BRK Valuation

[rb]

Vinod, instead of worrying about what the float will earn why don't you look at the float as a liability?

 

Then you can go about valuing the insurance business this way:

 

Investments + capitalized underwriting profits - PV of float liability - Goodwill.

[Thrifty3000]

At the end of 2019 my model assumed "normal" BRK.B look through earnings of $12 to $16 per share, and a growth rate of 6% to 8% (a base case of $14 per share look through earnings and 7% growth).

 

Since then I've impaired my estimates to incorporate post-covid margin compression for financials, and more modest growth expectations.

 

I now model for look through earnings of $8 to $12 per share, and growth of 5% to 7% (a base case of $10 per share and 6% growth).

 

For me, intrinsic value falls somewhere in the neighborhood of $240 per B share. I'm perfectly comfortable owning the stock at the current price, and would happily buy more below $160 per share.

 

Thanks for laying out the assumptions. Without knowing one's estimates of earnings, growth rate of earnings and discount rate, it is difficult to know where there is disagreement if any.

 

In the base case as of Q1, 2020, I have normalized earnings as $12.5 per share, excluding any earnings from float. Growth rate of 8.5%.

 

To me, it looks like any differences in IV estimates for BRK really boil down to

 

1) The value of float. This is harder to assess as it is dependent on interest rates and investment opportunities. One simple and likely very conservative way is to assume that $60-80 billion would always remain in cash and as long term interest rates may remain low for a long time, essentially are not worth much. So we end up valuing the rest of the float or about $50 billion.

 

Or assume it is going to earn some low rate on the total float and it ends up adding $0.5 to $1 per share.

 

2) The discount rate.

 

So you have ROE of about 8.5% which is consistent with the growth rate assumption as well with 100% earnings retention.

 

If you pick a discount rate of 7%, you have to have to look through 30 years to make it worth 1.5x book value.

 

The question I have for you is, if you assume a growth rate of 6%, why would it be worth even book value unless your discount rate is less than 6%?

 

Vinod

 

For a handful of index funds or index-like investments (aka Berkshire) that I hope to hold for decades I basically just capitalize growth in perpetuity. I know it's inelegant, but one benefit is I can do it in my head.

 

For Berkshire I do break down the groves in a spreadsheet, and slice and dice assumptions a number of different ways before landing on a look through earnings estimate.

 

(FWIW, when making an investment I assign a lot more significance to my look through earnings yield estimate than to my ability to forecast growth. Hence, Amazon hasn't made it into the 'ol portfolio just yet.)

 

I'm definitely comfortable with a long term BRK growth rate over 5%. The ROE, compensation structures, and capital allocation discipline ingrained in the culture give me that confidence.

 

(As an aside, it also doesn't hurt that:

 

a) Buffett pledged to give away 5% of his BRK holdings annually to the Gates Foundation

b) he is genetically hardwired to increase his wealth over the long term (with a margin of safety), so if history's greatest investor feels a minimum 5% long term growth rate for his own company is probably a safe bet, I'm more apt to accept that assumption.)

 

I do have trouble basing a long term growth assumption on a historical 8.5% ROE. They clearly can't reinvest organically at that rate, so it all comes down to acquisition prospects. And,

 

a) because the universe of sizable investment opportunities in Berkshire's circle of competence is now down to maybe a few dozen companies globally

b) because Buffett spent his entire life teaching/training an army of extremely well capitalized competitors (aka private equity) to do exactly what he does, while they're incentivized to pay higher prices, I can see scenarios where the elephant gun that used to be unloaded every 2 or 3 years, now gets shelved for 5 to 10 (or more) years at a time; leaving the next generation of management no choice but to accept lower growth while buying back shares or paying dividends.

 

I was comfortable with 7% last year. Now I'm more comfortable with 6%. I really really really hope it turns out to be at least 8.5%.

[vinod1]

Vinod, instead of worrying about what the float will earn why don't you look at the float as a liability?

 

Then you can go about valuing the insurance business this way:

 

Investments + capitalized underwriting profits - PV of float liability - Goodwill.

 

rb,

 

We can do BRK valuation different ways and I do that too. However, the only valuation approach on which I have confidence  is one based on owners earnings. When shit hits the fan, it is the only thing that gives me enough conviction on an investment. Nearly every investment mistake I made is because I misjudged earnings. I see that pattern among Buffett and other greats as well.

 

So my strong bias is to look at the earnings an investment can generate. Above, I am basically looking at all the earnings that BRK can generate (even though some are just look through) that accrue to the owners.

 

Coming back to float, you need to make some assumptions about it. Almost all the other parts of BRK are pretty simple to value. We can nit pick about the multiples a bit here and there, but there is not much ambiguity in most of the business. The thing that drives uncertainty in BRK valuation is float and its reinvestment opportunity. Even if you do not explicitly value float, you are basically making some assumption implicitly about its value. Better separate it out, that way you can change its value if something changes related to it.

 

Vinod

[vinod1]

At the end of 2019 my model assumed "normal" BRK.B look through earnings of $12 to $16 per share, and a growth rate of 6% to 8% (a base case of $14 per share look through earnings and 7% growth).

 

Since then I've impaired my estimates to incorporate post-covid margin compression for financials, and more modest growth expectations.

 

I now model for look through earnings of $8 to $12 per share, and growth of 5% to 7% (a base case of $10 per share and 6% growth).

