Post New

Has anyone here estimated Berkshire's Book Value Per Share (BVPS) as of Dec 31? The Sept 30th BVPS was up 20% over Sept 30, 2023. The June 30th BVPS was up 12% over June 30, 2023. The March 31st BVPS was up 14% over March 31st, 2023. BRK-A stock and the S&P 500 were up 25% in 2024. Without trying to estimate the growth of the business units and the return of the equity portfolio, I'm tempted to estimate that Berkshire's BVPS rose about 4% in Q4 and 16% in 2024 to $452K/A-share, but some of you may have better estimates. Thank you for your input.

I wanted to know what P/B was as of today. It looks from from Warren's history of repurchases that his estimate of IV is about 1.56x BV. Berkshire repurchased shares in March at 1.56x BV and issued shares in Sept at 1.56x BV,

If BVPS was up 16% in 2024 to $452K/A-share, then the current P/B as of 1/14/25 is 1.45. If BVPS was up 20% in 2024 to $467K/A-share, then the current P/B as of 1/14/25 is 1.50.

Has anyone here estimated Berkshire's Book Value Per Share (BVPS) as of Dec 31?

My estimate for end of year BV is $450,315 or $300.21 per B share. My last few quarterly estimates have been lower than actuals so I would not be surprised if it comes in higher and closer to your estimate.

Jeff

I don't make my own estimates, just look at what a few financial sites say about book value and P/B.

Here's the latest:
Wed Jan 15 08:17:23 AM CST 2025
From ycharts: P/B 1.544 for Jan. 14, 2025
From alphaquery: P/B 1.51
From macrotrends: P/B 1.51 January 14, 2025

For example alphaquery says:
Price to Book Ratio (Most Recent Quarterly Book Value per Share) 1.51
Book Value per Share (Most Recent Fiscal Quarter) $262.01

Here are the end-of-year from macrotrends for the last few years:
2019-12-30, 1.38
2020-12-30, 1.29
2021-12-30, 1.40
2021-12-30, 1.40
2022-12-30, 1.48
2024-12-31, 1.54

For example alphaquery says:
Price to Book Ratio (Most Recent Quarterly Book Value per Share) 1.51
Book Value per Share (Most Recent Fiscal Quarter) $262.01

Book Value per Share (Most Recent Fiscal Quarter) is $291.72 by my calculation.

Current P/B 1.56

End of year 2019-2024:
1.30
1.21
1.31
1.43
1.37
1.51? Guess

Thank you, Bluehorseshoe, rayvt and AdrianC. The average of the estimates you quote is a P/B as of Dec 31 of 1.51, or a BVPS as of Dec 31 of $451,000. That would put P/B as of Jan 14 at 1.50. I think these estimates are better than mine. I appreciate your help.

rrr12345

I look at the price-to-book value for a strategy that Jim (mungofitch) has mentioned a few time.

Here is what he posted on the old TMF board (R.I.P.):

When the P/B ratio drops below 1.35 sell some BRK stock (maybe half your position?) and use the proceeds to buy long dated deep in the money BRK call options.
The low strikes generally offer roughly 2:1 leverage
...
Then, next time the stock trades above 1.55 times known book per share, take the profits on the calls and use all the proceeds to buy plain stock again, ditching the leverage.
So, you're invested all the time, but pile on a little leverage when the outlook is particularly good.

I sold sold my last DITM call on 1/29/2024. The actual leverage on that trade was 2.27X.

"I think these estimates are better than mine."

Actually they are pretty close. The average of the estimates cited in the thread is a P/B as of Dec 31 of 1.51, or a BVPS of $451K/A-share, versus my guesstimate in the opening post of the thread of BVPS as of Dec 31 of $452K/A-share. For those of you who made estimates yourselves, would you mind going through the calculations? Many thanks.

Then, next time the stock trades above 1.55 times known book per share, take the profits on the calls and use all the proceeds to buy plain stock again, ditching the leverage.
So, you're invested all the time, but pile on a little leverage when the outlook is particularly good.

