Brevan Howard is a global macro hedge fund manager founded in 2002 by Alan Howard, with its flagship strategy launched in April 2003. It manages over $35bn and trades mostly discretionary directional and relative value strategies in fixed income, FX and equities. It also has a growing systematic business and BH Digital, its crypto and digital assets division. The main hubs are London, Jersey, Geneva, New York, Austin, Hong Kong, Singapore, Abu Dhabi and Bengaluru.
Junior hiring runs through a 10-week summer internship whose stated goal is to convert the best interns into the following year's graduate programme. The screen in front of it resembles a trading firm's process more than a bank's. You sit a coding test and 2 further online assessments, often before you have spoken to anyone.
Brevan Howard sends up to 3 online assessments from 3 vendors. These are a HackerRank coding test, a Predictive Index behavioural survey and an Ullix maths and statistics test. Candidates describe the Ullix test as easy and lose the process on the coding test, so put most of your preparation time into coding.
- 1Online assessments. HackerRank, Predictive Index and Ullix, sent together or spread across several weeks.
- 2HR interview. A short, non-technical call on motivation, markets interest and your CV.
- 3Market making game. Reported as the step between the assessments and the portfolio manager rounds.
- 4Portfolio manager and team interviews. Back-to-back conversations on markets, your projects and quantitative technique.
The order varies. Candidates report receiving all 3 assessments on the same day with no HR contact, while others went HackerRank, HR interview, then Ullix. The pace is slow. Assessments go out from mid October, offers start around February, and results are reviewed in batches, so a test can sit unopened for months. Candidates report interview invitations after weeks of silence, and about 3 months on average for internship processes, so a quiet month does not mean a rejection.
The 2027 programme
Applications for 2027 opened in early September 2026 and are listed as open until the end of April 2027. Apply early, because last cycle's assessments were already going out in mid October.
| Stream | Location | What it involves |
|---|---|---|
| Trading | London, New York | Work with discretionary and systematic investment teams, building data pipelines, models and trading signals, AI tooling and the core analytics library |
| Systematic Trading | New York | In 2026 this stream covered equities, macro, credit, commodities, digital assets and execution trading |
| Systematic Trading Technology Software Engineer | New York | A Java-first engineering seat on low-latency systems, paid at $150,000 annualised and prorated over the 10 weeks |
| Venture Capital | New York | Research and diligence on deep-tech startups |
Artificial Intelligence & Quant and Systematic Data Engineering internships were listed as opening later.
You need to be a penultimate-year undergraduate, or a 1st-year master's or PhD student, with your degree awarded before July 2028. The 2026 postings asked for a minimum 2:1 or 3.6 GPA, Python, and strong Excel and VBA. New York postings for the 2026 trading streams listed $12,500 to $16,600 per month.
The internship opens with 1 week of training before interns join a desk. It covers macroeconomics, FX, digital assets, interest rate derivatives, bonds, credit, trading strategies, risk management, Excel, Python and the latest AI tools. Interns describe a main programme in New York with an education focus, plus hands-on seats with specific portfolio managers, with the firm deciding who goes where.
Stage 1: Online assessments
| Assessment | What to expect |
|---|---|
| HackerRank | The coding screen for London and New York. Candidates report 75 minutes and 5 questions, made up of 2 Python, 1 SQL, 1 maths and 1 probability |
| Predictive Index | A short untimed behavioural survey built on adjective checklists. It measures working style and there is nothing to revise for |
| Ullix | Candidates report very basic maths and statistics questions, and describe it as the easy stage of a difficult process. Question counts and time limits are unreported, and you receive a scorecard link once the firm opens your results |
Treat that HackerRank shape as the working assumption, because the format can change between cycles. Candidates report reaching a first-round interview without solving the SQL question, so a complete set of answers is not required to progress. If you have sat the HackerRank or the Ullix test recently, tell us what you saw and we will update this guide.
