Hudson River Trading Quantitative Researcher mock interview

PhD-track 'Algorithm Developer (Quant Researcher)' role. ~7-week loop. OA → 1-2 phone screens → 4-6 round virtual onsite. C++ required. Spaghetti / coin / cube brainteasers. ~$651K Y1 reported (signing deferred over 2 years).

  • ~7 weeks
    End-to-end loop
  • ~$651K
    Y1 QR reported
  • 4-6
    Onsite rounds
  • Free
    First mock

Why HRT QR is structurally different — fundamentals depth over breadth, with deferred signing

Per HRT's HRTBeat post 'Engineering and Interviewing at HRT' (hudsonrivertrading.com/hrtbeat/engineering-and-interviewing-at-hrt), Quant Blueprint salary data, TheWallStreetQuants HRT question bank, and aggregated candidate reports, HRT QR is structured around fundamentals depth + C++ rigor + research engagement. The ~7-week loop: OA (3 LeetCode E-M, ~2.5h) → 1-2 technical phone screens with stats/probability brainteasers → final virtual onsite (4-6 back-to-back rounds mixing pure math, coding, research, culture).

Brainteaser depth gates the phone math rounds. HRT-specific patterns: spaghetti-strand looping (n strands), three-bowls-with-colored-balls, top-3 fastest of 25 bunnies using only relative race orderings, cube-painting unit-cube probabilities, 1000-coin Bayesian. Per TheWallStreetQuants, a specific QR phone-math fail mode is 'failing to prove method as n→N' — solving the small case but not generalizing flags missing mathematical maturity. Always derive the general case before the specific.

Compensation is high but structurally different. Per Quant Blueprint: QR Y1 ~$651K = $350K base + $185K signing + $116K performance bonus. CRITICAL: signing bonus is DEFERRED over 2 years, paid quarterly (per TeamRora). This is less favorable than peer top-tier funds where signing is upfront. HRT comp has 'normalized' to mid-range vs Citadel/Jane Street per TeamRora. C++ proficiency is REQUIRED for QR (Mid-Freq track Bachelor's-acceptable per Quant Blueprint listing). HRT's official rejection signals (per HRTBeat post) include 'cheating with LLMs' explicitly — they detect AI-generated code patterns.

What the loop looks like

Round-by-round

  1. 01

    Application + recruiter

    Pre-onsite
    30 min

    Background, PhD focus, motivation. HRT is famously private — public materials sparse. 'Why HRT' must reference concrete signals (HRTBeat blog, specific research interests aligned with mid-frequency trading).

  2. 02

    Online Assessment

    Pre-onsite
    ~2.5 hours

    Sent ~1 week after application. 3 LeetCode Easy-Medium coding problems, often including a brainteaser-style optimal-algorithm question. Cheating detection — HRT explicitly bans LLM use.

  3. 03

    Technical phone screen × 1-2

    Pre-onsite
    60 min each

    Probability/statistics brainteasers — coin flips, conditional probability, expected value. Must prove method as n→N for general cases. Going silent = fail.

  4. 04

    Virtual onsite × 4-6

    Onsite
    60 min each

    Back-to-back rounds: pure math/probability, coding (C++ required for QR), 1-2 conversations on research background and culture fit. Spaghetti / cube / 3-bowl brainteasers prevalent.

Question bank

Real interview questions reported by candidates

  • Probability

    "Jar of 1,000 coins, one is double-headed. You pick one and flip 10 heads in a row. P(double-headed)? Walk through Bayesian update."

    Source · thewallstreetquants.com HRT interview questions
  • Probability

    "Construct correlated random variables from independent Gaussians. Given target correlation matrix, derive the transformation."

    Source · thewallstreetquants.com HRT interview questions
  • Brainteaser

    "Brainteaser: n strands of spaghetti in a pot. Take random ends and tie them together until none free. Expected number of loops as a function of n?"

    Source · thewallstreetquants.com HRT interview questions
  • Brainteaser

    "25 bunnies racing 5 at a time on a track with no clock. Minimum number of races to find the top 3 fastest?"

    Source · thewallstreetquants.com HRT interview questions
  • Coding

    "Coding: implement two-sum where input is an array of doubles (not ints). Discuss precision issues, hash strategy."

    Source · thewallstreetquants.com HRT interview questions
  • Coding

    "Coding (C++): write a simple memory pool allocator. Discuss alignment, free list management, fragmentation."

    Source · thewallstreetquants.com HRT interview questions
  • Research deep-dive

    "Discuss your PhD research. What novel methods did you develop (not just tools you used)? What are limitations? How would you extend?"

    Source · hudsonrivertrading.com hrtbeat engineering and interviewing, rebellionresearch.com HRT interview
  • Behavioral

    "Why HRT specifically (vs Citadel, Jane Street, Two Sigma)? What about HRT's culture — 'built by coders, led by coders' — resonates with you?"

