Brookfield Corporation

IThe business

Brookfield Corporation, written BN, owns three businesses and says so in its own words: asset management, wealth solutions and a group of operating businesses in infrastructure, energy, private equity and real estate.³ It is a holding company with a very large asset manager in the middle. It lists in Toronto and New York under the same symbol, and its partners have built it over about 30 years, with a stated aim of 15% a year or more in returns to shareholders.³

The company reports in US dollars under IFRS, and so does this page, with US$ in front of every company figure. The shares trade in both cities, so the headline price and my values are in Canadian dollars, with US dollars beside them. The exchange rate is 1.3978, the Bank of Canada’s average on the price date.⁵ The financial year ends on 31 December. The latest figures are for the second quarter, to 30 June 2026; the third quarter has not been reported yet.³ The key figure is not net income, which was US$1.43B for the twelve months to June 2026, but distributable earnings, the company’s own measure of cash that reaches the parent, which was US$6.17B, or US$5.65B before realized carried interest and gains.³ I come back to the gap between the two in Financials and Valuation.

Three businesses and a balance sheet

Asset management

Brookfield Asset Management (BAM) has over US$1 trillion of assets under management.⁴ It earns fees for managing money for other people. The fee-bearing capital, the amount on which it charges fees, was US$672B at 30 June 2026, up 19% in a year. Fee-related earnings, which are fees less the cost of earning them, were US$3.2B over the twelve months, up 20% in the quarter, at a margin of 57%.⁴ BAM raised US$163B in the twelve months, a record US$77B of it in the second quarter.⁴ Brookfield Corporation owns 74% of BAM, 1,193M shares, and BAM’s own distributable earnings at that share are about US$1.94B a year, after US$126M of borrowing cost.⁴ Outside holders own the other 26%, which is why BAM is also a separate listed stock. Its price was US$51.47 at the price date.⁸

The second asset-management profit stream is carried interest, a share of the profit on a fund once the clients have had their minimum return. At 30 June the funds had US$12.5B of accumulated unrealized carried interest, US$6.8B net of costs and of the part that belongs to others, against US$520M of net carry actually realized in the last twelve months.⁴ The company also estimates a “target” of US$2.7B a year net, which it values at ten times, or US$27.3B.⁴ That ten-times figure is a straight-line guess at what funds would earn if every fund hit its target return, and it is the biggest single judgement in the company’s own estimate of worth. In 2025 the realized figure was US$560M, in 2024 US$403M.⁹

Brookfield also keeps US$10.6B of its own money in funds that BAM runs, mainly real estate, which paid it US$945M over the twelve months, and about US$2.8B of its own capital in technology and AI-linked businesses, which the supplemental lists as including SpaceX, Figure, a joint venture with OpenAI, and Anthropic.⁴ The company completed the purchase of the remaining 26% of Oaktree, its credit arm, on 31 July 2026 for about US$3B in cash and other consideration.¹⁰¹¹

Wealth solutions

Brookfield Wealth Solutions (BWS) is an insurance group. It sells annuities, pension buyouts and funding agreements, 96% of its liabilities, plus some property and casualty cover.⁴ Insurance assets were US$191B at 30 June, after the acquisition of the UK’s Just Group added US$45B. The business invests the premiums, 49% in public credit and 33% in private credit, and earns the difference between that yield and what it owes policyholders. The yield was 5.23% over the twelve months, the cost of the liabilities 3.29% and the “gross spread” 1.94% in all, with an average duration of nine years.⁴ Distributable earnings were US$1.76B over the twelve months and US$480M in the second quarter, up 23%.⁴ The property and casualty book ran at a combined ratio of 99% in the quarter, which means it roughly broke even before investment income.³ The company says the invested capital in the business is US$15.0B, on which it targets 15% or more a year.⁴

BWS was created about five years ago as a paired security to BN, and has gathered about US$30B of capital and close to US$200B of assets.¹² It is the part of Brookfield that is least like an asset manager and most like a lender with a very long, very large balance sheet.

Operating businesses

The operating businesses carry about US$50B of capital at the company’s values.⁴ Infrastructure is a 26% stake in Brookfield Infrastructure Partners (BIP), which owns utilities, transport, midstream and data assets; energy is 45% of Brookfield Renewable (BEP), hydro, wind and solar; private equity is the Brookfield Business Corporation (BBUC), 27% directly and 69% with BWS; and real estate (BPG) is wholly owned, with 34 of the world’s premier office and retail complexes, a group of 134 assets it plans to sell, and a homebuilding arm.⁴ Together they sent Brookfield US$1.55B in cash over the twelve months.⁴ That is a small part of what they earn, because most of it is reinvested in them or belongs to outside holders: BBUC, for one, paid US$24M against US$375M of operating FFO at Brookfield’s share.⁴ The Corporation values the listed three at their stock prices and real estate at fair value.

Management’s growth story sits here too. On the second-quarter call Bruce Flatt, the chief executive, described a US$100B AI campus in Kentucky on federal land, a further US$17.5B Department of Energy financing for Westinghouse, which has reactors under construction, and a US$25B Bloom Energy partnership for fuel cells at data centres.⁶³ These are very large numbers that sit mostly in funds and affiliates, so they add to fees before they add to Brookfield’s own earnings.