The only problem is if the stock remains low for too long and the option gets close to expiration. Then the leverage doesn't pay off nearly as much, if at all. The advent os LEAPs has helped in this regard, but they are also more expensive because of all that extra time premium.

"It looks from Warren's history of repurchases that his estimate of IV is about 1.56x BV. Berkshire repurchased shares in March at 1.56x BV and issued shares in Sept at 1.56x BV."

Another estimate of IV comes from the weighing machine method, where the trendline of a long term graph of price per share versus BV per share is used as the IV estimate. That IV estimate is now 1.53x BV. Today's stock price (Jan 15th) is 1.52x BV.

"Today's stock price (Jan 15th) is 1.52x BV."

Assuming BV is $451K/A-share.

For those of you who made estimates yourselves, would you mind going through the calculations?

Start with the end of Q3 BV of $291.72/Bshare

From there estimate the after tax ops earnings for the Q, after tax dividends from equities and return on the cash reserves, plus the after tax increase in the equity portfolio. I've been ignoring the quarterly insurance underwriting profits because they have historically not been that meaningful but you could put something in for those as well.

My estimates:

Shares Outstanding: 2.156B Bshare equivalents

Ops: $5.8B : $2.59/Bshare

Dividends: $0.786B : $0.36/Bshare

Cash Return: $2.75B : $1.27/Bshare

Equities: $9.2B : $4.27/Bshare

Total: $18.3B : $8.49/Bshare

Q4 BV Estimate: $300.21

The underwriting profit in Q4 2023 was $900M but the full year average from 2010 to 2022 was only $700M. Maybe there's another $0.40 to $0.50 that should be added for insurance but i happen to think my Q4 estimate for ops earnings is high so I'm not adding anything. The wonderful thing about forecasting is you are guaranteed to be wrong so just embrace it.

Jeff

If BVPS was up 16% in 2024 to $452K/A-share, then the current P/B as of 1/14/25 is 1.45. If BVPS was up 20% in 2024 to $467K/A-share, then the current P/B as of 1/14/25 is 1.50.

You have these switched. The *higher* BVPS should result in the *lower* P/B.

"My estimates:..."

Excellent. Thanks.

"You have these switched. The *higher* BVPS should result in the *lower* P/B."

Yes. Thanks.

I updated my chart „BRK-Price to WMAof BV 1,7" which shows how expensive BRK-A's is in relation to it's Peak(!) Book Value.

(Based on Jim's method he once explained in detail; only difference: I use "WMA16*1.7" as I think it's a better curve fit than his "WMA16*1.6")

https://drive.google.com/file/d/1Tid3g1RUZpC-qZe_z...

My updated (see my previous post) chart „BRK-Price to WMAof BV 1,7" contained a wrong data point, a more than $100k too high Book Value for end of 2022. As the chart uses "Peak BV" it carried that too high Peak BV through all of the following quarters. That resulted in the yellow trendline being too high for the last two years, so Berkshire in fact is more "overvalued" since two years than the previous chart suggested. Sorry.

To make up for that mistake I additionally added in the corrected chart the so often spoken about "channel" in which Berkshire since at least year 2000 quite reliably trades. The result:

https://drive.google.com/file/d/1TcWuqXsDUJTMqYI_T...

The two black lines mark the lower and upper borders of that channel. The area between covers practically all BRK prices, expressed as a multiple of (smoothed) Peak BV (Book Value).

The yellow line as in the previous chart is 1.7 x WMA16 (weighted moving average) of the higher value of "Peak BV" and "90% of smoothed Peak BV"

My Interpretation: BRK's price since 2000 practically always is in a channel, ranging from 1.2 x Peak BV to 1.7 x Peak BV. If it reaches the upper border it comes down again, if it reaches the lower border it comes up again. Both extremes do not persist.