 

For me, intrinsic value falls somewhere in the neighborhood of $240 per B share. I'm perfectly comfortable owning the stock at the current price, and would happily buy more below $160 per share.

 

Thanks for laying out the assumptions. Without knowing one's estimates of earnings, growth rate of earnings and discount rate, it is difficult to know where there is disagreement if any.

 

In the base case as of Q1, 2020, I have normalized earnings as $12.5 per share, excluding any earnings from float. Growth rate of 8.5%.

 

To me, it looks like any differences in IV estimates for BRK really boil down to

 

1) The value of float. This is harder to assess as it is dependent on interest rates and investment opportunities. One simple and likely very conservative way is to assume that $60-80 billion would always remain in cash and as long term interest rates may remain low for a long time, essentially are not worth much. So we end up valuing the rest of the float or about $50 billion.

 

Or assume it is going to earn some low rate on the total float and it ends up adding $0.5 to $1 per share.

 

2) The discount rate.

 

So you have ROE of about 8.5% which is consistent with the growth rate assumption as well with 100% earnings retention.

 

If you pick a discount rate of 7%, you have to have to look through 30 years to make it worth 1.5x book value.

 

The question I have for you is, if you assume a growth rate of 6%, why would it be worth even book value unless your discount rate is less than 6%?

 

Vinod

 

For a handful of index funds or index-like investments (aka Berkshire) that I hope to hold for decades I basically just capitalize growth in perpetuity. I know it's inelegant, but one benefit is I can do it in my head.

 

For Berkshire I do break down the groves in a spreadsheet, and slice and dice assumptions a number of different ways before landing on a look through earnings estimate.

 

(FWIW, when making an investment I assign a lot more significance to my look through earnings yield estimate than to my ability to forecast growth. Hence, Amazon hasn't made it into the 'ol portfolio just yet.)

 

I'm definitely comfortable with a long term BRK growth rate over 5%. The ROE, compensation structures, and capital allocation discipline ingrained in the culture give me that confidence.

 

(As an aside, it also doesn't hurt that:

 

a) Buffett pledged to give away 5% of his BRK holdings annually to the Gates Foundation

b) he is genetically hardwired to increase his wealth over the long term (with a margin of safety), so if history's greatest investor feels a minimum 5% long term growth rate for his own company is probably a safe bet, I'm more apt to accept that assumption.)

 

I do have trouble basing a long term growth assumption on a historical 8.5% ROE. They clearly can't reinvest organically at that rate, so it all comes down to acquisition prospects. And,

 

a) because the universe of sizable investment opportunities in Berkshire's circle of competence is now down to maybe a few dozen companies globally

b) because Buffett spent his entire life teaching/training an army of extremely well capitalized competitors (aka private equity) to do exactly what he does, while they're incentivized to pay higher prices, I can see scenarios where the elephant gun that used to be unloaded every 2 or 3 years, now gets shelved for 5 to 10 (or more) years at a time; leaving the next generation of management no choice but to accept lower growth while buying back shares or paying dividends.

 

I was comfortable with 7% last year. Now I'm more comfortable with 6%. I really really really hope it turns out to be at least 8.5%.

 

Thanks for the detailed response. Lots of good points.

 

What I am asking is what is your expected long term expected return on BRK? This is what people would call "required return".  If I understand it correctly, you seem to be saying you would be happy with 5%.

 

Vinod

[Thrifty3000]

At the end of 2019 my model assumed "normal" BRK.B look through earnings of $12 to $16 per share, and a growth rate of 6% to 8% (a base case of $14 per share look through earnings and 7% growth).

 

Since then I've impaired my estimates to incorporate post-covid margin compression for financials, and more modest growth expectations.

 

I now model for look through earnings of $8 to $12 per share, and growth of 5% to 7% (a base case of $10 per share and 6% growth).

 

For me, intrinsic value falls somewhere in the neighborhood of $240 per B share. I'm perfectly comfortable owning the stock at the current price, and would happily buy more below $160 per share.

 

Thanks for laying out the assumptions. Without knowing one's estimates of earnings, growth rate of earnings and discount rate, it is difficult to know where there is disagreement if any.

 

In the base case as of Q1, 2020, I have normalized earnings as $12.5 per share, excluding any earnings from float. Growth rate of 8.5%.

 

To me, it looks like any differences in IV estimates for BRK really boil down to

 

1) The value of float. This is harder to assess as it is dependent on interest rates and investment opportunities. One simple and likely very conservative way is to assume that $60-80 billion would always remain in cash and as long term interest rates may remain low for a long time, essentially are not worth much. So we end up valuing the rest of the float or about $50 billion.

 

Or assume it is going to earn some low rate on the total float and it ends up adding $0.5 to $1 per share.

 

2) The discount rate.

 

So you have ROE of about 8.5% which is consistent with the growth rate assumption as well with 100% earnings retention.

 

If you pick a discount rate of 7%, you have to have to look through 30 years to make it worth 1.5x book value.

 

The question I have for you is, if you assume a growth rate of 6%, why would it be worth even book value unless your discount rate is less than 6%?

 

Vinod

 

For a handful of index funds or index-like investments (aka Berkshire) that I hope to hold for decades I basically just capitalize growth in perpetuity. I know it's inelegant, but one benefit is I can do it in my head.

 

For Berkshire I do break down the groves in a spreadsheet, and slice and dice assumptions a number of different ways before landing on a look through earnings estimate.

