Post New

I like playing with numbers. We all have vices.

Apple's earnings per share and valuation multiples seem to have a gentle cycle. The earnings peak, then the P/E troughs, then the earnings pick up, then the multiples peak, then the price peaks, then it all starts over again.
Eyeballing the pattern from the last three cycles, a wild speculation: peak P/E based on trailing earnings might peak around May and, with an upswing in EPS underway, absolute price might peak around year end.
Place no weight on this forecast, it's just for fun.

One year price forecasts are a silly game, but here goes: Berkshire B shares might end the year in the vicinity of $390.
I spent a few minutes estimating that just to get a handle of which covered calls might be a particularly dumb idea.
I hope to be pleasantly surprised, but that's my number of the day.

Jim

Here's another magic 8-ball number--

A wild guess as to how many buybacks there will have been in Q4.
October: 200 A-share equivalents
November: 115
December: 100

This is based on a rather dodgy formula trying to correlate how many shares are purchased each month based on the average P/B during that quarter.

As an aside, I note that Mr Buffett has shown a strong preference for A shares lately relative to B shares. Since the buybacks seem sometimes to be limited by liquidity, this is a bit surprising.
In the last 18 months for which we have the data, the ratio of average price paid during months that both A and B shares were purchased was 1530:1.

Jim

"One year price forecasts are a silly game, but here goes: Berkshire B shares might end the year in the vicinity of $390."

would that be approximately growth in book with a steady multiple?

FWIW, I sold the 430 covered calls January 2025 at 7.77 on 1/3 of my non IRA holdings for “dividend” money and I sold the 390 covered calls January 2025 on my entire IRA position for 13.99. Proceeds for both sales are currently in cash earning 5% per year.
I don’t plan on selling any of my BRK for at least 5 years unless valuations become extreme (like P/B over 1.6 )
Cheers,

Brian

OOPS.

I meant the 410 calls at 7.77 on my IRA.

I do have some covered calls at 390 expiring in June 2024.

would that be approximately growth in book with a steady multiple?

No, it would be implicitly assuming an "average" valuation multiple at the end date.

I took a wild guess what both book and my own IV estimates will turn out to have been at end 2023 when the annual report comes out, and plugged those two into my sundry models of one year forward average real returns.
Most of the models use data only since 2008--the "modern era" of valuation multiples.
I then took the average of the real return forecasts, and added a guess of one year of inflation. I plucked 3.25% from the air.

Since these models are all of the general form "starting valuation level X gave average forward return Y in the past", they all implicitly assume that the trend of value growth will repeat.

Jim

Most of the models use data only since 2008--the "modern era" of valuation multiples.

I would argue that this valuation model may be flawed.
The low interest rate environment since the great financial crisis lowered the valuation of the insurance float. If interest rates remain more normalized (think like 1995 to 2005), Float is way more valuable and thus BRK should trade at higher P/B ratios. 5% on the float sure is nice compared to near zero since 2009.

The low interest rate environment since the great financial crisis lowered the valuation of the insurance float. If interest rates remain more normalized (think like 1995 to 2005), Float is way more valuable and thus BRK should trade at higher P/B ratios. 5% on the float sure is nice compared to near zero since 2009.

I'm not a big believer in this view.

It's the real interest rate that counts. Interest rates are higher, but inflation is higher too.

Plus, there is tax to pay on the NOMINAL interest earned, so there is actually a loss if both inflation and interest rates rise the same amount.

Year-on-year inflation rates don't tell the whole story easily--at the moment 12-month headline inflation is (unusually) lower than the 3- month T-bill rate--but overall I see no evidence that the interest rates being earned are sufficiently better in real terms than they were a couple of years ago to offer any improvement in the financial situation at Berkshire.
e.g., A portfolio of 3-month T-bills earned precisely nothing in real terms after 21% corporate tax from July 2022 to September 2023 while the T-bill rate averaged over 4.4%.
About the best you can say is that it hasn't been slowly losing real value they way it was for a couple of years before that.

The cash pile is worth a lot for sure, but nothing for its current real earnings. Only its purchasing power.

Jim

Buffet's preference for the A shares seems to coincide with his original guidance to shareholders that the A shares are the better deal as long as the premium paid is not more than 2% over the B shares.

I sold a January '25 covered call for $380 $24.80. It's trading a bit higher now. My call is covered by some shares I might have wanted to sell this year anyway.

I note that Mr Buffett has shown a strong preference for A shares lately relative to B shares. Since the buybacks seem sometimes to be limited by liquidity, this is a bit surprising.