For a full year now the share price is over the upper end of that range, an unusually long time, only comparable to around 2007/8 (so that I personally expect it to come down again --- which is why I still have some puts :)


Methodology (based on Jim):
- The basis for the whole chart: Quarterly share prices and quarterly book values, so for each there are only 4 data points per year
- For each quarter "Peak BV" (Peak Book Value) up to that date is used*
- This "Peak BV" now is interpolated/smoothed**
- Then the higher value of "Peak BV" and "90% of smoothed Peak BV" is taken, let's call it "VALUE"
- 1.7 x WMA16 (weighted moving average) of that "VALUE" results in the yellow line
- 1.2 x (smoothed) Peak BV results in the lower black line
- 1.7 x (smoothed) Peak BV results in the upper black line


*Inflation adjusted, so actually resulting in what Jim calls "Real Peak BV"; not important for the interpretation of the chart.

**This is a deviation from Jim, as he does not do that step; actually it changes nothing as you can see if you compare the newer chart with the one in my previous post, as I added that smoothing only to this newest chart; I can't see any difference in the yellow lines.

My comment

**This is a deviation from Jim, as he does not do that step; actually it changes nothing as you can see if you compare the newer chart with the one in my previous post, as I added that smoothing only to this newest chart; I can't see any difference in the yellow lines.

is confusing, as of course there is a difference in the yellow lines of the old and the corrected chart! But that concerns only the difference that results from correcting the as explained wrong peak book value data point. The smoothing itself does not result in a visible difference.

Nice work!

My Interpretation: BRK's price since 2000 practically always is in a channel, ranging from 1.2 x Peak BV to 1.7 x Peak BV. If it reaches the upper border it comes down again, if it reaches the lower border it comes up again. Both extremes do not persist.

For a full year now the share price is over the upper end of that range, an unusually long time, only comparable to around 2007/8 (so that I personally expect it to come down again --- which is why I still have some puts :)

Other valuation estimates point to the same thing. Berkshire is at the extreme upper end of its valuation range.

Why? Don't know, not seeing a value-boosting catalyst any time soon, quite the opposite. Investing all that cash could take many years.

I am very sorry, but I again want to correct something. As longer I looked at the chart as more I wondered why it claims that last year Berkshire's share price was practically always 1.7 x Peak BV or more --- as we all know the multiple was not that high (apart from maybe very short stretches). I just found that the reason is the "Inflation adjustment of BV" I said does not matter for our purposes. Wrong! It does matter, as it has a distorting effect. Without that adjustment and the resulting distortion (and the now better fitting WMA16*1.6 instead of 1.7):

https://drive.google.com/file/d/1CQrM07aVXdn7hXAa7...

So the channel in which Berkshire's stock price moves is the range 1.2-1.55 x Peak BV --- as often said here. And last year it was not mostly at or even over 1.7 Peak BV, but at the upper limit of that range. We know that it was partly over 1.55 (1.65 or so?), but that is not visible because of the low resolution of the chart, with only one data point for each quarter.

Sorry for all the confusion. That happens if one is too quick with clicking on "Send".

„BRK-Price to WMAof BV 1,7"

The main conclusion is that price tracks book value, which is reasonable, and which is proven by the data. The second conclusion is that P/B is sometimes higher than at other times, naturally.

There are numerous ways to analyze price versus BV. Plotting price and WMA of BV together against time is as good a way as any. Personally, after plotting price and BV together against tome, I like to then plot price versus BV (Mathematically, a log-log plot with a power law fit is appropriate when two variables move similarly against time.) That shows how well price follows BV. I can then plot price versus the trendline of the P vs BV plot to focus on the high and low excursions. Doing so makes it easier to see the magnitude of the excursions.

Another thing that I like to do is to use both the high and low stock price during the quarter, in order deliberately to add more scatter to the plots (For example, in Q3 the high price was 19% higher than the low price), and I like to use both the starting and ending book value for the quarter. None of these things are necessary for the analysis, but they do make it easier to see quantitatively how well price tracks BV, and and the magnitude of the excursions.

Anyway, as I say, plotting price and (WMA of BV)*1.7 together against time is perfectly fine, but plotting one against the other as well, and also plotting the ratio against time, makes it easier to see what's happening. Just FWIW.

Post New

Berkshire Hathaway FAQ

---

Contact Shrewd'm
Contact the developer of these message boards.