The structure, and why it is changing

The layers matter, because every one of them is a place where value can leak or hide. The parent owns 74% of a listed manager, almost all the earnings of a listed insurer whose shares are paired to its own, and minority stakes in three listed affiliates, while the affiliates pay fees and incentive distributions back to BAM, of which the parent owns 74%.⁴ Consolidated total assets are US$525.5B, of which non-controlling interests, the outside owners of subsidiaries, own US$120.1B of the equity and non-recourse lenders are owed US$250.3B.³ Only 6% of consolidated debt is issued by, or has recourse to, the Corporation.⁴ The partners, a group of current and former executives, own the Class B shares through a trust, which carry the right to elect half the board.⁷¹³

Shareholders approved a fix on 16 July 2026. A new parent, Brookfield Corporation Ltd. (“New BN”), will be formed in BWS’s home jurisdiction and will own BN and BWS. Each BN and BWS Class A share becomes one New BN share, listed in Toronto and New York as BN, and the old BN and BWS shares are delisted. Canadian taxable holders may elect exchangeable shares to defer tax. It is expected to close by year-end.¹³¹²⁷ The stated reasons are a single listed security, a route into US indexes, access for the insurer to the parent’s capital, which the company puts at about US$145B, and distributions that are expected to be returns of capital.¹²⁷ The circular expects about 2.52B Class A shares after closing, including the exchangeable ones, against the 2,366M diluted shares the company uses today.⁷⁴ I cannot reconcile those two counts from the documents, and I show what the higher count does to my values in Valuation.

The state of play now

The second quarter was good. Distributable earnings before realizations were up 15% a share, to US$0.61. Fundraising was a record, the insurer grew 23%, and Brookfield sold US$40B of assets in the first half, including One Churchill Place in London for £750M.³ The balance sheet kept its access: C$750M of ten- and thirty-year notes in April, which the company says drew four times the orders it needed, and US$600M of 5.65% notes due 2031 in September.⁶¹⁴ The share price closed at US$41.46 in Q4 2026, 16% below its January high and 15% above the low of 14 October.² The third-quarter results are due in November.

IIThe moat

The test of a durable business is what happened to its profit when conditions were bad. Brookfield has two hard tests in the last six years. In early 2020 the shares fell 52% from their high to the close of 23 March.² In 2022, as interest rates rose, the price fell from US$32.79 at the end of 2021 to US$20.97 a year later.¹⁵ Distributable earnings a share did not fall in either: they were US$1.22 in 2021, US$1.59 in 2022 and US$1.77 in 2023 on the company’s restated figures, before realizations R (reported).¹⁵ That is a good result. It says the fee, insurance and operating cash flows held. It does not say that the values held, and the values are where my doubt sits.

What does protect it

• Five sources of capital. Mr. Flatt lists them as public markets, institutions, private wealth, long-term insurance money and the company’s own balance sheet, and says the mix lets Brookfield pick the right one for each deal.⁶ The evidence is the US$163B raised in twelve months and a fee base of US$672B that grew 19%, against US$563B a year earlier.⁴
• Operating real assets. It does not only own pipes, wires and towers. It runs them, and the filings show contracted, inflation-linked cash flows: infrastructure operating funds from operations (FFO) rose 9% in the year, and energy 14% before its energy contracts.⁴
• Permanent capital. The listed affiliates and the insurer provide money that does not need to be refunded every ten years, and that is an unusual thing for an asset manager.⁴ It means Brookfield can keep buying when the market for fund money is shut.
• A fee base that compounds. Fee-related earnings were US$2.46B in 2024, US$3.00B in 2025 and US$3.20B over the twelve months to June 2026.⁹⁴ Fees come before profit, and once capital is committed it is tied up for years.
• Credit. Standard & Poor’s rates the senior debt A-, Moody’s A3, Fitch A- and DBRS A, all with a stable outlook.¹⁶ The weighted-average term of corporate debt is 15 years at 4.8%, and nothing falls due in 2026.⁴

Where it is thin

The values are the company’s own. More than half of the company’s US$176.5B of assets, before debt, is marks the company controls: US$28.0B of real estate at IFRS fair value, US$30.5B of wealth solutions at 15 times, and US$34.0B of carried interest at ten times a target.⁴ I treat each as a number to haircut and not as a fact. The real estate book, for example, pays a cash yield of 2.3% on the US$28.0B equity: US$638M over the twelve months.⁴ Appraisal discount rates on the 34 top properties run between 5.8% and 7.8%, so the values are as good as the appraisal rates.⁴

The IFRS earnings do not show the strength. Net income attributable to shareholders was US$1.43B over the twelve months and US$1.31B in 2025, against operating cash flow of US$11.0B on a consolidated basis.³¹⁷ Over ten years net income to shareholders added up to US$18.5B while total assets went from US$160B to US$519B C (calculated).¹⁷ A business with a clear edge would show IFRS profit rising with assets. The company’s answer is that depreciation, fair-value changes and the profit of outside owners distort the IFRS figure, and that distributable earnings are the cash that reaches owners. Both statements are partly true, and I treat distributable earnings as the number to test, not as a substitute for the audit.

Competitors can copy the model. Raising money from institutions and individuals, buying an insurer, and mixing credit with real assets is not secret. What cannot be copied quickly is the owned operating base. What can be copied is the bet that private markets will keep taking a larger share of retirement savings, which Mr. Flatt called an important source of long-term capital on the second-quarter call.⁶ If that source slows, everyone slows with it.

