Sized for a World That Ended

Asset management was a fee on a rising market. Its revenue peaks by 2028.

The ship remains. The wind has gone.

I spent close to forty years on the side of the table that writes the cheque.

I will not tell you what was said in those rooms. That is not mine to tell.

But I can tell you how the rooms changed.

In the early years the managers came to us. They waited. They explained themselves. By the end, in private equity and private credit above all, it was the investor who was told what he could have, and on what terms, and by when.

I never took it for rudeness. It was something more dangerous. It was the calm of people who had not been wrong in living memory.

And we let them. The sovereign funds, the pension boards, the endowments. We accepted the terms. We paid fees on money not yet invested. We came back for the next fund because the last one had been marked up, and because everyone else was coming back too. A manager is rarely more sure of himself than his clients allow.

We mistook asset management for a growth industry. For the most part it was a fee on a rising market. The market did the work. The industry sent the invoice.

Here is what I think happens next. The industry’s real revenue reaches its peak between 2026 and 2028. By the middle of the next decade it is running at least a fifth lower. Because its costs follow slowly, that takes away about half its profit. And before 2036 at least one of the great listed names in private markets loses most of its value.

Three waves bring it about. The index fund took the stock-picker. The drought of cash is taking the buyout fund. The machine will take the analyst, and with him the reason for the fee.

Nobody can prove a forecast. I can show you what mine stands on, give the other side its best case, and tell you what would knock mine down.

* * *

Start with how the money arrives.

In 2005 the world paid its asset managers about $131 billion. In 2025 it paid them $463 billion, on Boston Consulting Group’s estimate, and that leaves out the performance fees of private equity and hedge funds.

Take a fee of half a percent on $10 billion. It brings in $50 million a year. Now let the market lift those same assets to $12 billion. The fee becomes $60 million. No new client has come. The work has not grown by a fifth.

The extra ten million arrives before anyone has decided what it is for. Some is paid out. Some buys a rival. Some becomes a promise to a new hire. Within a few years, the firm has built its budget around it.

That is the excess I mean. Costs and expectations that need the next rise in prices to justify the last one.

And the growth came mostly from prices. McKinsey counts $97.1 trillion of growth in managed assets between the end of 2015 and the middle of 2026. Markets supplied $65.2 trillion of it. New money supplied $31.9 trillion. Two dollars in every three came from prices, not from clients.

And prices were lifted by things that will not come again. A Federal Reserve economist, Michael Smolyansky, found that lower interest bills and lower taxes explain roughly two-fifths of the growth in the real profits of large American companies outside finance from 1989 to 2019. Rates cannot fall again as they fell from the early 1980s. Taxes cannot be cut for ever.

Between 2010 and 2025 the industry’s revenue grew 5.1% a year. Its costs grew 5.4%.

It did become more efficient. Between 2010 and 2022 the managers in BCG’s sample cut their costs from 20.1 to 16.4 hundredths of a percent of the assets they ran. But what clients paid fell further, from 28.3 to 22.7. On that evidence the saving went to the client, and a little more went with it.

In this business, efficiency has tended to belong to the customer.

Then came 2022, the rehearsal. One bad year, for shares and bonds together. Revenue in that sample fell 11%. Costs stayed flat. Profits fell 27%. Markets recovered, and the lesson was put away.

* * *

The first wave is twenty years old. It came in plain sight.

In 2006 the average dollar in an American fund paid 0.80% a year. In 2025 it paid 0.32%. Index funds now hold more than half of America’s long-term fund assets. Over the ten years to 2024 investors pulled more than $2 trillion out of actively managed American funds, and most managers never felt it, because rising markets put more than $7 trillion back.

Rising prices hid the withdrawals.

Look at the firms where it showed. T. Rowe Price earned an operating margin of 48.4% in 2021 and 29.9% in 2025. Franklin earned 38.1% in 2015 and 4.8% in 2024. Invesco has written more than $3 billion off the value of fund businesses it bought.

Some of those margins have since come back. T. Rowe’s margin was 38.4% again in the third quarter of 2026. In that same quarter its clients withdrew another $28.3 billion. Markets rose and costs were cut. The clients did not come back.

