There is a conspicuous gap between the last Buffett Partnership letter, written in 1969, and the first Berkshire Hathaway letter posted on the company’s website, written in 1978. Recently I discovered several of the “missing” documents.

A consistent theme in Buffett’s early management of Berkshire is that capital from the textile operation was best redeployed in either marketable securities or business acquisitions. From 1969 to 1977 the textile operation averaged a return on capital of less than 3%, while the insurance and banking subsidiaries averaged well above 10%. Buffett’s refusal to shut down the Berkshire mills resulted in an immense opportunity cost compounded over nearly 20 years.

The moral seems to be that basing investment solely upon asset value (quite significant in Berkshire’s case) is not intelligent. This was a rewarding activity when security analysis was in its infancy, but a great deal has changed since then. One of the best criticisms of the “Graham approach”—which involves net working capital bargains, classic arbitrage, etc.—can be found in Victor Niederhoffer’s The Education of a Speculator:

“On the rare occasion when a true guru shares secrets of a recurring, well-defined systematic nature, the cycles are about to change. Better to go against. What looks good today is encapsulated in the market tomorrow and will change the expected profits, the probabilities, and the paths of least resistance in subsequent periods. A good bet is that all systems will stop working when you use them.”

This criticism extends to merger arbitrage, convertible arbitrage and liquidations, as well as to other approaches that are less systematic. Recent experience suggests that even value-oriented investors are unsafe.