Private credit. The company’s own letter says “much has been written lately about the risks in private credit.” It argues that its credit book sits in real asset finance, asset-backed lending and opportunistic credit, not in sponsor-backed loans to software companies, and that redemption requests at its business development company fell below 5% in the quarter when others saw high double digits.¹⁰ That is a claim from the company, and it matters because credit is 20% of the target carry and a growing part of the fee base, and 33% of the insurance portfolio is private credit.⁴

IIIThe competitive advantage

Brookfield competes for money in three places. For pension funds and sovereign funds it competes with the other large alternative managers: Blackstone, KKR, Apollo, Ares and Carlyle are the names I would put beside it. For individuals it competes with the same group plus the banks. For deals it competes with all of them and with the corporations that own the assets. I have not put a rival’s figures on this page, because each reports on its own basis, and a comparison on a mixed basis would mislead more than it helped.

What only an owner who operates can offer

The case Brookfield makes for itself is that it can do things a financial buyer cannot. On the second-quarter call Mr. Flatt gave the example of the US$100B AI campus in Kentucky on a federal site: the Department of Energy chose Brookfield, he said, because it can bring together land, power, transmission and capital at scale.⁶ The pieces exist inside the group. The property business has the land and the development skills. The energy business has generation. The infrastructure business has the transmission and data experience. The credit business can finance it, and the insurer can buy the long, safe part of the debt. That is a real combination. A pure manager has to rent each piece.

The same logic explains why the insurer exists. Wealth solutions invested over US$5B in real asset strategies in the quarter and US$16B over the twelve months, with an average net investment yield of 5.7% in the North American business.³ A buyer of long-dated, inflation-linked assets that wants 15% on its equity is a natural customer for what the group builds. It also means Brookfield is its own best customer, which I come back to under conflicts.

Where the edge is, and where I cannot see it

Some of the edge shows in the numbers. The fee base grew 19% in a year and the insurer 23%, while the cost of corporate debt was 4.8% at a 15-year term.⁴ Funds and affiliates raise money at the same time. The seventh private equity fund raised US$7B and the sixth infrastructure fund US$9B in the quarter, and the company says both are on track to be the largest of their series.³ A manager with no edge in a hard fundraising market does not get that.

Some of it I cannot see. The fund returns are given in terms of target and gross IRR; for the opportunistic real estate funds the company’s own table shows gross IRRs of 21%, 12%, 10% and 7% for the vintages of 2012, 2015, 2018 and 2021.⁴ The newest large fund is the weakest so far, though it is young and was invested at 2021 prices. Net returns to clients, which is what sets whether they come back, are not in the document I read. A manager’s edge is its clients’ net return, so on this point I am taking the fundraising numbers as a stand-in.

Conflicts of interest

Brookfield sits on several sides of its own deals. It owns the manager, holds stakes in the funds and affiliates the manager runs, and sells assets to funds it manages and to its insurer. The company says its wealth solutions business may buy investments “from the Corporation and its related party affiliate entities,” and that over the last twelve months it moved partial interests in six of its top properties into the insurer.¹⁰⁴ The affiliates in turn pay fees and incentive distributions to BAM: US$334M for BIP and US$156M for BEP over the twelve months.⁴ Each transaction may be fair. I cannot check the prices from outside, and neither can the policyholders whose money is on the other side. This is a cost of the model, and I do not value it, but I have kept it in mind when I choose how much to haircut the property and insurance marks.

Against its two nearest relatives, Brookfield offers the most reinvestment and the least cash to owners. Berkshire pays no dividend and keeps everything; Brookfield pays US$0.28 a share a year, a yield of 0.7%, and keeps the rest.³ That puts the weight on what management does with the retained money, which is the subject of the next section.

IVManagement

Bruce Flatt is chief executive. Nick Goodman is president, and the 2025 pay table lists him as president and chief financial officer. Connor Teskey runs BAM, Sachin Shah runs the insurer, and Justin Beber is chief operating officer.⁶⁷¹¹ They are partners, a group of current and former senior executives who, in the circular’s words, have owned a large part of Brookfield together for over 50 years.⁷ The team has been stable, and the bench is deep. What I want to know is what they are paid to do, how much of their own money is in, and what they did with the cash.

Who owns it, and who votes

As of 28 May 2026 the partners together owned about 471M Class A shares, 20% of the total, 455M BAM shares, 28% of BAM, and 40M BWS shares, 60% of BWS.⁷ At US$41.46 the Class A stake is about US$19.5B C (calculated). Mr. Flatt holds 27.0M Class A and BWS shares directly and through holding companies, about US$1.1B, plus his share of the larger holdings.⁷ That is a lot of personal money in the same shares as mine. It also means the partners are exposed to the same risks I describe.

The partners also own the Class B shares, 85,120 of them, through a trust. Those shares elect eight of the 16 directors; the Class A holders elect the other eight.¹³ Mr. Flatt has 41% of the beneficial interests in the trust, and eight others share the rest, and the circular says no single person controls it.⁷ So the outside shareholders vote for half the board, and the other half is elected by the insiders. I have no complaint about the people on the board, and I count the structure as a limit on what a dissatisfied owner can do.