* * *

So the ambition moved to private equity and private credit, where fees are fixed by contract for years and no index fund waits outside the door.

Over the three years to June 2026, American private equity funds returned 7.5% a year. The same money, put into the public market on the same dates, would have earned 20.3%. The fees are still charged on the promise of the earlier record.

Cash says the same thing. In 2025 the funds handed back about 12% of their value to investors by MSCI’s count, roughly half the usual rate. Bain counts 32,000 companies sitting unsold in buyout portfolios, carried at $3.8 trillion. Not all of them are overdue. A portfolio that returns half its usual cash has a problem, whatever its manager says about the value of what remains. And it is the manager who sets that value.

Think of the pension fund that counted on cash from an earlier investment before it signed for the next. Its report shows a respectable return. Its bank account does not. So it sells something it can sell, or it waits, or it commits less. The manager’s valuation committee and the pension’s treasurer are answering different questions.

Investors have noticed. Across the world 1,502 private equity funds closed in 2021. In 2025 the provisional count was 578. The money still comes. It goes to fewer funds.

Private credit is where the confidence is loudest. Consider how the interest is being paid. The Federal Reserve Bank of Boston studied the loan books of the lending funds known as business development companies. The share of loans paying some of their interest in more debt, not in cash, rose from about 6% in early 2022 to about 10% in early 2026. Realised losses are still small.

It has not hurt yet because of the contract. A mutual fund loses its fee the day the client walks. A buyout fund goes on charging for years on money promised long ago.

Consider a hypothetical buyout firm. It has a fund of $2 billion paying it 2% a year, which is $40 million. In the sixth year the fee steps down, as such contracts commonly provide, to perhaps $20 million. The plan was that a new fund of $2.5 billion would by then be paying another $50 million. Now let the new fund close two years late and at half the size. For two years the firm lives on $20 million where it had budgeted for $70 million. After that it has $45 million. No client has complained and no loss has been declared. The invoice has simply failed to arrive.

When the institutions slow down, a new door has to open. Apollo has set a target of more than $150 billion in its wealth business by 2029. Ares has set one of about $125 billion by 2028. Blackstone speaks of a trillion. The new client is the dentist, the retired engineer, the family with savings and an adviser.

This autumn, in the funds already sold to those clients, the exits narrowed. Investors in Blackstone’s private credit fund asked to hand back 181 million shares. The fund bought 90 million. In Apollo’s equivalent, a third of what was offered was taken. The first limits came in March. In May Ares said it was keeping its target.

I allege nothing improper. The limit was printed in the prospectus. The risk is that some buyers heard the word liquidity more clearly than the word limited.

The door is being held open by governments. In August 2025 the American president signed an order to bring private assets into the retirement accounts of ordinary workers. Britain and Europe are moving the same way. The industry’s lobby in Washington called the order “a great step” the day after it was signed.

For most of my career, when a private fund disappointed, it was a matter between professionals. What is being built now is different. The state has become the industry’s salesman, and the buyer is the voter. American workplace savings plans alone hold $15 trillion. Congress’s own auditors reported this month that nobody can say how much private equity 401(k) retirement plans already own. They also found that the research does not agree on whether it beats the stock market.

When the people writing the rules become too close to the people collecting the fees, I call it cronyism.

* * *

The third wave has not broken yet. It is the one I would fear most if I still had a business plan to defend.

The fee has always rested on one justification, and that was scarcity. To understand a company, to read a thousand loan agreements, to build the model and write the report, you needed rooms full of trained and expensive people. The client could not do it himself. So he paid.

That scarcity is ending. McKinsey reckons artificial intelligence could do work equal to between 25% and 40% of an average asset manager’s costs.

The industry reads that as good news. Lower costs, fatter margins. I read it the other way. In this business efficiency has tended to belong to the customer. When the cost of producing analysis falls, the price of analysis follows it down. A few houses with their own data, their own distribution, or capital that cannot leave may keep the saving. Most will hand it over.

