"PS if someone could put a link to the original log-log plot post on Shrewdm a few months ago, I have spent hours looking for it and haven't found it"

Thanks for sharing, Ralph. Here's the graph you asked for, updated to Dec 2025.

Google: Price per share vs BV per share '64-'25

May I make just a couple of points about your graph?

1. The GenRe purchase in mid 1998 wasn't the only event that could have caused the change in the rate of growth. The S&P 500 also changed its rate of growth, although beginning in early 2000. Here's some data:

period, Growth of Berk's BV, Growth of S&P 500 index

Oct '64 - Dec '80, 18.6%/yr, 1.4%/yr
Dec '80 - Dec '99, 28.8%, 10.2%
Dec '99 - Dec '20, 10.2%, 5.3%

My guess is that the cause of the S&P's change in growth rate was that the S&P's P/E had reached a high point in early 2000. Similarly, Berkshire's P/B reached a high point in mid 1998. Just another possible cause for the change in growth rate.

2. Just a nit-picky point. If the exponent in your exponential fit is 0.0931, then the growth rate is e^0.0931, or 9.76%/yr, not 9.31/yr%. I's sure that you know that.

I would be interested in other conclusions that you can make from comparing historical Price and Book Value. The most general conclusion that I would draw is that price tracks BV, and that that relationship is causal, not coincidence. As to the future, my guess is that the S&P can return to long term, nominal 10% growth (in the last five years it's been over 14%/yr), but that Berkshire cannot return to 19% growth. Berkshire is just too big to do so, especially if it restricts itself sure things, like BNSF, and eschews unsure things, like NVDIA.