Hedge fund, prop trading firm or market maker: how the employer types really differ

Most comparisons of these employers collapse into stereotypes: prop is fast and flat, funds are slow and hierarchical. That gets people into processes they do not want. The useful approach is to treat the differences as six separate axes and check where a specific firm sits on each one, because the labels do not travel together.

Axis one: whose capital is at risk

Two different questions hide inside the word "capital", and conflating them is the most common error in this comparison.

The first is whose money is in the trades. An investment manager deploys capital raised from outside investors under a mandate: there are investor reports, allocation decisions and redemption windows, and the mandate constrains what the strategy may hold. A proprietary trading firm trades its own balance sheet, so there is no client position to report and no redemption queue, and the definition of an acceptable drawdown is set internally.

The second is who owns the business. Trading on your own balance sheet does not mean the firm has no outside shareholders and does not raise capital. Citadel Securities announced on 11 January 2022 that Sequoia and Paradigm had made a 1.15 billion dollar minority investment in the firm (Citadel Securities, announcement, 11 January 2022). Principal trading firms also borrow, hold external credit facilities and take strategic investors. What distinguishes them from a fund is that outside investors own a stake in the business rather than the positions.

Citadel describes itself as an alternative investment manager working across major asset classes and geographies through five strategies: Commodities, Credit and Convertibles, Equities, Fixed Income and Macro, and Global Quantitative Strategies (Citadel, What We Do). Citadel Securities states that Citadel and Citadel Securities are separate and distinct firms focused on different aspects of the markets (Citadel Securities, Who We Are). Two entities, one founder, different businesses. If you are comparing offers, compare the entity, not the brand.

Institutional proprietary trading is not a retail funded-account product. The firms discussed here hire employees, pay salaries and put firm capital at risk. Retail funded-account or evaluation-challenge offerings ask an individual to pay a fee to attempt a simulated target. That is a different commercial arrangement, and it is not what an institutional prop seat is.

Axis two: what the firm does in the market

Market making is an activity, not a third capital structure. A firm quotes two-sided prices and earns the spread between bid and ask in exchange for carrying inventory risk; Citadel Securities describes exactly this business and notes that competitive spreads are often a small fraction of one per cent of a security's price, so the economics depend on pricing quickly at scale (Citadel Securities, What Is a Market Maker?).

The activity can sit alongside others in the same firm. Plenty of principal trading firms both quote and run directional or relative-value strategies, and a fund can take liquidity without ever providing it. Obligations vary too: a designated market maker on a particular venue or in a particular programme accepts quoting commitments defined by that venue and its regulator, while a firm quoting voluntarily in other markets does not carry the same duty. Ask which venues and which programmes, rather than assuming a common obligation.

Directional systematic trading is a different exercise again: the firm takes a view, holds it, and is measured on whether it was right. Multi-strategy managers run several of these side by side and allocate capital between them.

Axis three: holding period

Holding period shapes how quickly a trading result is known. It does not, on its own, determine how fast engineering ships or how engineers are assessed; a long-horizon manager can deploy daily and a low-latency firm can spend a quarter on one system.

A single firm can occupy several rows. Hudson River Trading says its trading teams are organised roughly by the time horizon at which they trade, then by asset class and region (Hudson River Trading, Engineering and Interviewing at HRT). Ask which horizon the seat you are interviewing for actually sits on.

Axis four: how teams are organised

This is the axis candidates most often get wrong, because the stereotype is weakest here.

Some managers run a pod model: independent teams with their own capital allocation and, often, information barriers between them. Others explicitly do not. HRT states that its trading group is not a pod system with siloed teams and IP barriers, and says this matters for engineers because they can work across all trading at the firm (HRT).

Scale cuts differently again. Jane Street describes a technology group that is small by design, with porous team boundaries (Jane Street, Technology at Jane Street). That is one firm's organisational choice, not a rule about prop firms.

The practical question is not "pod or not" in the abstract. It is: who decides what I work on next quarter, and who sees the result.

Axis five: what engineering is pointed at

Engineering is a first-class function at all three employer types, but the centre of gravity differs. HRT splits its engineering between the work needed to trade and the work needed to run research at scale, and characterises the difference as latency against scale (HRT). A multi-strategy manager with a long-horizon commodities book looks different again: Citadel describes atmospheric scientists working with high-performance computing specialists to turn very large daily data volumes into forecasts (Citadel, What We Do).

The useful read across all three: ask what the engineering organisation is measured on. Uptime and latency, researcher throughput, or product adoption inside the firm are three different jobs, and firms of every label run all three.

Axis six: how you are paid

Compensation mechanism is a question to ask, not to infer from the employer label. Establish how the bonus pool is determined, what it is measured against, how much is deferred and over what period, and what happens to deferred amounts if you leave. Treat any single advertised range as one observation.

A decision matrix you can actually use

Questions worth asking in every process

These are prompts to use in a conversation, not findings about the sector. The answers vary by firm and by team.

• Capital and ownership. Is the capital in these strategies the firm's own or an investor's, and does the answer change the constraints on the desk?
• Work ownership. Which parts of this system would be mine, and which are shared? Would I be aligned to one trading team or to a product used across the firm?
• Research to production. Who takes a model from a research result to something live, and who owns it afterwards?
• Reliability. Is there an on-call rotation, what is its scope, and what proportion of an engineer's week is operational?
• Evaluation. What does a strong first year look like in writing, and who assesses it?
• Compensation mechanism. How is variable pay calculated, what is deferred, over what period, and what happens on departure?
• Mobility. Has anyone moved between horizons, teams or functions here in the last two years? HRT says it encourages rotation across its engineering dimensions (HRT); ask any firm for a recent example rather than the policy.

Where to check the specifics

Our firm profiles group employers by business model and show the functions each one is currently advertising for, so you can test the axes above against a named firm rather than a category. The comparison pages put two employers side by side on the attributes we hold for both.