AI also makes it easier for clients to do the work themselves. The chief obstacle to a pension fund doing its own work was the size of the team it would need. That obstacle shrinks with every new release. British Columbia’s public fund already holds 49% of its private equity through direct investments. CalPERS has modelled what it saves by investing alongside its managers and not through them. On every billion dollars it would keep about $400 million in fees and profit share over the life of the investments, before its own costs. A fund that has built its own team does not take it apart because a former manager misses the business.

The index fund needed twenty years to halve the price of a fund. I do not think the machine will be so patient.

* * *

There is a strong case against this forecast.

New money is real. Net flows were 3.2% of assets in 2025, and McKinsey reckons the first half of 2026 ran closer to 5% a year. Pension assets across the OECD countries reached $78 trillion. Lending has moved out of the banks, and America’s private credit vehicles hold about $1.9 trillion. The strong are thriving: fee earnings rose 27% at KKR and 23% at Apollo. Over twenty years private equity is still ahead of the public market. American payrolls in the trade are rising, and nearly all the firms surveyed say the machines have barely changed their headcount. Bonds pay an income again. And when Wall Street lost its fixed commissions in 1975, volume did the rest. Lower prices, more business.

I accept every one of those facts. I do not accept what is built on them.

Six months of strong flows is not ten years of them. Money that moves from shares into bonds and cash pays the manager a thinner fee. KKR and Apollo are the winners of a concentration, and their growth tells you who is winning, not how large the prize is. Payrolls follow revenue. They do not lead it. As for 1975, it worked because trading boomed. The question is what booms now.

* * *

Revenue is assets multiplied by price. Suppose the price goes on eroding at 2.5% a year, about half the pace American funds have seen since 2006, and inflation runs at 2.5% as well. These are my assumptions. They are nobody’s data.

If markets give 3% a year and new money arrives at the 2025 rate, real fee revenue is about 12% higher in ten years. The industry is fine and I am wrong. If markets give nothing, real revenue falls about 17%. If the flows slow to 2% as well, it falls 26%.

I am betting that returns will disappoint over the next ten years, and that disappointing returns will make fresh money harder to win.

Poor returns do not need ten motionless years. A portfolio that falls by a quarter and then earns 3% a year for nine years ends slightly below where it began. The industry can cheer every recovery and still never recover the income its old budgets assumed.

I did not live through the last long winter as a practitioner. I inherited it. Later, when I sat on the investment and asset allocation committees of large sovereign funds, the elders in the room still carried the 1970s with them. One of them told me the story more than once. Nazem, he would say, in the seventies nobody talked about equities as the main asset, the one you overweight. They were a minority holding. The climate was too hostile. Yields were high, bonds paid, and owning shares was something close to a taboo.

The figures bear him out. American stock funds held $56 billion in 1972. Over the next two years they lost 44% of it. In seven of the eight years to 1979, investors took out more than they put in. The old peak was not seen again until 1983.

Those of us raised in the long summer took it for the climate. It was a season.

Returns will be poor because of the price already paid. Robert Shiller’s measure sets share prices against ten years of earnings. For American shares it has stood at about 40 since the spring of 2026. Before this year the only readings that high came at the top of the technology boom. In 1929 it was 33. In 1982, when the tailwind began to blow, it was about 7. The companies in the S&P 500 earn about 3.9% a year on the price of their shares. A ten-year loan to the American government pays about 5%. It is a rough gauge and not a prophecy. But the earnings behind a share now yield less than a loan to the Treasury, and the shareholder carries more risk.

I am outside the consensus. Vanguard expects American shares to earn between 4.8% and 6.8% a year. BCG expects the industry’s revenue to reach $657 billion by 2030.

Here is the forecast in a form that can be checked. Take the global revenue of asset management as BCG measures it, and adjust it for American inflation. That figure will reach its peak between 2026 and 2028. Its average for 2034 to 2036 will be at least 20% below its average for 2026 to 2028.

Twenty percent does not sound like an ending. Picture a firm with revenue of 100 and costs of 70. Profit is 30. Take a fifth off the revenue and leave the costs alone. Profit is 10. Cut the costs by a tenth, which in this trade means people, and profit is 17.

A defender will say the machine cuts the costs as well, and he is right. Take a quarter off the costs and profit hardly moves. So the whole question is which falls first, the price or the cost.