How they are paid, and what that pays them to do

Mr. Flatt’s pay for 2025 was US$5.74M from BN: a salary of US$375,000 and US$5.36M of “escrowed” shares. BAM paid him another US$375,000 and US$3.22M of escrowed shares.⁷ Mr. Goodman received US$8.53M, mostly escrowed shares plus a cash incentive, and Mr. Beber US$9.10M, which includes US$2.9M of advances under carried interest plans.⁷ Together the named officers received US$26.4M in 2025, 0.5% of FFO.⁷ The amount is small next to the businesses. What matters is the form.

Escrowed shares are shares in a private company that holds Brookfield stock. They vest over five years and can be held for up to 7.5 years, and their value to the executive is the price of the shares over that time.⁷ In the pay table I read, there is no fixed target such as distributable earnings a share or return on capital. The executives are paid in the thing I own, over a long hold. That is the right kind of pay. Its limit is that the share price can rise for reasons that are not about owners’ cash, such as a higher multiple on a promoted figure, and the plan does not tell the executive which is which. The company describes the escrowed plan as economically non-dilutive, because it cancels an equal number of shares when the holding companies are wound up, and options and escrowed shares together were 2.3% of the shares at the end of 2025.⁷ The company’s count of options, plan shares and exchangeable shares, 133M, is about 6% of the shares outstanding, and I include all of them in my share count.⁴

The shares did well for owners. Over the five years to the end of 2025 C$100 invested in BN with dividends reinvested grew to C$438, against C$211 for the S&P/TSX Composite.⁷ The plan shown at the 2026 Investor Day doubles core earnings again by 2031 and takes the company’s own value per share from US$67 to US$140.²¹

What they did with the money

Over a full cycle, three things stand out. The first is that Brookfield has twice handed owners a large piece of a business. In June 2021 it distributed the wealth solutions business, US$0.24 a share, and in December 2022 it distributed 25% of BAM, US$5.33 a share.¹⁵ That is a record of handing owners assets, not only keeping them. The second is the buyback. In the last two years the company bought 49.7M shares, 2.2% of those outstanding, for US$1.8B, and issued 14.4M under its plans, so the share count fell 1.3%, from 2,262.3M to 2,233.0M.⁴ That mostly offsets the stock plans and shrinks the count a little. The third is where the rest went.

The buyback was US$695M, 11% of earnings, at an average US$42.04, against a company estimate of US$66.64 a share for what the business is worth.⁴ If management believes its own number, the shares were available at 63% of it. If it does not, it is spending US$5.6B a year on wealth solutions and the operating businesses at prices it considers fair, which is a different statement. I do not say which. I say the two numbers do not fit together comfortably, and the stronger test of the company’s view is what it buys with its money when its own stock is that cheap.

What did the reinvestment earn? In the last year distributable earnings before realizations went up US$0.15 a share, from US$2.24 to US$2.39, while the company kept about US$2.06 a share of earnings C (calculated).⁴ That is a crude measure, since fee growth needs little capital and insurance capital takes years to earn, but it is the only one I can build from the documents. Over five years the company says it grew core earnings at 16% a year, and Mr. Goodman said it “marginally exceeded” the plan it set in 2021.²¹ The numbers I can check agree with that: US$1.22 a share in 2021, restated, and US$2.39 for the last twelve months, which is 16% a year.¹⁵⁴

Two deals are fresh. The remaining 26% of Oaktree cost about US$3B, in cash and other consideration, valuing all of Oaktree at roughly US$11.5B C (calculated).¹⁰ The company does not give Oaktree’s earnings in the documents I read, so I cannot say what multiple that is, and I treat the purchase as value-neutral at the price. The Just Group acquisition added US$45B of insurance assets, and its price was not in the documents I read.³

VFinancials

Ten years of consolidated figures tell you how large Brookfield has become and little about how much it earns for its owners. Total assets went from US$160B in 2016 to US$519B in 2025, equity with the outside owners’ share from US$70B to US$166B, and operating cash flow from US$3.1B to US$11.0B.¹⁷ Net income to shareholders was US$1.7B in 2016 and US$1.3B in 2025, with a low of negative US$0.1B in 2020 and a high of US$4.0B in 2021 R (reported).¹⁷

The reason is in how a holding company is built. Brookfield consolidates assets it owns a minority of, and the outside owners’ share comes out before the line that I own. The reconciliation shows the scale: of US$3.7B consolidated net income over the last twelve months, about US$2.3B belongs to outside owners, and the company adds back US$10.7B of depreciation before taking out the outside owners’ US$11.5B share of the adjustments.³ The result is distributable earnings of US$5.65B before realizations. I can follow the arithmetic and I cannot audit it, which is why the next section’s test is about cash that is paid out.

What distributable earnings are, and what they leave out

Distributable earnings are the cash that reaches the parent from its three businesses: BAM’s earnings at the parent’s share, the insurer’s earnings, and what the operating businesses actually distribute, less the parent’s interest and costs.³ For the operating businesses that is the key point. The cash that arrives is what the affiliates pay out, US$1.55B, and not their operating funds from operations, which are a good deal higher, because the affiliates keep most of their cash to reinvest.⁴ That keeps the measure close to real money. It also means the retained profit does not appear in distributable earnings until it is paid out, and that its value lives in the asset prices.