    Source · hudsonrivertrading.com hrtbeat, hudsonrivertrading.com careers
Common signals to fix

What gets you rejected at this level

  • Shallow fundamentals

    Per HRT's own published interview-prep post (hudsonrivertrading.com/hrtbeat/engineering-and-interviewing-at-hrt): rejections are most often due to shallow fundamentals. They explicitly value depth over flashy tricks. Generic answers without proof or derivation flag missing rigor.

  • Poor communication / poor listening

    Per HRT's official post: rejections cite 'poor communication, poor listening' as common signals. HRT interviewers want narrated reasoning AND active engagement with their hints. Going silent during phone math rounds when stuck = fail.

  • Over-engineering instead of writing real code

    Per HRT's official post: candidates who add abstractions, design patterns, or premature optimization instead of writing working code get flagged. HRT wants pragmatic engineering judgment. Don't 'architect' a solution when 'code' is asked for.

  • Cheating with LLMs

    Per HRT's own post: explicitly listed as a rejection signal. HRT detects LLM-generated code patterns. Don't use AI assistance during the OA or any phone round, even if not directly observed.

  • Failing to prove method as n→N

    Per thewallstreetquants.com: a specific phone-math fail mode for QR. Candidates who solve a small case but can't generalize the method to arbitrary n flag missing mathematical maturity. Always derive the general case before the specific.

How Applr's AI mock interview tracks HRT's QR rubric

Applr's HRT QR mock simulates the brainteaser-heavy phone math format — probability puzzles with explicit grading on prove-method-as-n→N (matching HRT's official failure-mode signal). Going silent during stuck moments gets flagged with practice prompts on narrating uncertainty.

Coding rounds prompt C++-specific patterns (memory pool, smart pointer semantics, custom allocators) — not just LeetCode correctness. Research deep-dive rounds prompt 'novel methods you developed (not just tools you used)' framing — generic 'I used XGBoost on this dataset' answers get rewrites that surface methodological contributions.

Adjacent loops

Related practice

FAQ

Do I need a PhD for HRT QR?

For the 'Algorithm Developer (Quant Researcher) PhD' track: yes — full-time PhD in math, physics, CS, statistics, or related is required (per HRT's own job posting at hudsonrivertrading.com/hrt-job/algorithm-developer-quant-researcher-2026-phds). HOWEVER, a separate Mid-Frequency QR posting requires only a Bachelor's in math/CS/stats/physics + C++ proficiency (per Quant Blueprint job listing). So PhD is not universal — confirm with recruiter which track you're applying to. Putnam/IMO not required in any public posting, but competition math/olympiad backgrounds are common among HRT hires.

What's HRT QR salary actually?

Per Quant Blueprint (quantblueprint.com/salaries/hudson-river-trading): QR Y1 total ~$651K = $350K base + $185K signing + $116K performance bonus. Algo Developer counterpart Y1 ~$553K. HRT's posted base range on job postings: $175K-$300K + discretionary perf bonus + signing. CRITICAL: signing bonus is DEFERRED over 2 years, paid quarterly (per TeamRora). This is less favorable than peer top-tier funds where signing is upfront. Per TeamRora: HRT comp has 'normalized' to mid-range vs Citadel/Jane Street. WSO reports avg QR ~$450K (broader sample).

Is C++ really required for QR?

For the QR track per Quant Blueprint job listing: yes, C++ proficiency is REQUIRED. HRT is a low-latency trading firm and the research-engineering bar reflects that. Python is also accepted in interview (per HRT's HRTBeat post 'Engineering and Interviewing at HRT'). For QR-track candidates: practice C++ memory management, STL, smart pointers, custom allocators before the loop. Python alone won't pass the C++ interview rounds.

How is the OA different from typical HackerRank?

Per Glassdoor and TheWallStreetQuants candidate reports: HRT's OA is 3 LeetCode Easy-Medium coding problems over ~2.5 hours, often including a brainteaser-style optimal-algorithm question. Sent ~1 week after application. Format is HackerRank-like but problem mix has more brainteaser flavor than typical FAANG OAs. Practice timed problems with optimal-complexity emphasis, not just correctness.

What's HRT's culture really like?

Per HRT's own materials (hudsonrivertrading.com/hrtbeat, hudsonrivertrading.com/careers): 'built by coders, led by coders.' Public materials are notoriously sparse compared to Citadel/Two Sigma — HRT prefers privacy. They explicitly value depth of fundamentals, clean communication, and grace. The 'No Jerks' filter applies — brilliant-but-unworkable candidates get cut. Be ready to demonstrate intellectual honesty (admit when you're stuck, ask clarifying questions).

Does HRT sponsor visas?

Yes for QR roles in NYC HQ and other US offices. Strong sponsor — international PhD candidates with US work eligibility have a clear path. Confirm timing with recruiter early — visa coordination can extend the timeline 2-4 weeks beyond standard 7-week loop.

Ready to practice for Hudson River Trading?

Start free →