I say the price. A client can move to a cheaper product in a day, and a big one can reopen a fee in a single meeting. Costs are leases, contracts, deferred pay and people, and they leave over years. The record is consistent with that. From 2010 to 2022 the price fell further than the cost. In 2022 revenue fell 11% and costs did not move at all.

Profit halves if costs fall by no more than about 7% while revenue falls by a fifth. That is my judgment. It is the part of this forecast I hold least firmly, and the part I would most like to be shown wrong on.

And that sum counts only the running costs. There is borrowed money in this story too. It does not sit with the traditional houses, which owe little. Business development companies borrow about a dollar for every dollar their investors put in. That is the funds’ debt and not the managers’ own, but it is the funds that pay the fees. Some of the wealth firms being rolled up into chains are bought with debt. Behind them all stand the banks. American banks have lent $1.7 trillion to financial firms outside banking, a wide category. More than $800 billion of it is drawn by private equity and private credit vehicles and their kind.

A fall in revenue meets the fixed costs first. Then it meets the debt.

There will be markers on the way. Before 2036 one of the seven large listed managers of private assets, Blackstone, KKR, Apollo, Ares, Brookfield Asset Management, TPG or Carlyle, will lose at least 60% of its market value from its peak. I cannot say which. Blackstone lost nearly 90% between its listing and early 2009. Several of these firms are no longer fee businesses alone. At Apollo, in the third quarter of 2026, earnings from its insurance arm exceeded its earnings from fees. That ties them more tightly to the credit cycle, and it lies outside the revenue I am scoring. The average American fund, which costs 0.32% today on Morningstar’s measure, will cost 0.25% or less on the same measure by 2030. And by 2030 at least one major country will place a new and material restriction on retail private funds, beyond the rules in force today.

And this would prove me wrong. If real revenue in any year from 2029 to 2036 is higher than the peak of 2026 to 2028, the date is dead. If the average for 2034 to 2036 comes in above four-fifths of the baseline, I missed. And if by 2030 costs are falling faster than fees and margins are widening, the engine of my argument has failed, whatever the calendar says.

* * *

Not everything shrinks. The cheap index fund survives. It is the thing doing the damage. The rare manager with a true skill in something scarce survives, and can charge for it. What does not survive is the wide middle. The firm that is neither cheap nor rare, and the teams it hired for a rise that will not come.

And the client? The client mostly wins. A lower fee is a pay rise for every pensioner who never asked for one. The world loses little by it. One industry faces a hard decade, and so do the people who planned their lives around its weather.

I think about those rooms still.

Now I would ask for two budgets. One for the world in the presentation. One for the years when the markets do not come to the rescue.

How I will keep score. I will use BCG’s measure of global asset-management revenue, which leaves out the performance fees of private equity and hedge funds, and convert each year’s dollar figure to constant purchasing power with the American consumer price index. The peak is the highest real figure in 2026 to 2028, with nothing higher from 2029 to 2036. The contraction is the average of 2034 to 2036 set against four-fifths of the average of 2026 to 2028. The two tests stand separately, and only revenue is scored. The halving of profit is my arithmetic on a firm with a 30% margin, not a measured test. If BCG restates its figures I will apply the restatement to both periods, and if the series can no longer be compared I will say the test cannot be scored. A fall caused by currencies alone would not prove my explanation. For the listed-manager marker I will count a share price that closes at least 60% below a peak set after this essay is published, for twelve month-ends in a row. The retail marker means a rule in America or Britain, in force by the end of 2030, that narrows who may buy these funds, how much they may hold, or how they may withdraw.

Sources

Figures from consultancies and data providers are estimates. Company figures come from the companies’ own filings. The forecasts, the arithmetic and the judgments are mine.

Industry revenue, cost and revenue growth, and the 2030 projection. Boston Consulting Group, Global Asset Management 2026 (April 2026).

Costs and prices per unit of assets, 2010 to 2022; the 2022 fall in revenue and profit. Boston Consulting Group, Global Asset Management 2023 (May 2023), Appendix 1.

Growth of managed assets from markets and from flows; net flows in 2025 and 2026. McKinsey, Asset Management 2026 (12 October 2026).

