Three Worthwhile and Difficult Conglomerates
Board: Berkshire Hathaway
I agree wholeheartedly with Jim and Euro's points; getting to understand Leucadia is a worthwhile enterprise, but difficult. After Berkshire and Fairfax, Leucadia is my 3rd largest holding, so obviously I have a weak spot for worthwhile, difficult conglomerates, even if these 3 are very different from each other.
Berkshire boring and reliable, it's unlikely to grow very fast, but it's very safe, very conservative. Buffett never sells its subsidiaries and rarely sells its major public equity stakes. I am less and less enamored of the company, but the price just keeps getting better and better, so what's a guy to do? Berkshire has a fantastic track record, particularly in the 70s and 80s, when prices were so low, although not so great lately, and likely to get even less great when the torch gets passed. With Buffett at 80, that is likely to happen fairly soon.
Valuation: You can get a very rough idea of valuation by looking at earnings, with market cap currently at about 17 times earnings. However those earnings are highly dependent on the timing of the sale of investments, and on insurance mega-cats, so looking at book value and analysis like the 2-column approach are also helpful. On the basis of BV or its cousin, BV + float, or BV + some % of float, Berkshire is unusually cheap right now. Looking at investments and operating companies separately (the two-column approach), the company is even cheaper. About half the company's value is held as insurance investments, the biggest ones being Coca-Cola, Wells Fargo, American Express, Procter and Gamble, Munich Re, Kraft, Johnson and Johnson, and US Bancorp. You can back out these insurance investments (careful, you can't back out ALL investments, since most of the non-insurance investments are counterbalanced by liabilities), by assigning them market value and backing out their dividend income, to avoid double counting. What's left is an operating stub that is only about 5 times earnings, despite the fact that current earnings are probably at a below-normal level because of the housing slowdown.
So bottom line, the company is dirt cheap, but it's the operating half of the company that is attractive; the other half being public equities, means Berkshire is heavily exposed to the equity markets, not to mention the derivative position, and I think the company is not likely to grow much in the future.
Fairfax is like Berkshire in some ways, primarily in that it is an insurance-based conglomerate, allowing some safe leveraged equity bets with the float. Watsa is much less conservative than Buffett, and doesn't mind going short the market, most famously with his CDS bets against housing in the years prior to the 2007-2008 crash. If you are worried about a serious economic slowdown and even deflation, then Fairfax is a better bet than Berkshire, since Watsa has basically no market exposure now, much to the detriment of recent performance. Watsa is 60 years old, so succession is less of a concern.
Valuation is more difficult. Earnings are actually negative in the last 12 months, largely because the equity markets are way up and the hedges that Watsa put in place, too early as it turns out, more than wiped out all the equity gains last year - in other words, the market is up more than the $4-5 bn worth of stodgy equity holdings (primarily Wells Fargo, Johnson and Johnson, US Bancorp and Kraft; obvious similarities with Berkshire!). Book value is not so helpful either; Fairfax has a market cap of $15 bn, and equity of only $8.1 bn, a much higher ratio than Berkshire's (189/160=1.18). So how to give a value to Fairfax? That is a subject for another post, since this one is getting long already, and we haven't even got to Leucadia, the ostensible subject.
Different from the other two, this conglomerate is not based on insurance, so the only leverge they use is a moderate amount of debt, $1.5 bn compared to $6.9 bn equity. They do not hold big public equity stakes in things like Wells Fargo, JNJ and Kraft, they prefer to make private deals for small companies, often troubled assets or risky financing of a far-fetched enterprise like setting up a new iron-ore mine and building the railroad to the sea, and the harbor, and going up against the big 3 iron ore producers to do this (ie Fortescue, FMG), or a California winery, or a Barbados utility, or a pre-paid telephone card company (STI), or a blood substitute company that no one else believes in but the company's CEO and his aging mother. Some of these deals work out fabulously (FMG), and some of them don't (STI). The principals, Cummings and Steinberg, are 70 and 67 years old, so they could be doing this for a while if they want to.
Valuation is a challenge, but it's not as bad as Fairfax. As tzuri suggests, you can start with the 3 big equity stakes, Jefferies, Inmet and Fortescue. LUK's market cap is $8.4 bn, so when you subtract those 3 positions (only JEF pays a meaningful dividend, $16 million last year), you get 1.05 + 0.78 + 1.60 = 3.43 billion. That leaves $5 bn to be accounted for.
As tzuri says, the FMG note is in the next tier, and that's a marathon question in itself. Suffice it to say that the it is a $100 bond (or note), with a funny kind of interest: 4% of the after-royalty revenues coming from 2 mines, Christmas Creek and Cloud Break if memory serves. These payments go on until 2017, payable twice a year; since the project has gone well, there's lots of iron ore in there, and prices are good, the payments are very high; just for the last 6 months of 2010, the note generated 'interest' of $72.9 million. Unfortunately, this may be too good to be true, since FMG is now claiming they can dilute this interest by issuing more such notes, diluting the same 4% revenue stream, although this is of course hotly contested by Leucadia. So there is a question mark over these payments which might otherwise be worth a billion or so. I suspect Leucadia will prevail here, but lets say $500 million to be on the safe side. That's still a lot better than the $160 million the company has this note on their books for.
Still another $4.5 billion to go, but I have to take a long break and do some real work, so I'll leave off for now, comments welcome on any of this of course.