PDT Partners: Careers, Culture & Interview Guide 2026

The Quietest Name in Quant

In 1993, a mathematician named Peter Muller set up a small trading group inside Morgan Stanley called Process Driven Trading. Over the next two decades it became one of the most consistently profitable systematic trading operations anywhere, reportedly generating hundreds of millions of dollars a year for the bank at its peak. When post-crisis regulation made proprietary trading desks untenable inside banks, Muller took the group independent. PDT Partners spun out of Morgan Stanley in 2012 and has run as a standalone fund ever since.

Most people in finance have never heard of it. That is deliberate. PDT has roughly 300 employees, offices in New York and London, no marketing to speak of, and a founder who was once better known for playing keyboard in the New York subway than for running one of the best track records in quantitative finance. Scott Patterson's 2010 book The Quants gave outsiders a rare glimpse of the group; since then, PDT has said very little in public.

For candidates, that silence cuts both ways. PDT is one of the most selective employers in the industry, and there is far less public information about its process than for Citadel or Two Sigma. This guide assembles what is publicly known and anecdotally reported: what the firm does, how it is structured, how the interview process reportedly works, and what it pays.

---

PDT at a Glance

• Founded: 1993 by Peter Muller, inside Morgan Stanley; independent since 2012
• Type: Systematic hedge fund (quantitative, research-driven)
• Headquarters: New York, with a London office
• Size: Roughly 300 employees
• Style: Mid-frequency statistical strategies across global markets, run almost entirely by models
• Roles they hire: Quantitative Researcher, Software Engineer, and a small number of infrastructure and business roles
• Reputation: Extremely secretive, PhD-heavy, top-of-market pay, very low staff turnover

---

What PDT Actually Does

PDT is a systematic fund in the classic sense: researchers build statistical models that forecast returns, and those models trade with limited human intervention. The firm is usually grouped with Renaissance Technologies, D. E. Shaw and Two Sigma as one of the research-led quant houses, as opposed to the multi-strategy pod shops like Millennium and Citadel where humans run discretionary books alongside quants.

Public detail on strategy is thin, which is itself informative. What is consistently reported is that PDT trades liquid instruments across equities and futures globally, holds positions over horizons of roughly hours to weeks rather than microseconds, and reinvests heavily in research rather than growing assets aggressively. Capacity discipline is part of the story former employees tell: the firm has reportedly returned capital rather than dilute returns.

If you want the broader landscape before going deeper, our quant hedge fund guide maps where firms like PDT sit relative to the rest of the industry.

---

Culture: Small, Academic, Long-Tenured

Three things come up repeatedly in accounts from people who have worked at or interviewed with PDT.

It runs like a research lab. A large share of the research staff hold PhDs in mathematics, physics, statistics or computer science, and the working culture is closer to an academic department than a trading floor. Ideas are debated on evidence. Seniority matters less than whether your backtest survives scrutiny.

It is genuinely small. Around 300 people is tiny for a firm with PDT's history. Compare Two Sigma at roughly 2,000 employees or D. E. Shaw at over 2,500. The practical consequence is that individual researchers own meaningful pieces of the production system, and there is nowhere to hide.

People stay. Turnover is reportedly among the lowest in the industry. That is partly pay, partly culture, and partly the fact that PDT's collaborative model (a shared research platform rather than siloed pods) means your work compounds over years. The flipside for candidates: headcount growth is slow, and open seats are rare.

Muller himself remains an unusual figure by hedge fund standards - a published songwriter and a serious poker player who once took a multi-year sabbatical from Morgan Stanley to travel and write music. The firm's low-ego, slightly bohemian streak is often traced back to him.

---

Roles and Who Gets Hired

PDT hires into two main technical tracks.

Quantitative Researcher. The core job: finding, testing and productionising predictive signals. Most researcher hires come from PhD programmes in quantitative fields or from other top research shops. Publication records, olympiad results and strong research references all reportedly carry weight. Machine learning and statistics dominate the skill set; finance knowledge is not expected at entry.

Software Engineer. PDT's trading is fully systematic, so the engineering platform is the firm. Engineers work on execution systems, research infrastructure, and the data pipeline. The bar is comparable to top-tier technology firms, with an emphasis on writing careful, correct code rather than shipping fast.

There is no large graduate intake and no famous internship programme on the scale of Jane Street's or Citadel's. Hiring is small-batch and year-round, which means the process can feel slow. Candidates coming from academia should treat the application like a research job: a CV that reads like a strong academic record plus evidence of practical coding ability.

---

The Interview Process, As Reported

PDT does not publish its process, so the outline below is assembled from candidate accounts on forums and anecdotal reports. Treat it as indicative rather than definitive; formats change and vary by role and seniority.

Stage 1: CV screen and recruiter call

A short call covering background and motivation. For PhD candidates this often focuses on the thesis: what you did, why it mattered, what you would do differently.

