No. of Recommendations: 26
Price to Book has been a quick, simple method for valuing Berkshire for a long time. Jim Mungofitch suggested using Peak Book in the denominator to improve the estimate.
I looked at quarterly, inflation-adjusted Book Value of Berkshire since 2000 and plotted it here, along with the least-squares trend line -->
https://ibb.co/kMydfJtA couple of things popped out at me. First, Book Value has gotten noisier than in the distant past. There were two dips below the trend line that happened in 2009 and 2012. These returned to the trend line within a few years. (This might not have happened if we ended up with the Great Depression 2.0, but fortunately that didn't happen.) But those short-lived drops below the trend line are what made peak PBV fit the data better by staying closer to the trend line.
But the last few years have seen book value become much noisier and moved away from the trend line in both directions. This was probably caused both by the easy money that's been available, as well as changes in how book value is calculated. But if the reported book value is indeed becoming noisier in BOTH directions, peak price-to-book may not produce a more reliable indicator.
In fact, I submit that the long-term trend line of book value would be a better value to use than the reported book value. Book is likely to grow consistently and reliably as long as there is no structural change to Berkshire. It certainly seems to have mean-reverted to the trend in the past and presumably will continue to do so going forward. I plan to use trend book, rather than the reported value. I'll watch for signs that book growth is slowing, but even if that happens, it will likely change very slowly.
According to my calculations, nominal Book Value has been growing over the past 20 years at 10.2% per year (2.46% per quarter). Here are the nominal trend line quarterly values for the next few years.
2022 EOY 311176
2023 318830 326672 334707 342940
2024 351375 360018 368873 377946
2025 387242 396767 406526 416525
2026 426770 437268 448023 459043
2027 470334 481902 493755 505900
2028 518343 531093 544156 557540
2029 571254 585305 599701 614452
2030 629565 645051 660917 677173
2031 693829 710895 728381 746296
2032 764653 783461 802731 822475
2033 842706 863433 884671 906431
2034 928726
The nominal values are easy to use because you can calculate them in advance as far into the future as you want.
But it's probably better to use real (inflation-adjusted) values. I show the inflation-adjusted trend line growing at inflation plus 7.95% per year (1.93% per quarter). Once inflation and BRK's book values for the end of 2022 are reported, I plan to update my spreadsheet and then calculate the values every month starting from the trend book value on 31 Dec 2022. I'll report those starting values when I get them if anyone is interested.
BTW, we're used to updating BRK's book value quarterly with a lag when it's reported. A side benefit of using trend book is you can calculate the value monthly or at any other period you want and it has no lag.
Thoughts?