Would he perhaps be looking ahead with an eye towards the cumulative benefit of retiring a small percentage of the pool of A shares (with their outsize voting rights) each year?

The math wrt voting control will change once all of his shares have ultimately been liquidated by the Gates Foundation. Institutional buyers will presumably be content spending 98c of their capital on B shares when A shares would cost $1.00, but it would still be nice to have as many A shares as possible in the hands of old stubborn individual shareholders

--sutton

“” The cash pile is worth a lot for sure, but nothing for its current real earnings. Only its purchasing power.“” However, the end of super cheap easy money makes our cash more valuable if a super target becomes available.

It's the real interest rate that counts. Interest rates are higher, but inflation is higher too.

Plus, there is tax to pay on the NOMINAL interest earned, so there is actually a loss if both inflation and interest rates rise the same amount.


This is a very insightful and convincing argument. However: there’s a lot of percentage gain, when you’re starting at less than zero on $160b. From about 2010 to 2015, 3 month Treasuries were earning about 0%, with inflation at about 2%. Then from 2015 to 2-2.5%, and inflation was

sorry, please ignore this premature postulation - full post coming soon…

It's the real interest rate that counts. Interest rates are higher, but inflation is higher too.

Plus, there is tax to pay on the NOMINAL interest earned, so there is actually a loss if both inflation and interest rates rise the same amount.

Sorry for that false start.

Jim makes a very insightful and convincing argument. However: there’s a lot of percentage gain, when you’re starting at less than zero return, and it makes a difference, when you’re working with $160b.

So from about 2010 to 2015, 3 month Treasuries were earning about 0%, with inflation at about 2%. Then from 2016 to 2019, there was a brief spike in these treasuries to almost 2.5%, with inflation still at about 2%. Then the COVID madness, with treasuries back to zero, until about the beginning of 2022, and inflation around 2% (2020), 5% (2022) and 8% (beginning of 2023). And in 2023, inflation has dropped back to about 2-3%, with treasuries at just over 5% now.*

So we have had 4 different periods, with the difference between treasuries and inflation about -2% from 2010 to 2015, +0.5% (2016-2019), -2%, -5% and -8% (2021, 2022 and early 2023), and now about 2.7% (late 2023).

It is quite true that we pay taxes on the 5.2% current rate, not the 2.2% spread, so maybe we only keep half of that spread (5.2% less 1.3% tax, so 3.9% post tax, less 2.5% inflation, so we’re at plus 1.4%.

But still, +1.4% is so much better than the negative spread we’ve had for so long, now, almost 14 years, averaging about -2%. Although float has kept slowly increasing during that time, for comparison’s sake, -2% on $160b would be -$3.2b, and +1.4% would be +$2.2b. So the same float generates $5.4b more in earnings. Give that a 15x multiple, wouldn’t it be fair to say that the $160b in float is worth $80b more, at current treasury and inflation rates, compared to the past 14 years?

Regards, DTB

*P.S. I have been unable to find a graph showing the exact spread between 3-month treasuries and inflation over time, say in the last 20 years, or the average rates from year to year, so I have just eyeballed the data. Perhaps someone with access to this data (a Bloomberg terminal, for instance?) could sharpen this analysis. TIA

I have been unable to find a graph showing the exact spread between 3-month treasuries and inflation over time, say in the last 20 years...

I have a data set for 3-month T bill rate and CPI. Headline CPI isn't perfect, but it will do to get the broad strokes. (bit fan of pinnacle.com)
If you simplify and assume that you reinvested each calendar quarter for three months, the real value of your portfolio over time in the last few years would look like this (after 21% corporate tax, the current rate for US corporations)
e.g., for the last decade, real portfolio value at each quarter end

2013-12-31 1000.00
2014-03-31 992.90
2014-06-30 980.06
2014-09-30 980.11
2014-12-31 987.19
2015-03-31 993.61
2015-06-30 980.70
2015-09-30 980.31
2015-12-31 982.81
2016-03-31 984.14
2016-06-30 972.01
2016-09-30 969.71
2016-12-31 968.18
2017-03-31 960.49
2017-06-30 957.65
2017-09-30 956.45
2017-12-31 953.77
2018-03-31 947.67
2018-06-30 941.14
2018-09-30 942.61
2018-12-31 947.10
2019-03-31 948.80
2019-06-30 940.98
2019-09-30 943.04
2019-12-31 944.15
2020-03-31 941.69
2020-06-30 950.16
2020-09-30 937.59
2020-12-31 936.58
2021-03-31 926.88
2021-06-30 905.57
2021-09-30 891.11
2021-12-31 877.22
2022-03-31 859.48
2022-06-30 835.16
2022-09-30 827.06
2022-12-31 822.97
2023-03-31 826.73
2023-06-30 827.72
2023-09-30 826.45
2023-12-31 835.45


To summarize: Slow loss of real value from the start of this set at end 2013 to end 2020, at a rate (inflation and tax adjusted) of -0.93%/year.
A big "sudden" loss of -10.8% in the next six months to mid 2022 (not annualized rate, just a one-time haircut)
Perfectly flat (net) since then. Down a pinch then up a pinch.

