Written by GPT-5.6 Sol under Leo's direction. Human-directed Workbench essay, 1 October 2026.
We got here by asking why somebody earning serious money in finance would move to Dubai.
If you're already a portfolio manager at a big multi-manager fund, the answer can get pretty clean. Same markets, same firm, same book, much friendlier tax treatment. Great.
Then one question ruins the clean little story:
How many people ever become that person?
The $5 million portfolio manager does an enormous amount of rhetorical work whenever people talk about finance careers. He is the equivalent of describing acting by pointing at Leonardo DiCaprio, except the route to PM often begins with a job that is itself considered elite and already filters hard.
You get the target school, the internship, the bank, the right group, the right exits, the buy-side seat, the track record, the capital allocation, and then maybe one day the compensation graph goes completely insane.
A lot of people disappear somewhere along the way.
So the career question gets more interesting once you stop asking, "How much can finance pay?" and start asking:
What is the expected bargain for a very capable person who could have spent the same decade becoming excellent at something else?
The $5 million person is doing too much work
Junior banking and senior hedge-fund investing barely feel like the same career.
A young investment banker can spend a huge amount of time executing somebody else's transactions: models, decks, diligence, process management, comments, revisions, client requests, another turn of the deck at 1:00 in the morning.
The hours are real. In eFinancialCareers' 2026 compensation and lifestyle survey, based on more than 2,700 finance professionals, M&A respondents reported 67.3 hours a week on average. Finance-technology respondents reported 43.8. (eFinancialCareers)
Then people talk about the payoff and quietly jump ten years forward to a successful hedge-fund PM whose compensation is linked to P&L.
Those jobs have a completely different economic logic. The 2026 eFinancialCareers data puts hedge-fund analysts around $230,000 in average compensation, more senior staff around $840,000, and PM/MD-level respondents above $2 million on average, with enormous dispersion around all of those numbers. (eFinancialCareers)
Great prize.
Very different seat.
The analyst who spent two years rebuilding merger models did not receive a coupon promising eventual admission to the $2 million club.
The feeder job gives you training, credentials, a network, credibility, and access to the next contest.
Then the next contest begins.
The funnel keeps asking for a new proof
Elite careers often filter repeatedly. Finance makes the sequence unusually visible because the thing being entrusted to you keeps getting more expensive.
At the front end, major hedge funds such as Millennium, Citadel, and Point72 routinely accept less than 1 percent of applicants to their graduate programs. (eFinancialCareers)
First somebody trusts you with grunt work.
Then analysis.
Then recommendations.
Then a piece of a portfolio.
Then meaningful risk.
Eventually, if everything goes well, somebody may trust you with hundreds of millions or billions of dollars and pay you according to what happens.
Each step asks for a different proof.
A current Dubai discretionary-macro PM listing asks for at least five years of portfolio-management experience, a demonstrated Sharpe ratio above 1.5, a detailed investment track record, and a proposed strategy and process. (eFinancialCareers)
Caxton's official Associate Portfolio Manager program gives a nicer view of the intermediate stage. The firm says roughly a quarter of its current PMs came through the program, which recruits promising investors, sell-side talent, and quantitative analysts and tries to turn them into people who can actually run capital. The program operates in Dubai, London, New York, and Singapore. (Caxton)
So even after reaching an elite bank or fund, the most lucrative endpoint remains another selection problem.
You have to become the person whose judgment somebody is willing to express in dollars of risk.
For somebody who loves the game, it's a remarkable career.
It is a pretty strange lottery ticket if the main attraction is hearing that somebody at the end can make $5 million.
What do you get to keep?
The compensation question started bothering me less than the human-capital question.
Suppose you spend ten years becoming excellent at something.
What lives in your hands and head afterward?
A software engineer can make software. They can automate an annoying task, build a product, contribute to an open-source project, start a company, write a game, understand why the computer is doing something stupid.
A carpenter can look at wood and make a table.
A mechanical engineer can reason about machines.
A statistician can look at a dataset and investigate what happened.
A physician acquires a giant body of practical knowledge about bodies, disease, diagnosis, treatment, and care.
A mathematician gets a language for abstraction and proof that keeps appearing in places nobody warned them about.
Even philosophy, which society loves using as the mascot for impractical education, gives you a few thousand years of arguments about knowledge, ethics, logic, language, mind, politics, aesthetics, and what the hell anybody thinks they are doing here. You can read it, argue with it, write into it, teach it, carry it around forever.
Finance absolutely gives you real skills too.
Accounting. Valuation. Probability. Risk. Incentives. Negotiation. Business models. Capital allocation. Contracts. Market behavior. Reading management. Seeing where the cash actually goes. Learning to ask what assumptions are hiding inside a number.
A good investor can become extraordinarily perceptive about companies and people.
But the highest-value edge can become very attached to the game in which you learned it.
Imagine an excellent long-short equities analyst covering European industrials. They know management teams, suppliers, accounting quirks, consensus expectations, broker models, channel checks, historical multiples, position sizing, catalysts, earnings revisions, and which sentence on a call was quietly terrifying.
Inside the right fund, this person may be worth millions.
Now take away the fund's capital, execution systems, data, management access, broker relationships, colleagues, risk budget, and the ability to express the view at institutional scale.
A lot of valuable knowledge remains. The person can understand businesses beautifully. They can manage their own money well. They may be an excellent operator, adviser, or entrepreneur.