Interest, taxes and profit growth, 1989 to 2019. Michael Smolyansky, Federal Reserve Board, FEDS 2023-041 (June 2023).

American fund fees, 2006 and 2025. Morningstar, US Fund Fee Study (19 May 2026).

Index funds’ share of American fund assets. Investment Company Institute, Active and Index Investing, September 2026.

Outflows from active funds and market appreciation, ten years to 2024. Morningstar, Active Management Isn’t Going Away Anytime Soon (March 2025).

T. Rowe Price margins, 2021 and 2025. T. Rowe Price, 2025 Form 10-K and 2021 results release.

T. Rowe Price margin and client flows, third quarter of 2026. T. Rowe Price earnings release (29 October 2026).

Franklin margins, 2015 and 2024. Franklin Resources, fiscal 2026 Form 10-K, which restates fiscal 2024; fiscal 2015 results release.

Invesco write-downs. Invesco, 2023 and 2025 Forms 10-K.

Private equity returns against the public market. Cambridge Associates, US Private Equity benchmarks, second quarter 2026.

Private equity distributions in 2025. MSCI, The distribution drought in private equity persists (11 June 2026).

Unsold companies in buyout portfolios. Bain & Company, Global Private Equity Report 2026, release of 23 February 2026.

Private equity funds closed, 2021 and 2025. PitchBook, 2025 Annual Global Private Market Fundraising Report.

Loans paying interest in kind. Federal Reserve Bank of Boston, Early Warning Signals in Private Credit? (5 August 2026).

Apollo’s wealth target. Apollo Global Management, Investor Day (1 October 2024).

Ares’s wealth target. Ares Management, analyst presentation (September 2025).

Blackstone’s wealth ambition. Forbes (17 February 2026).

Blackstone Private Credit Fund, third-quarter 2026 tender result. SEC filing (2 November 2026).

Apollo Debt Solutions, third-quarter 2026 tender result. SEC filing (5 November 2026).

First limits on withdrawals, March 2026. MarketScreener, Ares private credit fund limits investor withdrawals (24 March 2026).

Ares keeping its target, May 2026. Bloomberg Law (1 May 2026).

The executive order on retirement accounts. The White House (7 August 2025).

The industry lobby’s response. American Investment Council (8 August 2025).

Assets in American workplace plans. Investment Company Institute, retirement statistics, second quarter 2026.

Private equity in 401(k) plans. US Government Accountability Office, GAO-27-106779 (released 10 November 2026).

Artificial intelligence and managers’ costs. McKinsey, How AI could reshape the economics of the asset management industry (16 July 2025).

British Columbia’s direct private equity. BCI, performance overview, fiscal 2026.

CalPERS on co-investment savings. CalPERS, private equity review (June 2026).

Pension assets in OECD countries. OECD, Pensions Outlook 2026.

Private credit vehicles and their borrowing. Federal Reserve, US Private Credit Vehicles (9 November 2026).

KKR fee earnings. KKR, third-quarter 2026 results.

Apollo fee earnings and insurance earnings. Apollo, third-quarter 2026 results.

American payrolls in portfolio management and investment advice. US Bureau of Labor Statistics, Table B-1a.

Firms’ headcount and artificial intelligence. EY, GenAI in Wealth and Asset Management Survey (16 September 2025).

Fees on equity funds and money market funds. Investment Company Institute (25 March 2026).

The end of fixed commissions, 1975. US Securities and Exchange Commission, News Digest (6 June 1977).

American stock funds in the 1970s. Barry Barbash, SEC, Remembering the Past (4 December 1997).

The Shiller price-earnings measure. Robert Shiller’s data, as published by multpl.com.

Earnings yield on the S&P 500. multpl.com.

Ten-year Treasury yield. Federal Reserve Bank of St. Louis, FRED.

Vanguard’s return forecast. Vanguard, market outlook return forecasts.

Debt-funded wealth firms. WealthManagement.com, on Mariner and Moody’s (22 October 2026).

Bank lending to financial firms outside banking. Federal Reserve Bank of Boston (4 November 2026).

Blackstone’s share price, 2007 to 2009. Blackstone, 2010 Form 10-K; ABC News (10 March 2009).