Two things distributable earnings leave out are worth stating. Stock pay is added back: US$118M at the corporate level and US$56M at BAM over the twelve months.⁴ I treat it as a cost in Valuation. And the measure treats the insurer’s earnings as cash, though the insurer retains them to support growth, which is why the parent put US$2.76B into it over the same period.⁴

Cash earned against cash paid

The chart below is the point of this section. Net income to shareholders was US$1.43B. Distributable earnings before realizations were US$5.65B, 4.0 times as much, and with realized carry US$6.17B. The company paid owners US$1.29B, 21% of it, in dividends of US$596M and buybacks of US$695M.⁴ That is 90% of reported net income. It is also 1.3% of the market value of the company, US$98.1B on the diluted count C (calculated).⁴² The dividend alone is US$0.28 a share, a yield of 0.7%.³

The other 79% was put back. It went into the insurer, the operating businesses and funds, and the whole bet on the page is on what it earns there. A fair test is to ask whether the businesses that received it are worth more than they cost. The company’s answer is yes, by its own marks. Mine, in the next section, are lower, and that is the reason for the range.

Leverage, in three layers

The parent’s debt is modest. Corporate borrowings of US$14.7B and preferred shares of US$3.5B at market are 16% of the sum of those and the market value of the equity, and the company puts corporate debt at 13% of capitalization at the June price and 21% on book values.⁴ Distributable earnings before realizations cover corporate interest and preferred dividends 7.1 times, counting back the US$747M of interest and US$181M of preferred dividends they are struck after C (calculated).⁴ The ladder is long: US$352M due in 2027, US$1.05B in 2028, US$1.0B in 2029, US$750M in 2030, and US$11.1B in 2031 and later.⁴ The parent also issued US$600M of 5.65% notes due 2031 in September.¹⁴ Debt of US$14.3B has grown from US$5.7B in 2017 at about the same pace as assets.¹⁷

The bigger numbers are underneath. Consolidated debt to capitalization is 50%.⁴ Non-recourse debt of US$250.3B is owed against particular assets and funds. “Non-recourse” means the lender cannot reach the parent, which is a real protection. It does not stop the parent losing the equity it has in an asset, and it does not stop the parent choosing to put money in to protect a relationship or a rating. The insurer is the third layer, with liabilities of a different kind: US$167.3B of invested assets against US$19.1B of group capital.⁴ A loss of 1.15% of those assets would equal a year of the insurer’s earnings, and a loss of 11.4% would use up its capital C (calculated).⁴ Both are far from what the company reports today, and the insurer holds 33% of its portfolio in private credit, whose marks I cannot see.

The share price

Over ten years the shares rose from US$12.30 on 3 November 2016 to US$41.46, 13% a year before dividends and after the 2022 distribution of 25% of BAM, which is not added back C (calculated).² The last five years were slower: US$32.55 in November 2021 and US$41.46 now, 5% a year. The shares have been volatile. They fell 52% to March 2020, 36% from the end of 2021 to the end of 2022, and 16% from the January high of US$49.17 to now.²¹⁵ The low this year was US$35.95 on 14 October, and the shares are up 15% from it. On the last trading day before I priced this they rose 5.6%, on volume of 8.2M shares, the highest of that week, with BAM up 6.6% and the affiliates up 2% to 5%.²⁸ I do not know why, and the third-quarter results had not been reported.

VIValuation

First, how this goes badly wrong

Before any numbers, I ask what would hurt an owner most. I count three ways.

The manager’s price falls and fundraising slows. The BAM stake is 63% of the market value of Brookfield, US$61.4B of US$98.1B C (calculated).⁸²⁴ The market pays 31.6 times its annualized earnings.⁴ If private markets stop taking new money, or retirement savers do not arrive in numbers, the growth in fees goes first and the multiple after it. At 18 times, my bear, the stake is worth US$34.9B, US$26.5B or US$11.20 a share less than today, 27% of the share price C (calculated). Fundraising has been strong, so I give this one real weight, and I cannot know when it turns.

The marks are wrong. Of the company’s US$157.7B estimate of worth, US$62.0B is real estate and carried interest, which the company values itself, and US$30.5B is the insurer, at 15 times.⁴ The real estate pays 2.3% in cash on its book equity, the carried interest is realized at US$520M against a US$2.7B target, and the gross returns of its flagship property funds fell from 21% to 7% across four vintages, the newest still young.⁴ If the real estate is worth 65% of book and carried interest a third of what the company says, the shares lose about US$14 a share against the company’s own figure E (estimate). That is my bear for those parts.

The insurer’s credit goes wrong, or its spread shrinks. The insurer’s invested assets are 8.8 times its capital, and a third of them are private credit.⁴ A fall of 1.15% in their value costs a year of its earnings. Private credit is the part of the market the company’s own letter says has been questioned.¹⁰ Rates are higher than they were: the ten-year Treasury yielded 4.94% on 30 October.²² If the spread falls from 1.94% to 1.5% on the same assets, the annual earnings fall by about US$0.7B, 38% of the insurer’s total C (calculated).⁴

These three are related, because a bad credit market hits fundraising, property values and the insurer together, and I do not treat them as separate bets.