How to prepare
- Python. The reported questions sit closer to data handling than to puzzle algorithms, so practise joins, cleaning, aggregation and summary statistics. The problems below cover it, with more in the coding practice catalog.
- SQL. Joins, GROUP BY with aggregates, HAVING filters and basic window functions. Write the queries by hand against a small dataset until the syntax costs you no time.
- Probability. Work the probability questions and the cards and coins collection.
- Maths and statistics. Drill speed on the Math Trainer, and use the probability and statistics course to cover the ground the Ullix test draws on.
- Time. 75 minutes across 5 questions is 15 minutes each. Read all 5 first, bank the maths and probability, then code.
Answer the Predictive Index honestly and consistently.
Stage 2: HR interview
A short call on motivation and background with no technical content. Reported questions are consistent across offices and years.
- Why Brevan Howard?
- Why are you interested in the markets?
- How do you follow the news?
- What makes you stand out from other applicants?
Candidates report that the interview after the assessments tests knowledge of the firm itself, and that interviewers question the fit of a purely technical background.
How to prepare
Build your answer on something specific to Brevan Howard, such as the macro mandate, the rates and FX focus, or the mix of discretionary portfolio managers and systematic teams. Follow markets daily through central bank decisions, inflation prints and moves in the yield curve, and be able to say what you read and what view you formed from it. If your background is purely technical, prepare a concrete account of how you became interested in markets.
Stage 3: Market making game
Candidates report being told by HR that a market making game follows the assessments and comes before the portfolio manager interviews. The format has not been described publicly, so prepare for the standard version. You quote a two-way price on an uncertain quantity, the interviewer trades against you, and you requote.
Practise on the dice game and the card game, and read our guide to make me a market if you have never quoted out loud.
Set a width that reflects your real uncertainty, and move your market after every fill. If the interviewer keeps lifting your offer, the true value is probably above it.
Stage 4: Portfolio manager and team interviews
The final rounds put you in front of the people you would work with. Candidates report 2 hours of 4 interviews at 30 minutes each with analysts and economists, covering machine learning, economics and finance. Reported questions include "What do you think of inflation?". Quant research candidates report being asked to explain their projects in detail. For AI and machine learning seats, candidates report online coding, live coding, a group interview and a final HR stage, with around 2 weeks between stages.
What they cover
- Macro views. Expect an open question on inflation, rates or a current policy story, and expect follow-ups on your reasoning.
- Your CV and projects. Every line is open to questioning, including the choices you made and what you would do differently.
- Machine learning and statistics. Expect these for quant and research seats. The validation course and our backtesting interview questions cover what systematic teams probe.
- Rates and FX fluency. This is our recommendation, based on what the firm trades. Know duration, convexity, the shape of the yield curve, swaps and carry. The Futures Basis Trainer drills fair value and cost of carry.
- Probability. Keep the brainteaser bank warm for quantitative follow-ups.
How to prepare
Bring 1 trade idea you can defend for 10 minutes. It should cover the instrument, the thesis, what is already priced in, the catalyst, what would prove you wrong and how you would size it. A modest idea with clear reasoning beats an ambitious one that falls apart at the second follow-up.
Key tips for success
- Apply in September or October, since assessments go out in waves and reviews happen in batches.
- Give the coding test most of your preparation time, and practise under a 75-minute clock.
- Do not wait on Brevan Howard. Keep other applications moving, because a gap of several weeks between stages is normal here.
- Read about markets every day from the moment you apply, because the HR call and the final rounds both test it.
Closing remarks
Brevan Howard combines a quantitative screen with a macro conversation. The assessments test Python, SQL, probability and basic statistics, and the interviews test whether you have real views on rates, inflation and FX and can defend your own work. Candidates who prepare only one side tend to stall at the other.
The process is slow and the order of stages is inconsistent, so judge your progress by what you have completed and not by how long you have been waiting. Work through everything we have for Brevan Howard, and read our full interview guide for how these stages compare across firms.