 

(FWIW, when making an investment I assign a lot more significance to my look through earnings yield estimate than to my ability to forecast growth. Hence, Amazon hasn't made it into the 'ol portfolio just yet.)

 

I'm definitely comfortable with a long term BRK growth rate over 5%. The ROE, compensation structures, and capital allocation discipline ingrained in the culture give me that confidence.

 

(As an aside, it also doesn't hurt that:

 

a) Buffett pledged to give away 5% of his BRK holdings annually to the Gates Foundation

b) he is genetically hardwired to increase his wealth over the long term (with a margin of safety), so if history's greatest investor feels a minimum 5% long term growth rate for his own company is probably a safe bet, I'm more apt to accept that assumption.)

 

I do have trouble basing a long term growth assumption on a historical 8.5% ROE. They clearly can't reinvest organically at that rate, so it all comes down to acquisition prospects. And,

 

a) because the universe of sizable investment opportunities in Berkshire's circle of competence is now down to maybe a few dozen companies globally

b) because Buffett spent his entire life teaching/training an army of extremely well capitalized competitors (aka private equity) to do exactly what he does, while they're incentivized to pay higher prices, I can see scenarios where the elephant gun that used to be unloaded every 2 or 3 years, now gets shelved for 5 to 10 (or more) years at a time; leaving the next generation of management no choice but to accept lower growth while buying back shares or paying dividends.

 

I was comfortable with 7% last year. Now I'm more comfortable with 6%. I really really really hope it turns out to be at least 8.5%.

 

Thanks for the detailed response. Lots of good points.

 

What I am asking is what is your expected long term expected return on BRK? This is what people would call "required return".  If I understand it correctly, you seem to be saying you would be happy with 5%.

 

Vinod

 

My current “normalized” look through earnings estimate is $10 per B share. My long term expected growth rate is 6%. I usually use a discount rate of 9% or 10%.

[Munger_Disciple]

My current “normalized” look through earnings estimate is $10 per B share. My long term expected growth rate is 6%. I usually use a discount rate of 9% or 10%.

 

The $10 of current earnings are not distributable earnings, but retained earnings. In other words, the only way Berkshire can achieve growth rate of 6% in your model is by retaining the earnings. So you should re-work your model to come up with an intrinsic value estimate that incorporates this fact. It is not equal to $10/(10%-6%) = $250 per B share as you seem to be implying.

[Thrifty3000]

My current “normalized” look through earnings estimate is $10 per B share. My long term expected growth rate is 6%. I usually use a discount rate of 9% or 10%.

 

The $10 of current earnings are not distributable earnings, but retained earnings. In other words, the only way Berkshire can achieve growth rate of 6% in your model is by retaining the earnings. So you should re-work your model to come up with an intrinsic value estimate that incorporates this fact. It is not equal to $10/(10%-6%) = $250 per B share as you seem to be implying.

 

The company earns $10 in year 1. It retains and reinvests that $10 earned in year 1, and as a result, it earns an additional $.60 in year 2, for total year 2 earnings of $10.60. It does the same in year 3. It reinvests the $10.60 it retained in year 2, and earns $11.24. That pattern continues in perpetuity. In 24 years, for example, it will be earning $40 per share according to the model. Why does that need to be reworked? Seems straightforward.

[Munger_Disciple]

<i> In 24 years, for example, it will be earning $40 per share according to the model.</i>

 

Ok, what is your estimate of intrinsic value then? Please show the steps you use to arrive at it. You cannot use the standard DCF model because there are no cash flows that accrue to the owner during the 24 years; they are reinvested back in the company.

[Poor Charlie]

My current “normalized” look through earnings estimate is $10 per B share. My long term expected growth rate is 6%. I usually use a discount rate of 9% or 10%.

 

The $10 of current earnings are not distributable earnings, but retained earnings. In other words, the only way Berkshire can achieve growth rate of 6% in your model is by retaining the earnings. So you should re-work your model to come up with an intrinsic value estimate that incorporates this fact. It is not equal to $10/(10%-6%) = $250 per B share as you seem to be implying.

 

The company earns $10 in year 1. It retains and reinvests that $10 earned in year 1, and as a result, it earns an additional $.60 in year 2, for total year 2 earnings of $10.60. It does the same in year 3. It reinvests the $10.60 it retained in year 2, and earns $11.24. That pattern continues in perpetuity. In 24 years, for example, it will be earning $40 per share according to the model. Why does that need to be reworked? Seems straightforward.

 

Your’re capitalizing retained earnings rather than distributed earnings.  The value of $10 in earnings growing by 6% a year depends on how much has to be reinvested to produce the 6% growth.  Your economic assumptions ($10 initial earnings, 100% retention, 6% growth) and your valuation assumptions ($10 initial earnings, 0% retention, 6% growth) are not the same thing.  Using your 10% discount rate, you get the following present values for Berkshire’s operating earnings (i.e., the value of Berkshire excluding cash and securities):

 

 

(a) $10 earnings, 100% retention, 6% growth (using your terminal year of 24): ($40.49/.10) / 1.10^24 = $41.11

 

(b) $10 earnings, 0% retention, 6% growth (in perpetuity): $10 / (.10-.06) = $250

 

 

Under your economic assumptions of 100% retention, 6% growth and a 10% discount rate, Berkshire would be destroying value.  In fact, each dollar retained would be worth only 60 cents. 