What an owner can really take out

The company paid owners 90% of IFRS net income and 21% of distributable earnings. The earnings the company defines would be worth US$2.54 a share to an owner if every dollar reached him, and the cash that did is US$0.55. That is a low payout, and not an unreasonable one for a company that says it puts money to work at 15% or better and has grown earnings 16% a year. The cash you receive while you wait does not carry the price. The growth has to, and the growth is bought with the other 79%.

So I look at the company two ways. The first is the sum of the parts, which is the standard way to value a holding company, and says what the parts would fetch. The second asks what an owner receives and when. If the parts are worth what I say, then the retained money is already in them. If they are not, then the second view is the more honest one.

The method

I value each part separately, at the 30 June 2026 balance sheet and Q4 2026 market prices, and then subtract what the parent owes. Cash earned from 1 July to now, about US$1 a share, is in neither the balance sheet nor my cash flows, and I have not added it.

BAM. The parent’s share of BAM’s earnings is US$1.94B a year after borrowing costs. I grow it for five years, then five, then for ever, discount at 10.5% (the 4.94% ten-year Treasury plus 5.5 points) and take all of it as cash.⁴²² The base gives US$47.2B, or 24.3 times, against the market’s 31.6. The growth rates are my judgement: 12% a year for five years is below the 20% growth in fee-related earnings over the last year and the 17% a year the company plans for total earnings, and above the 7% growth in earnings a share over the last twelve months.⁴²¹

Carried interest. I do not use the company’s ten times a target. I value the stream I expect to be realized: net US$0.70B in 2027, growing 8% for ten years, discounted at 12% because it is lumpy and depends on markets. That gives US$10.5B, a third of the company’s US$34.0B. As a check, my stream adds up to US$10.1B over ten years, which is more than the US$6.8B of unrealized carry accumulated net today, so it already assumes new funds pay as well as old.⁴

Direct investments. 92% of the IFRS value of US$10.6B. Listed affiliates. Today’s market prices for the 207M BIP, 309M BEP and 89M BBUC units, at 95%, a small block discount.⁸⁴ Real estate. 82% of the US$28.0B of IFRS equity. The book pays a cash yield of 2.3%, and it is revalued by the company at discount rates of 5.8% to 7.8%, so I take a fifth off.⁴

The insurer. I use 10.5 times annualized earnings of US$1.92B, which is four times the second quarter’s US$480M. This is equal to US$20.2B, or 1.06 times the US$19.1B of group capital. The multiple is the return on equity applied to the US$15.0B of capital the company invests: earnings of US$1.92B are 12.8% of it, and with 4% growth and a 10.5% required return a 12.8% return is worth about 1.35 times the capital.⁴ The bear’s 7.5 times is a 10% return, the bull’s 13.5 times is the company’s 15% target.

Debt and stock pay. I subtract corporate borrowings of US$14.7B, preferred shares at market of US$3.5B and working capital of US$0.7B, and US$2.5B for stock pay, US$174M a year capitalized at the 10.5% rate less 3.5% growth. The company adds that stock pay back to its earnings, and I do not.⁴ I divide by 2,366M shares, the company’s count with all its options and plan shares, and convert at 1.3978. The Oaktree purchase for US$3B is value-neutral in my base, since the price is about what the earnings are worth.¹⁰

The other view: the cash an owner receives

For the second view I run distributable earnings a share as a whole. They start at US$2.51, US$2.61 or US$2.74 in the bear, base and bull, after normalized carry and stock pay. They grow 6%, 12% or 17% for five years, then 4%, 8% or 10% for five, then 3%, 3.5% or 4% for ever. To grow at a rate g, the company has to keep g divided by r of its earnings, where r is the return on retained money: 15% for the first five years, 13% for the next five and 12% after. Owners get the rest. That makes the payout small at first and most of the value a terminal value. At 10.5% the cases give US$28, US$34 and US$42 a share, or C$39, C$47 and C$59.

That is lower than the sum of the parts by about 27% in the base. The gap is the real issue with Brookfield. It comes from the parts being valued at today’s market and book prices, which already include the market’s own assumption about the reinvestment of retained money, while the second view charges a fixed 10.5% for it. It also comes from the end value: the second view ends at about 10.5 times earnings, since a business growing 3.5% a year for ever and discounted at 10.5% is worth no more than that, while the market pays 17 times today. I do not know which is closer to the truth. My base uses the sum of the parts, and my margin of safety is set where the two meet.

This table is what I find most useful. The market is paying about US$38B for the insurer, the property book, the funds and all future carry together, while the company marks the same things at US$104B and I mark them at US$64B. If the BAM stake and the listed affiliates are worth what they trade for, the price implies the rest is worth 37% of the company’s figure. If BAM is worth my US$47B and not US$61B, the rest is worth about US$52B, closer to my base. So the question is not whether the stock is cheap in the abstract. It is how much of BAM’s price is real and how much of the rest is.

Three of these need only one part to be wrong. BAM at 18.7 times, the insurer at 4.8 times, or real estate at 43% of book would each explain the price on their own, and none is impossible. The fourth line is the harsh one: for the cash view to justify the price, distributable earnings a share must grow 17.7% a year for five years and 11.8% for five more. That is about the company’s own five-year plan, followed by 12% a year for five years more. At my 10.5% the base cash flows equal the price only at 9.5%. In short, the price asks little from the parts and a great deal from the retained money.

