How Quant Firms Split Profit Between the House, the Pod, and the Individual
Citadel's funds produced $56.8bn in gains, $17bn went to employees, and the arithmetic in between decides what you earn...
Citadel’s three largest funds started 2021 with $23.6bn and produced $56.8bn of gains by the end of September 2024. Investors kept $30bn. The firm collected $7.5bn in management and performance fees. Another $17bn left the funds as pass-through expenses, close to 90% of which was employee compensation. Those figures only became public because a $1bn bond offering forced Griffin’s firm to publish a prospectus, and they remain the most detailed public accounting of how a multi-manager platform divides what it earns.
For daily Quant articles and market breakdowns:
Most descriptions of that division are wrong. The common version is 2 and 20 with a slice carved out for the trader, which describes almost nothing about how the money actually moves. Three separate claims sit on the same P&L. Each one is priced differently, each carries different risk, and each is funded from a different pocket.
Which claim you hold determines your entire compensation profile. It explains why a quant generating identical alpha at Millennium and at Jane Street ends up with completely different outcomes, and why the same payout percentage can be worth double at one firm compared with another.
Three claims on one P&L
The house is the general partner. Israel Englander at Millennium, Ken Griffin at Citadel, Dmitry Balyasny, and Steve Cohen at Point72 hold a claim on aggregate fund performance plus the residual economics of the management company itself. Their revenue is a function of assets under management and firm-level net return.
The pod holds a contractual claim on its own P&L, expressed as a payout percentage. That claim survives a bad year at the firm. A pod up 8% inside a fund that finished flat still gets paid in full.
The individual below portfolio manager level holds no contractual claim at all. Analysts and quant researchers receive an allocation from a pool the PM controls, sized at the PM’s discretion.
The mechanic that matters sits in how the first two claims are funded. Pod payouts are not carved out of the 20% performance fee. They are charged to limited partners as an expense, which changes the economics for everybody involved.
What the house actually charges
Headline management fees at the large platforms bear almost no relationship to what investors end up paying.
Citadel wraps pass-through costs into a management fee that has floated between 3% and 6%, with investors effectively reimbursing the full cost of running the fund. Millennium moved to a structure charging roughly 1% of assets or 20% of gains, whichever is larger, which guarantees the firm several hundred million dollars in a flat year. Point72 has stated in regulatory filings that no limit applies to the pass-through expenses it can charge.
A review of filings across five platforms managing more than $200bn found the disclosed lists of eligible pass-through items expanded sharply over the past decade. Recent categories include artificial intelligence spend, compliance, internal referral payments, and the cost of terminating staff.
The aggregate effect is measurable. Multi-strategy funds retained an average of 59% of every dollar they generated in 2023, up from 46% two years earlier.
The house has engineered its revenue stream to be largely independent of performance. Investors carry the cost base whether the year is strong or weak, which is why platform economics get valued like an operating business.
For full access to daily quant articles (12.5$ a month):
The pod’s contractual number
Payout ratios across the platforms cluster in a 15% to 20% band on pod P&L for a senior seat. Established PMs recruited through a competitive process negotiate higher. 25% to 30% appears at newer platforms and at firms trying to buy a track record they cannot build internally.
The number climbed for a structural reason worth understanding properly. Under a pass-through model, the platform does not net losing pods against winning pods before paying bonuses. A PM who makes $80m gets paid even if the fund overall lost money, because that bonus is charged to investors as an expense.
Netting risk transferred from the manager to the limited partner. Once the manager stopped bearing that cost, it could afford to bid the payout ratio higher without touching its own margin.
The people funding the talent war are the pension funds, endowments, and sovereign wealth allocators writing cheques to these platforms. That is the reason fee pushback has intensified even in years when the platforms outperformed.
The arithmetic that actually lands
Take a senior equities seat running $1bn of capital that generates $100m in trading P&L for the year. The money passes through two distinct cost layers before any of it reaches a person.