 

Here’s another way to look at it: If Berkshire needs to retain 100% of its earnings to grow by 6% a year, they’re earning 6% on equity.  Berkshire is levered 2:1 and the liabilities cost zero (roughly).  You’re therefore assuming Berkshire will earn just 3% on the asset side.  Of course, assets that produce operating earnings only make up half the balance sheet, but you get the idea.

[Thrifty3000]

My current “normalized” look through earnings estimate is $10 per B share. My long term expected growth rate is 6%. I usually use a discount rate of 9% or 10%.

 

The $10 of current earnings are not distributable earnings, but retained earnings. In other words, the only way Berkshire can achieve growth rate of 6% in your model is by retaining the earnings. So you should re-work your model to come up with an intrinsic value estimate that incorporates this fact. It is not equal to $10/(10%-6%) = $250 per B share as you seem to be implying.

 

The company earns $10 in year 1. It retains and reinvests that $10 earned in year 1, and as a result, it earns an additional $.60 in year 2, for total year 2 earnings of $10.60. It does the same in year 3. It reinvests the $10.60 it retained in year 2, and earns $11.24. That pattern continues in perpetuity. In 24 years, for example, it will be earning $40 per share according to the model. Why does that need to be reworked? Seems straightforward.

 

Your’re capitalizing retained earnings rather than distributed earnings.  The value of $10 in earnings growing by 6% a year depends on how much has to be reinvested to produce the 6% growth.  Your economic assumptions ($10 initial earnings, 100% retention, 6% growth) and your valuation assumptions ($10 initial earnings, 0% retention, 6% growth) are not the same thing.  Using your 10% discount rate, you get the following present values for Berkshire’s operating earnings (i.e., the value of Berkshire excluding cash and securities):

 

 

(a) $10 earnings, 100% retention, 6% growth (using your terminal year of 24): ($40.49/.10) / 1.10^24 = $41.11

 

(b) $10 earnings, 0% retention, 6% growth (in perpetuity): $10 / (.10-.06) = $250

 

 

Under your economic assumptions of 100% retention, 6% growth and a 10% discount rate, Berkshire would be destroying value.  In fact, each dollar retained would be worth only 60 cents. 

 

Here’s another way to look at it: If Berkshire needs to retain 100% of its earnings to grow by 6% a year, they’re earning 6% on equity.  Berkshire is levered 2:1 and the liabilities cost zero (roughly).  You’re therefore assuming Berkshire will earn just 3% on the asset side.  Of course, assets that produce operating earnings only make up half the balance sheet, but you get the idea.

 

For an investment that doesn’t pay dividends the cash flows consist of

 

a) the initial investment

b) the amount received upon exit.

 

The ROI is a function of

 

a) earnings growth

b) earnings multiple at entry vs earnings multiple at exit

 

^that’s Jack Bogle 101

 

The longer you hold the investment the less of a factor multiple expansion/compression becomes, while the ROI gravitates toward the earnings growth rate (there is a nuanced relationship between ROE and ROI). It’s why paying a fair price for a great business over the long term works out better than paying a great price for a fair business.

 

^that’s Chuck Munger 101

 

I said I can quickly estimate the intrinsic value of a long term index-like investment by using a perpetual growth formula in my head. I didn’t say the result is what I would pay for the investment. I did mention I would get excited about buying at prices below $160 per share - or 16 times my estimated look through earnings (my opinion of a fair price for a good business).

 

If I buy and hold for several decades I would expect my ROI to equate to the long term growth rate.

 

In looking ahead over the next 2 to 6 decades I’m not giving Berkshire any benefits of the doubt regarding a future performance anywhere near past performance - even the recent past.

 

The company is gigantic, it has limited investment opportunities of scale, extreme competition for those investment opportunities, and its leadership is in transition.

 

Personally, I’m concerned the company isn’t being handed over to “leaders.” I think it’s being handed off to managers. There’s a big difference.

 

On the positive front you have Ajit and Todd. Ajit is the real deal when it comes to insurance. And, I’m generally excited about the prospects of capital being deployed by Todd Combs, as I get a sense he can learn to be as shrewd as Buffett.

 

I think Ted and Greg, however, are faking it until they don’t make it.

 

No manager will grasp Berkshire or the investment world the way Buffett did in his prime - when it was easier to grow. And, none of the next generation of senior management has shown any ability to communicate like a leader (David Sokol was the closest - but awkwardly arrogant). I suspect they will communicate like operators not wanting to screw up. And, thus begin the journey toward bigness and dumbness.

 

As far as the 2 to 1 leverage goes, Buffett has set an expectation for float to peak, and even decline, though slowly, in the not too distant future. Markets do saturate. I don’t necessarily believe it will decline, I certainly don’t model for it, but declining leverage would create a drag.

 

Estimating earnings growth of 6% and discounting those earnings at 10% is not destroying value. It’s producing an estimate of what I would pay for a 10% return on an earnings stream growing at 6%.

 

Retaining $10 to earn $.60 in perpetuity only destroys value if the multiple others are willing to pay for that $.60 is less than 16.67.

[Munger_Disciple]

My current “normalized” look through earnings estimate is $10 per B share. My long term expected growth rate is 6%. I usually use a discount rate of 9% or 10%.