The first grid shows how much two judgements carry. Each 3 turns on BAM’s multiple moves my value by about C$3.5 a share, and each 1.5 turns on the insurer by about C$1.7. Only three cells, all in the 18-times row, fall below the price, and the lowest, with the insurer at 7.5 times, is C$54, because the real estate, the funds and the carry stay at base. The second grid is flatter, because most of the value sits in parts I value off market prices or books.

The share count matters. The circular expects about 2.52B Class A shares after the combination, including exchangeable ones, against the 2,366M I use. On 2.52B my base would be US$43.25, or C$60, instead of C$64. I cannot reconcile the two counts from the documents, so I treat the lower one as the base and the higher as the risk. The price would still be 4% below that lower value.

The margin of safety, and the price against the range

My margin-of-safety price is C$45, US$32. It is 70% of my base value. It sits at my bear of C$44 and close to the C$47 that the cash view gives, which is why I chose it. At C$45 you would pay 12.7 times the earnings an owner could take. The shares last closed below C$45 in April 2025.¹ At C$57.95 there is no margin of safety. The price is 9% below my base, 33% below my bull and 32% above my bear. My weighted value, at 25% on the bear, 50% on the base and 25% on the bull, is C$65.

The price reflects a real debate. One side says the retained money compounds at 15%, the manager’s growth is structural, the insurer is a machine for turning long liabilities into spread, and that the stock trades at 0.62 of the company’s own value per share: US$41.46 against US$66.64.⁴ That side would call my base too low. The other side says the values are the company’s own marks, the owner gets 21 cents of every dollar, the structure is hard to see through, and that the combination into New BN is the sort of move a company makes when it wants something, which is index inclusion and access to the parent’s capital.¹²⁷ That side would call my bear more likely than I do. The market could be right to pay less than the company’s marks, and a price 9% below my base is not a very large concession to that view. I give both sides some weight, and I end with a base a little above the price.

Quality, and price

Those are separate questions. The business is of high quality in the ways that can be checked: it makes real cash, grew its fee base 19%, owns assets it operates, is run by owners and has a long, cheap debt ladder. The quality of the reported value is lower, since more than half of the assets are the company’s own marks. At this price I get a good business at a price near my value, with no margin of safety yet. I do not call it cheap or dear.

The next best use of the same dollar

The ten-year Treasury paid 4.94% on 30 October and the Canadian ten-year 3.78%.²²²³ On this site, the closest holding companies are Berkshire Hathaway, whose price is 16% above my base and whose base return is 6.6% a year to 2036, and Prosus, whose price is 22% below my base and whose base return is 9.8%.¹⁹²⁰ Brookfield’s base return from today’s price is 10.5% a year to 2036 and 12.9% over five years, with a bear of 2.1% and a bull of 14.9% E (estimate). That is more than the bonds and Berkshire and a little above Prosus, and it comes with a wide range and with 79% of the earnings kept out of your hands. Brookfield’s base return is about my required return, with a price below value to start: it earns its keep, and no more.

The bull case needs many things to go right together, so I multiply them out. In the bull, BAM earnings grow 15% for five years and the stake holds a 31 times multiple (I say about 35% likely), carried interest rises to US$1.0B a year and keeps growing at 12% (30%), the insurer earns 15% on its equity (40%), and real estate is worth its book (50%). Together that is 2% if they are independent. They are not, because a strong market lifts all four, so I call it about 8% E (estimate). My 25% weight on the bull is generous on that arithmetic.

What I cannot judge

Five inputs cannot be known from outside. The first is how much carried interest will be realized, and when. The second is the quality of the insurer’s private credit, which is marked by the company and cannot be seen line by line. The third is whether the real estate appraisals will stand if rates stay high. The fourth is the fair price of the transactions between the parent, its funds and its insurer, which the parent sets on both sides. The fifth is the tax cost of the sum of the parts: the parent’s own deferred tax liability is US$0.2B, but the sum of the parts is not a plan to sell, and I have not tried to price the tax on a sale.⁴ I show ranges for each and avoid a single precise answer.

Where this could be in 5 and 10 years

For the base to be right at five years, distributable earnings a share have to be near US$4.6 and BAM’s share of earnings has to reach US$3.4B. The company’s plan is higher: US$11.5B of distributable earnings before realizations in 2031, or about US$4.9 a share on today’s count.²¹ My base value then is C$100, and the bull, which tracks the plan’s 17% growth, C$136. The company’s plan value of US$140 a share is C$196, which is 19.7 times its planned US$16.8B of total earnings on today’s count C (calculated).²¹ My bull cannot reach that, because I cap how fast value can grow.

VIIThe case for the price

What most people miss

Three things cut in Brookfield’s favour and get little attention. First, take away the two parts that trade, BAM and the three listed affiliates, and the market is paying about US$38B for the insurer, the property book, the funds and all of the future carried interest, which the company marks at US$104B C (calculated).⁴⁸² Second, the parent’s debt is smaller than the headline. Only 6% of consolidated debt has recourse to it, the term is 15 years on average, nothing is due in 2026 and the senior debt is rated A- to A3.⁴¹⁶ Third, earnings held up through the bad year for the stock. Distributable earnings before realizations went from US$1.22 a share in 2021 to US$1.59 in 2022 on the company’s restated figures, while the price fell 36%.¹⁵