Stage 2: Technical phone screens

One to three calls, typically 45 to 60 minutes each. For researchers, expect probability and statistics questions, some at a level noticeably harder than the standard interview canon, plus questions probing how you think about noisy data, overfitting and model validation. For engineers, live coding and algorithms with attention to correctness and edge cases.

Stage 3: Onsite (or virtual onsite)

A longer day of four to six sessions. Reported components for research candidates include:

• A research presentation. PhD candidates are often asked to present their own research to a room of PDT researchers, who probe assumptions hard. This is reportedly the most heavily weighted part of the process.
• Probability and statistics rounds. Estimation problems, stochastic processes, conditional expectation puzzles and Bayesian reasoning.
• Machine learning and data analysis. Practical questions: how would you detect regime change, how do you validate a model when the data is autocorrelated, when does cross-validation mislead you.
• Coding. Usually Python-centric for researchers, with an emphasis on clean handling of real data rather than algorithmic gymnastics.

Stage 4: Decision

Timelines are reportedly slow relative to peers - weeks rather than days - consistent with a firm that hires rarely and deliberately.

For preparation, the probability component overlaps heavily with the standard quant canon. Work through our probability interview questions and the green book study guide, then go a level deeper: PDT interviewers reportedly enjoy questions that extend a standard puzzle in an unexpected direction, and the candidates who do well are the ones who can keep reasoning past the memorised answer.

---

What PDT Reportedly Pays

PDT does not disclose compensation, and because headcount is small there is far less levels.fyi and forum data than for larger firms. The consistent anecdotal claim is that PDT pays at or near the top of the market, in the same conversation as Renaissance and the top pod-shop seats, precisely because it hires so few people and wants to keep them.

Indicative estimates from forum discussion and reported offers, all approximate:

Because PDT runs a collaborative research model rather than pods, pay is reportedly driven by overall fund performance and individual contribution rather than a formulaic share of a personal book. In strong fund years the bonus pool expands substantially. For cross-industry context, see our quant researcher salary guide.

---

PDT vs Renaissance, Two Sigma and D. E. Shaw

The comparison candidates most often want:

The realistic takeaway: Two Sigma and D. E. Shaw hire at scale and are the practical route in for most strong graduates. PDT and Renaissance hire a handful of people a year and are best treated as targets of opportunity rather than a plan.

---

What This Guide Cannot Tell You

An honest limitation section matters more for PDT than for almost any other firm, so here it is.

Because PDT is private, small and quiet, nearly everything above the level of basic corporate history rests on anecdote: candidate reports, former-employee accounts and secondhand forum discussion. We cannot verify the interview stages, the pay figures or the internal structure against official sources, because there are none. Specifics also go stale; a process reported by a 2023 candidate may not match what you face in 2026.

What we can say with confidence is structural. The firm is small, research-led and slow-hiring, and every credible account agrees the technical bar sits at the top of the industry. Prepare for the hardest version of the process and you will not be caught out by an easier one.

---

Compensation & recruiting notes

Pay ranges in this guide are illustrative estimates from publicly discussed bands and anecdotal reports - not official figures from the employer. Packages vary widely by role, office, performance, and year. Hiring processes change; nothing here guarantees an interview, assessment format, or offer.

---

Frequently Asked Questions

What is PDT Partners known for?

PDT (Process Driven Trading) Partners is a systematic hedge fund founded by Peter Muller inside Morgan Stanley in 1993 and independent since 2012. It is known for a long, strong track record, an unusually small and academic culture, and extreme secrecy about its strategies.

How hard is it to get a job at PDT Partners?

Very hard, primarily because of volume. With roughly 300 employees and low turnover, PDT hires a small number of people each year, mostly PhDs in quantitative fields and experienced engineers. The technical bar is reportedly comparable to Renaissance and the top research seats at D. E. Shaw.

Does PDT Partners hire graduates without a PhD?

Reportedly yes for engineering roles, where a strong computer science background matters more than a doctorate. For quantitative research roles, the intake skews heavily toward PhDs, though exceptional undergraduates and masters candidates with olympiad-level records do appear.

What does the PDT interview process involve?

Candidate accounts describe a recruiter screen, one to three technical phone rounds, then an onsite with probability and statistics, machine learning discussion, coding, and (for PhD candidates) a presentation of your own research. The research presentation is reportedly weighted heavily.

How much does PDT Partners pay?

PDT does not disclose pay. Anecdotal reports place new PhD researchers at roughly $400,000 to$700,000 in year-one total compensation, with senior researchers earning seven figures in good fund years. Treat all figures as estimates.

Is PDT Partners a hedge fund or a prop firm?

A hedge fund: it manages external capital (alongside significant internal capital) and charges fund fees. The trading itself is fully systematic, which makes it feel closer to a research lab than to a discretionary fund. Our quant hedge fund guide explains the distinction in detail.

How should I prepare for a PDT interview?

Build depth in probability, statistics and machine learning beyond the standard interview canon, be ready to present and defend your own research, and sharpen practical Python. Working through quant interview questions is a sensible baseline, but expect follow-ups that push past the textbook answer.