 

The $10 of current earnings are not distributable earnings, but retained earnings. In other words, the only way Berkshire can achieve growth rate of 6% in your model is by retaining the earnings. So you should re-work your model to come up with an intrinsic value estimate that incorporates this fact. It is not equal to $10/(10%-6%) = $250 per B share as you seem to be implying.

 

The company earns $10 in year 1. It retains and reinvests that $10 earned in year 1, and as a result, it earns an additional $.60 in year 2, for total year 2 earnings of $10.60. It does the same in year 3. It reinvests the $10.60 it retained in year 2, and earns $11.24. That pattern continues in perpetuity. In 24 years, for example, it will be earning $40 per share according to the model. Why does that need to be reworked? Seems straightforward.

 

Your’re capitalizing retained earnings rather than distributed earnings.  The value of $10 in earnings growing by 6% a year depends on how much has to be reinvested to produce the 6% growth.  Your economic assumptions ($10 initial earnings, 100% retention, 6% growth) and your valuation assumptions ($10 initial earnings, 0% retention, 6% growth) are not the same thing.  Using your 10% discount rate, you get the following present values for Berkshire’s operating earnings (i.e., the value of Berkshire excluding cash and securities):

 

 

(a) $10 earnings, 100% retention, 6% growth (using your terminal year of 24): ($40.49/.10) / 1.10^24 = $41.11

 

(b) $10 earnings, 0% retention, 6% growth (in perpetuity): $10 / (.10-.06) = $250

 

 

Under your economic assumptions of 100% retention, 6% growth and a 10% discount rate, Berkshire would be destroying value.  In fact, each dollar retained would be worth only 60 cents. 

 

Here’s another way to look at it: If Berkshire needs to retain 100% of its earnings to grow by 6% a year, they’re earning 6% on equity.  Berkshire is levered 2:1 and the liabilities cost zero (roughly).  You’re therefore assuming Berkshire will earn just 3% on the asset side.  Of course, assets that produce operating earnings only make up half the balance sheet, but you get the idea.

 

+1

 

It is really dumb for a company to retain 100% of earnings and grow at a lower rate than its discount rate (alternately, the opportunity cost for shareholders). Such a management action destroys shareholder wealth. This is the main reason Buffett said the condition for earnings retention is to be able produce more than $1 of market value for each dollar retained.

 

[Thrifty3000]

My current “normalized” look through earnings estimate is $10 per B share. My long term expected growth rate is 6%. I usually use a discount rate of 9% or 10%.

 

The $10 of current earnings are not distributable earnings, but retained earnings. In other words, the only way Berkshire can achieve growth rate of 6% in your model is by retaining the earnings. So you should re-work your model to come up with an intrinsic value estimate that incorporates this fact. It is not equal to $10/(10%-6%) = $250 per B share as you seem to be implying.

 

The company earns $10 in year 1. It retains and reinvests that $10 earned in year 1, and as a result, it earns an additional $.60 in year 2, for total year 2 earnings of $10.60. It does the same in year 3. It reinvests the $10.60 it retained in year 2, and earns $11.24. That pattern continues in perpetuity. In 24 years, for example, it will be earning $40 per share according to the model. Why does that need to be reworked? Seems straightforward.

 

Your’re capitalizing retained earnings rather than distributed earnings.  The value of $10 in earnings growing by 6% a year depends on how much has to be reinvested to produce the 6% growth.  Your economic assumptions ($10 initial earnings, 100% retention, 6% growth) and your valuation assumptions ($10 initial earnings, 0% retention, 6% growth) are not the same thing.  Using your 10% discount rate, you get the following present values for Berkshire’s operating earnings (i.e., the value of Berkshire excluding cash and securities):

 

 

(a) $10 earnings, 100% retention, 6% growth (using your terminal year of 24): ($40.49/.10) / 1.10^24 = $41.11

 

(b) $10 earnings, 0% retention, 6% growth (in perpetuity): $10 / (.10-.06) = $250

 

 

Under your economic assumptions of 100% retention, 6% growth and a 10% discount rate, Berkshire would be destroying value.  In fact, each dollar retained would be worth only 60 cents. 

 

Here’s another way to look at it: If Berkshire needs to retain 100% of its earnings to grow by 6% a year, they’re earning 6% on equity.  Berkshire is levered 2:1 and the liabilities cost zero (roughly).  You’re therefore assuming Berkshire will earn just 3% on the asset side.  Of course, assets that produce operating earnings only make up half the balance sheet, but you get the idea.

 

+1

 

It is really dumb for a company to retain 100% of earnings and grow at a lower rate than its discount rate (alternately, the opportunity cost for shareholders). Such a management action destroys shareholder wealth. This is the main reason Buffett said the condition for earnings retention is to be able produce more than $1 of market value for each dollar retained.

 

Seriously? It's not ITS discount rate it's MY discount rate!

 

I think you might be confusing discount rate with cost of capital.

 

PS. -1

[Poor Charlie]

For an investment that doesn’t pay dividends the cash flows consist of

 

a) the initial investment

b) the amount received upon exit.

 

The ROI is a function of

 

a) earnings growth

b) earnings multiple at entry vs earnings multiple at exit

 

^that’s Jack Bogle 101

 

If the only cash you get is when you sell the investment (the terminal year), then why are you capitalizing the yearly earnings [$10 / (.10 - .06) = $250]?  You wouldn’t capitalize the reinvested interest when valuing a zero-coupon bond, so why would you capitalize the reinvested earnings when valuing an equity investment?

