I expect Berkshire to underperform at some point, but not today
At Berkshire's annual meeting this weekend, Warren Buffett will be inviting a prominent short seller to discuss his bearish thesis. In fact, it seems that Doug Kass is actually short the stock. In contrast, Berkshire is currently my largest stock position.
In March, 2008, Kass wrote a bear thesis on Berkshire. Some of his more prominent and currently relevant reasons include that the field of good investors is getting more crowded, that the law of large numbers is working against Berkshire at an increasing rate, that overall stock market returns themselves are expected to be lower than they once were, that Berkshire had then made a lot of reinsurance contracts/derivative investments that were poorly disclosed and that could blow up. All but the last are still true. The last reason didn't turn out, but certainly reinsurance can be pretty volatile.
Stepping back a bit, it's important to realize that Berkshire is not a firm that generates the majority of its earnings from its investments. Morningstary estimates that a bit over a quarter of Berkshire's 2013 earnings will come from investments (see the table called look-through earnings). Most of Berkshire is an operating company. Many of Kass' criticisms aimed at the firm's potential to earn money through its investments are correct, imo, but they don't address the majority of what the firm actually does right now.
Berkshire has definitely been a conglomerate for some time, but it's recently added on a number of capital-intensive operations like a railroad and a utility, plus it's been buying small manufacturing firms. When Morningstar assigns economic moats (i.e. when Morningstar evaluates whether a company has durable competitive advantages), railroads, utilities and manufacturing companies overwhelmingly tend to get narrow moats, not wide ones. Exelon is a wide moat utility, and GE is a wide moat manufacturer, but those are exceptions that prove the rule. In fact, a lot of manufacturers do not warrant a moat at all. In contrast, M* thinks that Berkshire has a wide moat.
That should be the core of a bearish thesis on Berkshire, imo. The company is increasingly going into sectors where it will not have a durable competitive advantage over competitors. Now, against that, the insurance business throws off a lot of float, since it runs at an underwriting profit (if an insurer of any type runs at an underwriting loss, it can make up the difference from investments, but inherently that means it's in a vulnerable position). Berkshire can use a lot of the float to reinvest in the capital-intensive businesses. In that sense, Buffett was a master strategist as well, building a company which could take advantage of synergies in that way. But such advantages can't last forever. Buffett is sui generis - or maybe he just got lucky, as Bill Gross observed, and came of age with a boatload of money in an era when it was easy to make big gains. Buffett can't be replaced, and the firm will have to rely on the leadership and operational skills of lesser people. Its advantages over similar property/casualty, utility, railroad and manufacturing firms will decline.
That said, that doesn't suffice for me to be actually short the stock. It does suffice for me to say that Berkshire is in a very gradual decline, and that I'll gradually be reducing my position size, probably just by buying other companies and putting money into a Vanguard Target Date fund. I am not sure if Kass is actually short the stock now - he may have made a bunch of money shorting Berkshire in 2008 as the financial crisis unfolded, but I would not short the stock now. Even if Berkshire's advantages dwindle as I described, it would take a lot more for me to start shorting the stock. It would have to be something like persistent poor management, especially at its core insurance operations, or a massive disruption in the property/casualty industry that left Geico with large operating losses. After all, shorting stocks is not costless. You pay interest. You may need to meet a margin call if the stock rises too much. On the other hand, buying a stock and holding on to it for a very long time, or buying a cheap target date fund and holding it for a very long time, is very, very cheap.