The part worth millions can become much more context-dependent.
I keep coming back to a stupid little test:
What can you do with the skill on a Sunday afternoon?
The engineer can build.
The carpenter can make something.
The statistician can investigate.
The writer can write.
The finance person can absolutely read a company, model it, think about a deal, or make an investment decision. Yet the rare edge that justified giant institutional compensation may require an institution around it before it becomes economically gigantic again.
This is where the career starts feeling a little insular.
Finance is unusually second-order
Finance sits one level above a huge amount of ordinary productive activity.
Somebody designs a chip.
Somebody discovers a drug.
Somebody builds an apartment tower.
Somebody runs a grocery chain.
Somebody writes enterprise software.
Finance decides what the company is worth, who supplies the capital, how much the capital costs, which risks get transferred, whether the acquisition happens, who owns the claim, and which project gets funded next.
Those decisions have enormous real-world consequences. Capital allocation can determine which factory gets built and which company dies.
The intellectual activity is often about somebody else's activity, though.
Public markets add another layer. A good investor asks more than whether a company is good.
What does everybody else believe?
Which assumptions already sit inside the price?
Where is consensus wrong?
Which future event changes the belief distribution?
How crowded is the position?
What will other investors do when the number prints?
You become good at businesses, and then good at beliefs about businesses, and then good at beliefs about other people's beliefs about businesses.
It can be fascinating.
It can also become a beautifully elaborate game whose deepest expertise makes the most sense to people already playing it.
Money becomes both product and scoreboard
Engineering usually leaves an object behind.
The bridge stands.
The program runs.
The machine works.
Science leaves a result somebody else can test.
Medicine leaves a patient who is alive, healthier, diagnosed, treated, relieved, or at least better understood.
A portfolio manager leaves a return series.
The return series is real. It may represent extraordinary judgment under uncertainty. It may be one of the cleanest scoreboards any intellectual profession gets.
It also means the profession's output, status system, and compensation can all collapse toward the same unit.
You made money.
Therefore you were good.
Because you were good, you get more capital.
With more capital, a small edge can make more money.
Because you made more money, your compensation gets larger.
Pretty soon the money is the product, the score, the proof, and the reward.
No wonder the culture can feel self-contained from outside.
The person inside may be doing incredibly difficult work. They can spend all day synthesizing economics, company behavior, politics, psychology, statistics, market microstructure, and risk.
Then somebody asks what they made.
"Eight percent above benchmark with a 1.7 Sharpe."
Inside the profession, the sentence can carry enormous meaning.
Outside it, the most legible translation is still:
"I made a lot of money with money."
There are expansive versions of the career
A biotech investor can learn a shocking amount of biology.
A semiconductor analyst can end up understanding fabrication, packaging, capital equipment, supply chains, architecture, and the tiny collection of companies that keep the whole industry alive.
A commodities trader can develop an intimate knowledge of shipping, storage, refineries, harvests, mines, pipelines, and physical bottlenecks.
A quant can build serious mathematics, statistics, programming, optimization, and experimental habits.
A credit investor can become fluent in contracts, corporate behavior, distress, and what happens when cheerful assumptions meet a maturity wall.
Private equity can teach a person how companies are bought, financed, governed, and sometimes improved.
Finance can become a passport into dozens of other domains because money touches all of them.
The broader mind comes from caring about the underlying world, though.
A person can use finance as an excuse to learn everything.
A person can also become unbelievably good at a narrow internal move: building the model, turning the deck, parsing the covenant, timing the catalyst, optimizing a particular signal, surviving the process.
Because I Fucking Wanted It makes a related point about careers and craft: desire decides which distinctions you keep paying to learn. You Will Find Another Cursed Thing to Care About argues that obsession itself can travel when a specific technical craft gets automated.
Finance rewards obsession too.
The object of obsession simply deserves inspection.
The career bet
Any profession with a magnificent right tail can hypnotize ambitious young people.
Look at the partner.
Look at the founder.
Look at the surgeon with the private practice.
Look at the principal engineer.
Look at the hedge-fund PM.
Look at the person who survived every filter and then got very, very good.
Now run the tape from the beginning.
How many years did it take?
How many people wanted the same seat?
How much control did you have over advancement?
How much of your life did the apprenticeship consume?
How valuable were the intermediate outcomes?
What could you do if you left halfway through?
What kind of knowledge accumulated even if the jackpot never arrived?
A career is also a machine for turning years into a mind.
Finance can produce a formidable one: numerate, commercially sharp, probabilistic, alert to incentives, comfortable with uncertainty, and capable of making decisions while the evidence stays incomplete.
It can also train somebody exquisitely for a game whose highest rewards exist inside the game.
So I like one question more than "Which career pays the most?"
Imagine the compensation equalized for a minute.
After ten years, which mind would you want to have?
Would you want to know how to build software?
How to design machines?
How to practice medicine?
How to prove things?
How to understand organisms?
How to write?
How to reason about markets, businesses, risk, and capital?
Pick the mind first.
Then put the money back in.
If markets themselves fascinate you, finance can be a fantastic answer. The work gives you a living scoreboard, endless uncertainty, real stakes, and a world that refuses to stay solved.
If the jackpot is the fascination, the funnel deserves much more attention than the guy standing at the end of it.