 

The longer you hold the investment the less of a factor multiple expansion/compression becomes, while the ROI gravitates toward the earnings growth rate (there is a nuanced relationship between ROE and ROI). It’s why paying a fair price for a great business over the long term works out better than paying a great price for a fair business.

 

^that’s Chuck Munger 101

 

I said I can quickly estimate the intrinsic value of a long term index-like investment by using a perpetual growth formula in my head. I didn’t say the result is what I would pay for the investment. I did mention I would get excited about buying at prices below $160 per share - or 16 times my estimated look through earnings (my opinion of a fair price for a good business).

 

If I buy and hold for several decades I would expect my ROI to equate to the long term growth rate.

 

In looking ahead over the next 2 to 6 decades I’m not giving Berkshire any benefits of the doubt regarding a future performance anywhere near past performance - even the recent past.

 

The company is gigantic, it has limited investment opportunities of scale, extreme competition for those investment opportunities, and its leadership is in transition.

 

Personally, I’m concerned the company isn’t being handed over to “leaders.” I think it’s being handed off to managers. There’s a big difference.

 

On the positive front you have Ajit and Todd. Ajit is the real deal when it comes to insurance. And, I’m generally excited about the prospects of capital being deployed by Todd Combs, as I get a sense he can learn to be as shrewd as Buffett.

 

I think Ted and Greg, however, are faking it until they don’t make it.

 

No manager will grasp Berkshire or the investment world the way Buffett did in his prime - when it was easier to grow. And, none of the next generation of senior management has shown any ability to communicate like a leader (David Sokol was the closest - but awkwardly arrogant). I suspect they will communicate like operators not wanting to screw up. And, thus begin the journey toward bigness and dumbness.

 

As far as the 2 to 1 leverage goes, Buffett has set an expectation for float to peak, and even decline, though slowly, in the not too distant future. Markets do saturate. I don’t necessarily believe it will decline, I certainly don’t model for it, but declining leverage would create a drag.

 

Not worth commenting on.

 

Estimating earnings growth of 6% and discounting those earnings at 10% is not destroying value. It’s producing an estimate of what I would pay for a 10% return on an earnings stream growing at 6%.

 

Retaining $10 to earn $.60 in perpetuity only destroys value if the multiple others are willing to pay for that $.60 is less than 16.67.

 

If Berkshire is earning 6% on incremental retained earnings, which is what you’re implying (6% growth * 100% retention = 6% ROEs), and you assume the market will discount cash flows at 10%, which is also what you’re implying [$10 earnings / (10% discount rate – 4% growth rate)], each dollar retained is indeed worth 60 cents.  For example, retained earnings at the end of year 24 for would increase by $508.16 [($40.49 ending earnings – $10 beginning earnings) / 6% ROE] but the market value would only increase by $304.89 [($40.49 ending earnings – $10 beginning earnings) / 10% discount rate].  This means that each dollar retained is worth 60 cents ($304.89 increase in market value / $508.16 increase in retained earnings).

 

Part of the problem is you’re conflating the market’s discount rate with your discount rate (or more specifically, your required rate of return).  Rather than plugging the wrong numbers into a formula it seems you don’t fully understand, try thinking about it without using a formula at all.  For instance, “I get $40.49 in cash flow every year after year 24.  The market will capitalize cash flows at 6.25% (using the 16x you used above, but pick whatever multiple you want), which means I can exchange my $40.49 cash flow stream for an upfront payment of $647.83 at the end of year 24.  I want a 10% compounded return (what you say you’d pay for an “earnings cash flow stream growing at 6%”), so I’d be willing to pay anything less than $65.77 for this now ($647.84 / 1.10^24).”

 

 

 

[Jurgis]

It is really dumb for a company to retain 100% of earnings and grow at a lower rate than its discount rate (alternately, the opportunity cost for shareholders). Such a management action destroys shareholder wealth. This is the main reason Buffett said the condition for earnings retention is to be able produce more than $1 of market value for each dollar retained.

 

OT? Something useful out of this discussion.  8) ./bow

 

Let me see if I understood this correctly in a broader context. Let's take a company that produces FCF. But it does not pay it out some way. Instead it keeps it and does not invest it (at high/reasonable ROI). Then really we cannot just DCF that FCF at the time it was produced since we are not really getting it. We'd have to assume it's only paid out later (if ever) and add additional time periods (and discounts) to DCF model to estimate when it is paid out.

 

Of course, we could still assume that if we owned the whole business then we'd be able to use the FCF as we wish. And we can do the basic DCF based on that "wholly owned" assumption. It's just that in reality the business won't be worth as much as our basic DCF indicates.

 

Pretty interesting, n'est-ce pas?  8)

 

I agree that Thrifty3000 is confused, but I'm not sure how to explain to them why their calculations are incorrect.

[Munger_Disciple]

Thrifty3000,

 

It is clear to me that you don't understand (and perhaps don't want to understand) how to think about retained earnings that enable future growth and free cash flow that can be distributed to shareholders. I belatedly realize that there is no point in replying to your posts and I am sure the feeling is mutual. So let's leave it at that.

 

-MD

[Poor Charlie]

Thrifty3000,

 

It is clear to me that you don't understand (and perhaps don't want to understand) how to think about retained earnings that enable future growth and free cash flow that can be distributed to shareholders. I belatedly realize that there is no point in replying to your posts and I am sure the feeling is mutual. So let's leave it at that.