The bull case

In plain words: private markets keep taking a growing share of the world’s savings, and Brookfield is one of the few firms that can place that money in real assets and also own and run them. Its insurer turns long liabilities into spread. Its funds earn fees on a base that grew 19%. Its own money compounds at 15% or better. At the 2026 Investor Day on 17 September, Mr. Flatt said the business “should compound without too much effort” at 16% from US$67 a share to US$140 in 2031, and Mr. Goodman put numbers on it: distributable earnings before realizations of just over US$11.5B, a 17% yearly growth rate, and US$16.8B in total with carry and capital allocation, which he said puts the stock at six times earnings five years out.²¹ On the Q2 call Mr. Flatt called AI infrastructure, electricity and nuclear “one of the most compelling investment opportunities we’ve seen in years.”⁶

One named holder I can document is Pershing Square Capital Management, whose Form 13F for 31 March 2026, filed 15 May, lists 59,697,208 Brookfield Corp shares worth US$2.416B, 17.6% of its reported US$13.7B portfolio and the largest line in the filing.²⁴ Its filing for the June quarter was a notice only, with no holdings table, so I cannot confirm the June position.²⁵ A 13F states a holding on one date. I read no statement by that firm on this company for this page, so I count it as a holding and not as a reason. I found no named analyst view that I could verify, so the bull case here is management’s own.

Against reality

If Brookfield hits its plan, distributable earnings before realizations of US$11.5B in 2031, the price is possible, and then some. My bull, which grows earnings at the plan’s 17%, gives C$136 in 2031 against C$57.95 now. The distance from today to the plan is the point. From US$5.65B in the last twelve months, US$11.5B in 2031 needs about 14% a year for five and a half years. The last twelve months grew 6%, from US$5.31B, and the last four years 14% a year.⁴ Fee-bearing capital, up 19%, and the insurer, up 23% in the quarter, are running ahead of that. The operating businesses went the other way: their distributions fell from US$1.69B to US$1.55B in a year, and corporate costs and interest are heavier.⁴ For total earnings of US$16.8B the climb is about 20% a year, and it leans on realized carry and on gains from selling assets. Those were US$0.52B and nothing in the last twelve months.

So the price is possible with less than the plan under my sum of the parts, which needs only my base, and it is hard to reach under the cash view, which needs about the plan itself. I put the plan at about the odds of my bull: perhaps one in twelve that all of it happens as written, because the pieces move together.

What a 10x would take. Ten times C$57.95 is C$580, a market value of about US$981B on 2,366M shares, which is ten times today’s US$98B. At the 17 times distributable earnings it trades on, that needs about US$58B of earnings a year, 9.4 times the US$6.17B of the last twelve months and 3.4 times what the 2031 plan calls for. In ten years it would take a 26% rise in the price every year. For comparison, the shares rose 13% a year in the last ten years, and distributable earnings a share about 16% a year in the last four and a half years.²¹⁵⁴ Brookfield would have to grow its fee base and its owned assets several times over, and be paid the same multiple on all of it. A tenfold rise from here is close to impossible, and I would not build anything on it.

Where I come down

What moved me was the price of the rest: if BAM and the affiliates are worth their prices, the market pays a third of the company’s marks for everything else, and the debt at the parent is modest. What did not move me was the company’s US$66.64 a share and the US$140 plan, which depend on carried interest at ten times and on property and insurance values only the company can set. What would raise my values is realized carry above US$0.7B a year, an insurer that earns 12.8% or more on its capital with a combined ratio under 100%, property sales at or above book, and a share count near 2.4B after closing. A value of C$64 is a base, not a forecast of the price.

VIIIRisks, and what would change my view

The three ways this goes badly wrong are at the start of Valuation. These are the rest.

• Private credit. The company’s own letter says much has been written about its risks, and that its book sits in real asset finance, asset-backed lending and opportunistic credit. The insurer holds 33% of its portfolio in private credit and credit is 20% of target carry.¹⁰⁴
• Real estate. The property book is US$28.0B of equity at appraisal discount rates of 5.8% to 7.8%, with loan-to-value of 45.6% in the top 34 and 51.6% on the offices. Higher rates push appraised values down, and the cash yield is 2.3%.⁴
• The combination. It needs court and regulatory approval, and the agreement lets the deal fail if holders of more than 5% of BN shares dissent.¹³⁷ The new parent will sit in a jurisdiction chosen partly for index inclusion, and its shares and its distributions will be treated differently in tax by country.⁷
• The share count. About 2.52B shares after closing against 2,366M diluted. If the higher count is right, my base falls 6%.⁷⁴
• Dilution through pay. 133M options, plan shares and exchangeable shares, about 6% of the shares outstanding, already counted.⁴
• Related-party dealing. The insurer may buy from the parent and its funds, and the funds pay fees to a manager the parent owns 74% of. I cannot check the prices.¹⁰⁴
• Control. The partners’ Class B shares elect half the board. Outside holders have no vote on that half.¹³
• Concentration in one story. AI infrastructure, power and nuclear are where management points for growth: the Kentucky campus, Westinghouse and Bloom.⁶³ If that spending slows, fundraising and development both slow with it.
• Currency. The company reports in US dollars, earns in many currencies including the pound through Just Group and the Canadian dollar through its preferred shares, and I convert my values at 1.3978.⁴⁵
• The cost of the model being copied. Other managers also buy insurers and market to individuals. If fees compress, the 57% margin does too.⁴