I believe it was as much an empirical / practical point as anything else.

With BRK, if you are using P/B as a measure of overvaluation or undervaluation, peak book has served as a better indicator of when BRK has been undervalued. The reason is because because the fluctuations in book value down have typically been transient. Given the way BRK made their capital allocation decisions, which tends to be conservative and protective of downside, the usual history is that BRK returns to peak book and surpasses it within a reasonably short period of time. So P/Peak book serves as a better indicator of moments of undervaluation that may make good opportunities for investment.

Some of the people who have questioned this typically tend to take a purist view that P/B in real time is the true measure of value and so why wouldn't you use that? I don't think Jim has ever tried to argue that this purist view might be right in theory, but for BRK in particular, with its large diversified based and conservative capital investment approach, its not a very practical signal of undervaluation. Rather, you would probably be better off using P/Peak Book unless you think something has fundamentally and dramatically changed in the prospects of the major components that make up book value.

Hope that helps and hope that is accurate, Jim.