 

-MD

 

I actually enjoy conversations on valuation math.  I’m always looking to improve my understanding, so I welcome people poking holes in my posts.  I just ask that you put forth coherent arguments, which we don’t seem to be getting from Thrifty.   

[Thrifty3000]

Thrifty3000,

 

It is clear to me that you don't understand (and perhaps don't want to understand) how to think about retained earnings that enable future growth and free cash flow that can be distributed to shareholders. I belatedly realize that there is no point in replying to your posts and I am sure the feeling is mutual. So let's leave it at that.

 

-MD

 

I actually enjoy conversations on valuation math.  I’m always looking to improve my understanding, so I welcome people poking holes in my posts.  I just ask that you put forth coherent arguments, which we don’t seem to be getting from Thrifty. 

 

As soon as I sober up I'll see if I can clarify. Haha. Geez, I feel like I'm being baited.

 

But, fine, if it makes you trolls happy...

 

Normalized Per-Share Look Through Earnings - Year 0: $10.31

 

Projected Earnings:

 

Year 1: $10.93

Year 2: $11.58

Year 3: $12.28

Year 4: $13.02

Year 5: $13.80

 

Present Value of Projected Earnings - Years 1 Through 5 (10% Discount Rate): $46.19

 

Residual Value:

Residual Earnings (Year 5 Earnings x 1.06): $14.62

Capitalization Rate: 6%

Capitalized Residual Value: $243.75

 

Present Value of Capitalized Residual Value (10% Discount Rate): $151.35

 

Total Present Value of Projected Look Through Earnings: $46.19 + $151.35 = $197.54

 

And there you have it. $197.54 is the precise intrinsic value of a Berkshire Hathaway B share down to the penny. I can carry it out to more decimals if you like. Haha.

 

Obviously, you can model it out for more than 5 years, play with all the variables, add more inputs and variables, and derive pretty much whatever intrinsic value you want (why Warren Buffett has never wasted time on a DCF model, and why I feel slightly disgusted for wasting my own time).

 

For me, the most important inputs are the normal earnings estimate and growth estimate. Naturally, those are the two numbers I originally shared, and are the two items I wish others would share and intelligently discuss (that goes for all investments on COBF), especially if they aren't just pulled out of the air, and are the result of quality research and personal experience.

 

If you want to add value while attacking my (or anyone's) work, personally I'd prefer you challenge me to defend the effort and assumptions that went into deriving the look through earnings and growth estimates.

 

Another way to be a hero would be to post your own estimates, and subject yourself to a little scrutiny.

 

Bring it on!

 

 

 

 

[Poor Charlie]

Thrifty3000,

 

It is clear to me that you don't understand (and perhaps don't want to understand) how to think about retained earnings that enable future growth and free cash flow that can be distributed to shareholders. I belatedly realize that there is no point in replying to your posts and I am sure the feeling is mutual. So let's leave it at that.

 

-MD

 

I actually enjoy conversations on valuation math.  I’m always looking to improve my understanding, so I welcome people poking holes in my posts.  I just ask that you put forth coherent arguments, which we don’t seem to be getting from Thrifty. 

 

As soon as I sober up I'll see if I can clarify. Haha. Geez, I feel like I'm being baited.

 

But, fine, if it makes you trolls happy...

 

Normalized Per-Share Look Through Earnings - Year 0: $10.31

 

Projected Earnings:

 

Year 1: $10.93

Year 2: $11.58

Year 3: $12.28

Year 4: $13.02

Year 5: $13.80

 

Present Value of Projected Earnings - Years 1 Through 5 (10% Discount Rate): $46.19

 

Residual Value:

Residual Earnings (Year 5 Earnings x 1.06): $14.62

Capitalization Rate: 6%

Capitalized Residual Value: $243.75

 

Present Value of Capitalized Residual Value (10% Discount Rate): $151.35

 

Total Present Value of Projected Look Through Earnings: $46.19 + $151.35 = $197.54

 

And there you have it. $197.54 is the precise intrinsic value of a Berkshire Hathaway B share down to the penny. I can carry it out to more decimals if you like. Haha.

 

Obviously, you can model it out for more than 5 years, play with all the variables, add more inputs and variables, and derive pretty much whatever intrinsic value you want (why Warren Buffett has never wasted time on a DCF model, and why I feel slightly disgusted for wasting my own time).

 

For me, the most important inputs are the normal earnings estimate and growth estimate. Naturally, those are the two numbers I originally shared, and are the two items I wish others would share and intelligently discuss (that goes for all investments on COBF), especially if they aren't just pulled out of the air, and are the result of quality research and personal experience.

 

If you want to add value while attacking my (or anyone's) work, personally I'd prefer you challenge me to defend the effort and assumptions that went into deriving the look through earnings and growth estimates.

 

Another way to be a hero would be to post your own estimates, and subject yourself to a little scrutiny.

 

Bring it on!

 

My posts had nothing to do with what I think Berkshire is worth.  I was simply trying to point out that $10 in retained earnings aren’t the same thing as $10 in distributed earnings.  You can’t capitalize earnings like you did (and continue to do) unless those earnings are distributed.  Would you capitalize the reinvested interest when valuing a zero-coupon bond?  If not, then why would you capitalize the retained earnings of Berkshire (Berkshire, which retains all its capital, is effectively a zero-coupon equity)?  Aren’t they the same thing?