Top-line costs come off first. Financing, stock borrow, commissions, exchange fees, market data, and capital charges are deducted from gross P&L before the payout percentage applies. Assume $5m, which leaves $95m as the pool base.
The contractual payout applies to that figure. At 20%, the pod’s total compensation pool is $19m.
Bottom-line costs then come out of the pool itself. Team base salaries, analyst bonuses, dedicated technology spend, and any support headcount the PM negotiated onto the book all sit here. Assume a team of four with $600k in combined base salaries and $4m distributed to analysts. The PM’s personal number lands near $14m before tax.
Base salary is almost always a draw against that pool rather than an addition to it. Salaries at the platforms cap around $150k to $175k even for portfolio managers, because the firm has no interest in paying fixed money for a variable output.
Whether a cost sits above or below the line is worth more than two or three points of payout percentage on most books. A PM at 18% with financing charged top-line frequently out-earns a PM at 22% who absorbs it.
Inside the pod, discretion replaces contract
The PM receives a pool and allocates it. No formula governs what an analyst receives. No analyst holds a contractual entitlement to a share of P&L in the way the PM does.
Attribution inside a pod is a judgment call the PM makes about who generated which idea, and that judgment is rarely written down anywhere.
A strong analyst in year two or three on a profitable pod at a top platform realistically clears $400k to $700k. The band is wide because it depends on book size and on the PM’s read of contribution. In a weak year, some PMs smooth junior bonuses upward out of their own take to hold a team together, which is a retention decision made by an individual rather than a firm policy.
Below that level, compensation compresses hard. Junior research and support seats land in a $100k to $150k range split roughly evenly between base and bonus, with P&L linkage that is almost entirely indirect.
A junior at a platform carries exposure to two performance outcomes at once. The pod has to make money, and the PM has to be willing to share it.
Autonomai Recruitment places engineers and quantitative researchers into trading technology seats across London and New York. Their recruiters cover defined lanes of the stack, from low-latency C++ and FPGA work through to systematic research and infrastructure. If you want visibility on live mandates before they hit job boards, that is where to start.
The structural point behind those mandates is that a platform seat and a prop seat pay for different things, which becomes obvious once you look at how the firms without external capital run their compensation.
The prop shop model inverts the structure
Jane Street runs no external capital, no pods, and no individual P&L attribution. Every employee is paid from firm-wide profit.
The firm booked $39.6bn in net trading revenue in 2025 against roughly 3,500 employees, producing $31.2bn in EBITDA. That works out to close to $9m of profit per head. Applying the firm’s historical compensation ratio near 20% of revenue implies average compensation around $2.3m across the entire staff. The first quarter of 2026 delivered $16.1bn in trading revenue and $10.3bn in net income, more than double the same quarter a year earlier.
Hudson River Trading produced roughly $12.3bn in 2025 revenue with around 900 employees. Citadel Securities produced roughly $12.2bn. New graduate packages at these firms run $300k to $700k, and IMC converted its 2024 intern class at approximately $425k total compensation.
If you are working out which seat structure suits your background, QuantFinanceWiki.com maps the paths properly. The platform covers firm profiles across the multi-managers and prop shops, role roadmaps by function, and a large interview question bank organised by firm and by round.
The reason that mapping matters is that turnover across these platforms is high enough that the seat you accept is rarely the seat you retire from.
The trade is explicit. You surrender the right to capture your own P&L, and in exchange you stop carrying single-book termination risk.
A pod PM down 7.5% at Millennium is terminated by rule. A trader on a firm-wide profit share who has a flat year at a firm that printed still gets paid a serious number.
Neither structure is superior. They price different things. Pod economics pay for isolated, measurable alpha and charge you the full cost of failing to produce it. Prop economics pay for durable contribution to a shared engine and charge you the upside of your best year.
Systematic pods sit somewhere between the two
Quantitative pods at the multi-managers complicate the picture. A systematic book carries higher fixed costs in data, compute, and research headcount, and those costs are frequently charged bottom-line against the pod’s own pool.
