book smarts
I'll admit that I've always been intimidated by Berkshire. I've never been sure how to value it -- so much size, so much scope, so much Warren Buffett, so much opportunity for me to screw it up. Other analysts, many I'm sure possessing greater insight and sharper number-crunching skills than mine, are less reserved. 

But many of the opinions they proffer are centered on book value, which I've always found a little odd.

I rarely hear an analyst lead an argument with book value when opining about lesser companies. More often, the analysis leads with P/E multiples, profit margins, earnings trends, revenue growth, new product development. Book value is often a tertiary topic, if it's mentioned at all.

But with Berkshire, it's different: Barron's Jan. 11, 2010 edition featured a Berkshire-centric article, which included a very prominent book-value-centric opinion: "After rising just 3% in 2009, the stock, which is way below its late 2007 peak of $147,000, fetches a mere 1.2 times our estimate of the company's year-end 2009 book value of $84,500 a share -- compared with an average 1.65 times in the past decade. The stock rarely has been cheaper, relative to book value, in 15 years."

This focus on book value is understandable. After all, you could label Berkshire a financial company. What's more, Berkshire has posted a remarkable streak of adding assets that generate persistently higher earnings, which is more than you can say about Dell (NASDAQ:DELL), which trades at a rich 5.3 multiple of book value, or American International Group (NYSE:AIG), which trades at a well-deserved 0.7 multiple, or even perhaps Blackstone Group (NYSE:BX), which trades at a comparable multiple to Berkshire, yet has nowhere near Berkshire's history nor its record of earnings persistence.

Value and potential
But is a low price-to-book value a precursor to a higher share price, as Barron's suggests? In 1995, Berkshire's average price-to-book value multiple was 2.6, more than double what it is today. Berkshire was a much smaller, much more nimble entity then. In 1995, total revenue was a mere $4.5 billion, producing net income of $795 million. By 2000, total revenue had swelled to $34.9 billion, while EPS had more than trebled to $3.3 billion.

You could easily argue that Berkshire deserved the premium, particularly when considering its stock price -- the most important measure for shareholders. Berkshire's shares closed 1995 at $32,100, and then more than doubled over the subsequent five years to close 2000 at $72,400.

But the book value premium eroded along the way. The average book value multiple was only 1.5 in 2000, a much lower multiple compared to 1995.

So how would Berkshire shareholders fare over the subsequent five years? One would think admirably, given the implied value in the company's price-to-book value multiple. In 2005, total revenue grew to $81.7 million, while net income increased to $8.5 billion. Impressive, to be sure, but the share price failed to maintain the torrid pace, increasing only to $88,620.

Today, Berkshire's shares trade around $100,000 each. According to Barron's, they sport a 1.2 price multiple to book value.

Buy, sell, hold, or consider the alternatives
Does that mean Berkshire is undervalued? The price doubled in five years from a high book-value multiple in 1995. The stock nearly doubled from a low book-value multiple in 2005, before trading lower to its current 15% premium.  

As we've learned over the years, it's never a good idea to bet against Warren Buffett, but it is a good idea to be grounded in reality. Reality says this isn't 1995, when Berkshire's book value per share was around $14,000. Today, it's around $84,000 a share. That's a lot more girth, a lot more notoriety, and a lot more potential for disjointed economies of scale.

That said, Berkshire could be undervalued. If you are unsure (like me), or think Berkshire's overvalued and that book-value is still important, here are two holding companies that might be worthwhile surrogates: WR Berkley (NYSE:WRB), with a price-to-book value multiple of 1.1, and Leucadia National (NYSE:LUK), with a multiple of 1. Both remind me of Berkshire circa 1980: They are little less unwieldy, a little easier to analyze, and a lot less dependent upon a famous CEO.

What's more, both have easily outpaced Berkshire in the all-important measure of price appreciation over the past decade.