 

If we can’t agree retained earnings ≠ distributed earnings, then I’m sorry for wasting your time.  I was just looking to join a conversation about something I find interesting. 

 

All the best

 

[mcliu]

He doesn't get it. But it's a nice demonstration of commitment bias.

[Thrifty3000]

Thrifty3000,

 

It is clear to me that you don't understand (and perhaps don't want to understand) how to think about retained earnings that enable future growth and free cash flow that can be distributed to shareholders. I belatedly realize that there is no point in replying to your posts and I am sure the feeling is mutual. So let's leave it at that.

 

-MD

 

I actually enjoy conversations on valuation math.  I’m always looking to improve my understanding, so I welcome people poking holes in my posts.  I just ask that you put forth coherent arguments, which we don’t seem to be getting from Thrifty. 

 

As soon as I sober up I'll see if I can clarify. Haha. Geez, I feel like I'm being baited.

 

But, fine, if it makes you trolls happy...

 

Normalized Per-Share Look Through Earnings - Year 0: $10.31

 

Projected Earnings:

 

Year 1: $10.93

Year 2: $11.58

Year 3: $12.28

Year 4: $13.02

Year 5: $13.80

 

Present Value of Projected Earnings - Years 1 Through 5 (10% Discount Rate): $46.19

 

Residual Value:

Residual Earnings (Year 5 Earnings x 1.06): $14.62

Capitalization Rate: 6%

Capitalized Residual Value: $243.75

 

Present Value of Capitalized Residual Value (10% Discount Rate): $151.35

 

Total Present Value of Projected Look Through Earnings: $46.19 + $151.35 = $197.54

 

And there you have it. $197.54 is the precise intrinsic value of a Berkshire Hathaway B share down to the penny. I can carry it out to more decimals if you like. Haha.

 

Obviously, you can model it out for more than 5 years, play with all the variables, add more inputs and variables, and derive pretty much whatever intrinsic value you want (why Warren Buffett has never wasted time on a DCF model, and why I feel slightly disgusted for wasting my own time).

 

For me, the most important inputs are the normal earnings estimate and growth estimate. Naturally, those are the two numbers I originally shared, and are the two items I wish others would share and intelligently discuss (that goes for all investments on COBF), especially if they aren't just pulled out of the air, and are the result of quality research and personal experience.

 

If you want to add value while attacking my (or anyone's) work, personally I'd prefer you challenge me to defend the effort and assumptions that went into deriving the look through earnings and growth estimates.

 

Another way to be a hero would be to post your own estimates, and subject yourself to a little scrutiny.

 

Bring it on!

 

My posts had nothing to do with what I think Berkshire is worth.  I was simply trying to point out that $10 in retained earnings aren’t the same thing as $10 in distributed earnings.  You can’t capitalize earnings like you did (and continue to do) unless those earnings are distributed.  Would you capitalize the reinvested interest when valuing a zero-coupon bond?  If not, then why would you capitalize the retained earnings of Berkshire (Berkshire, which retains all its capital, is effectively a zero-coupon equity)?  Aren’t they the same thing?

 

If we can’t agree retained earnings ≠ distributed earnings, then I’m sorry for wasting your time.  I was just looking to join a conversation about something I find interesting. 

 

All the best

 

If you can’t understand your mom ≠ a girlfriend then I cant help you. Haha

 

Everyone on this forum understands earnings retained by a company aren’t distributed. I’m dying to hear what other brilliant insights you have.

 

I recommend taking your ample free time to read some of Buffett’s annual letters about how to consider retained / look through / owner earnings.

[LC]

Put it this way.

 

CashCo dividends out $50 this year and $50 next year, then immediately goes BK.

 

InvestCo retains $50 this year and dividends out $100 next year, then also goes BK.

 

What is each worth.

[mcliu]

If you can’t understand your mom ≠ a girlfriend then I cant help you. Haha

 

Everyone on this forum understands earnings retained by a company aren’t distributed. I’m dying to hear what other brilliant insights you have.

 

I recommend taking your ample free time to read some of Buffett’s annual letters about how to consider retained / look through / owner earnings.

You are double counting the earnings without considering how growth is generated.

[Thrifty3000]

Put it this way.

 

CashCo dividends out $50 this year and $50 next year, then immediately goes BK.

 

InvestCo retains $50 this year and dividends out $100 next year, then also goes BK.

 

What is each worth.

 

Let’s assume a 10% discount rate...

 

CashCo: Pre-tax PV of year 1 dividend = $45; year 2 = $40.5. Sum = $85.5

 

InvesCo: Pre-tax PV of year 2 dividend = $81

 

Am I getting the hang of it?

 

Now, my question for you. How come the retained earnings didn’t generate additional value in year 2?

[LC]

You can play with the #s however you like, the point is that companies which can grow earnings without reinvesting capital (and therefore can distribute it to owners) are much more valuable.

[Thrifty3000]

You can play with the #s however you like, the point is that companies which can grow earnings without reinvesting capital (and therefore can distribute it to owners) are much more valuable.

 

Ok, I totally agree with you there. I feel like that’s where my 6% growth assumption comes into play. I’ve seen growth assumptions from others as high as 8.5%.

 

But, I’m pretty sure Buffett would not have stopped buying back shares in Q1 if he thought long term growth of 8.5% was feasible.