So, the celebration should be that the big sudden loss in real purchasing power in the first half of last year has ended, and the recent flat spot is at least better than when we were losing real value gradually for the ~7 years up to the pandemic. But no positive value creation.

I agree with other comments that the real value is the optionality. From time to time there is an opportunity that only someone with a big cash pile can take advantage of. Cash is the flip side of a bear market, and shares the same forgotten property: that's actually how you make most of your money, it's just that most of the time you forget that.

Jim

One year price forecasts are a silly game, but here goes: Berkshire B shares might end the year in the vicinity of $390.


With 2023 likely to come in as an uncharacteristically poor year for after tax Ops earnings (energy+rail+MSR), what are your thoughts for 2024? Trend would seem to indicate $27B but I had been penciling in getting about half way there to maybe $24B next year. I personally think rail will continue to struggle, energy should get back towards trend as the PacifiCorp issues are resolved, but MSR will likely be flat at best.

My own optimistic outlook was ending 2024 in the $370-$380 range with very wide error bars around that obviously.

Jeff

So, the celebration should be that the big sudden loss in real purchasing power in the first half of last year has ended, and the recent flat spot is at least better than when we were losing real value gradually for the ~7 years up to the pandemic. But no positive value creation.

I agree with other comments that the real value is the optionality. From time to time there is an opportunity that only someone with a big cash pile can take advantage of.


OK, thanks. Your numbers are much better than my eyeballing and handwaving. And for the sake of argument, I will grant that this big slug of cash may someday enable a big, transformative purchase. That requires a lot of faith, with the dwindling number of candidates big enough to make so much cash necessary, and Buffett's lack of success in acquiring anything big and helpful in the last 10-15, the frustrating persistence of high prices for public companies, and the dwindling number of years that Buffett is likely to have to make this elephantine purchase. But let's just grant the argument that it's always good to have options, and this one may work someday. On top of that, it is still important to see the economic performance of that $160b in float, while we're waiting.

And your better numbers lead me to much the same conclusion. We have gone from about -1% return on the short-term treasuries for 8 years (practically the whole dollar amount of the float being invested in short-term treasuries), to one bad year of -11%, and now up to +1.5% in the last year. So over the last 10 years, that gives us -16.5% (average -1.6%). And now, finally, there's hope that we might be back to the usual positive spread, currently at +1.5%. That roughly 3% swing, from -1.5% to +1.5%, if it holds up, means that that big chunk of $160b in float may be worth a lot more than we might have thought, last year, when it seemed to be just losing money. 3% of $160b is about $5b, and even for a company as big as Berkshire, with $31b in operating earnings, that change in float return is a pretty big deal.

dtb

A wild guess as to how many buybacks there will have been in Q4.
October: 200 A-share equivalents
November: 115
December: 100

My guesses were considerable underestimates for October and November, but at least the December guess was pretty much spot on:

The actual numbers:
October: 1815.0 A-share equivalents
November: 2145.4
December: 103.0

The total for the year was extremely similar to the total in 2022, both much lower figures than the 2020 or 2021 rates.

The highest monthly average price paid per B share remains the September figure of $357.22. The figure has been rising maybe 2.50 - 4.00 per month in the last couple/few years. (For some months there are A shares purchased but not B shares. If you want to see what management might might have paid for B shares, the average price paid per A is hovering around 1540 times the average price paid per B, so you can convert using that number.)

Jim

Jim talking about BRK-A share buybacks:
If you want to see what management might might have paid for B shares, the average price paid per A is hovering around 1540 times the average price paid per B, so you can convert using that number.)

That 3 months of buybacks erases 2% of the votes outstanding!

So BRK board is willing to pay a 3% premium to concentrate voting power in the very long term buy-and-holds? My interpretation of why they would pay a premium to buy A's instead of B's in the buyback.

Do you have thoughts about the reasons?

If they weren't trying to preserve voting rights for someone, couldn't they just go to a single class of stock trading 1500 B's for each A?

R:

Post New

Berkshire Hathaway FAQ

---

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