An implicit assumption of the efficient market hypothesis is that identical assets should sell for identical prices. Every now and then, however, this rule is clearly violated (Palm and 3Com being one relatively well-known example), illustrating the market's puzzling and somewhat exasperating inability to perform simple arithmetic.
NL Industries (NL) is a holding company that owns 30% of Kronos (KRO), a titanium dioxide producer with a market cap of 1.85B, and 87% of CompX International (CIX), a lock-making company with a market cap of 214.5M (these figures are at the time of writing). The market cap of NL Industries should therefore be at least 0.30(1850M)+0.87(214.5M)=741M. NL's actual market cap is 536M, which would translate to a 38% gain if NL's price were to cover the difference.
Free money certainly sounds promising, but it's not necessarily a free lunch. There are two reasons why going long on NL might not work.
First, it is possible that it is Kronos and CompX which are overpriced, not NL that is underpriced. I find this possibility unlikely, due to the fact that the market is valuing Kronos and CompX directly instead of calculating a sum of products as it is (or perhaps I should say isn't) doing for NL. Nevertheless, it is a possibility that has to be acknowledged.
Second, there is no guarantee that the market will ever actually correct the pricing difference. In a rational market such pricing differences would not occur, but the market is clearly not rational, and can remain irrational for much longer than we would prefer.
To investigate the second possibility, I've compiled a table that shows the various mispricings of NL Industries over the past 5 years. (Note: I accounted for the fact that NL owned 36% of Kronos in 2008.) Market caps are in millions.
|Year||NL Mkt Cp||CIX Mkt Cp||KRO Mkt Cp||NL Actual Worth||Diff (%)||NL Price ($)|
As you can see, NL was actually over-priced at the end of 2008, and violently corrected in 2009, losing 49% of its value. At the beginning of 2010, NL was still slightly over-priced, but the market cap of Kronos grew 157% that year while CompX grew 60%, so NL's price grew 61% anyway, and became significantly under-priced.
It's important to note that even though NL only holds 30% of Kronos, NL's composition is actually 75% Kronos, 25% CompX, due to the fact that Kronos is a larger company. This means that NL's price is affected by the outlook of Kronos much more than by the outlook of CompX.
As you can see, at the beginning of this year NL was under-priced by 53%, but NL is currently under-priced by only 38% while NL's price is actually down 9.2% this year. The reason for this is that shares of Kronos have fallen nearly 20% so far this year on two negative quarterly earnings in a row. The question of whether there will be anything left to gain from the mispricing of NL this year thus hangs on the performance of Kronos for the remainder of the year.
I think that there is good reason to be more optimistic about the second half of the year for Kronos, and will discuss the situation and outlook for Kronos more thoroughly in a separate article.
However, even if we assume that Kronos falls another 20% for the remainder of this year, NL would still be under-priced by 18%. On the other hand, if Kronos rises 20%, then NL would be under-priced by 60%. Given these numbers, it seems that there is a very nice upside and relatively limited downside to going long on NL.