Quant SME

Robson Bale Ltd; Greater London; On-site; GBP 175,000 - 185,000; Full time

Job summary

Robson Bale Ltd is seeking a Quant SME for a contract role on-site in London, 5 days a week, at £950–£1000 per day (Umbrella). The position centers on building and maintaining a Python library that downstream services depend on, with a focus on multi-asset portfolio statistics and time-series data engineering using pandas.

You will also work with Snowflake data warehousing, write SQL against vendor tables, and interface with Thomson Reuters/Bloomberg tools.

Qualifications

• Experience with Thomson Reuters and Bloomberg applications, and Excel Add-ins.

Responsibilities

• Build and maintain a Python library used by downstream services.
• Develop and maintain time-series portfolio statistics with pandas at scale.
• Write SQL against Snowflake/vendor tables for curated data.
• Collaborate with front-end/middle-tier teams on API implementations.
• Contribute to CI/CD pipelines and testing frameworks.

Skills

Python 3.11+

DSL parsing

Pandas/Numpy

Portfolio statistics

CI/CD testing discipline

Linting/Static typing

Software architecture

API design

Team collaboration

AI-assisted development

Tools

Snowflake

SQL

GitLab CI/CD

pytest/Hypothesis

ruff

pyright

numpy

scipy

cvxpy

numba

Bloomberg Thomson Reuters tools

Job description

Quant SME – Contract - 5 days per week on site in London - £950-1000 per day via Umbrella

This role is 5 days per week on site in London

Experience:

• Exposure to Thomson Reuters, Bloomberg applications, Excel Add-ins, etc.
• Hands-on experience in computing multi-asset portfolio statistics
• Building/maintaining a Python library (not app work): packaging, versioning, releasing a package that downstream services depend on. This is the core of the job.
• Time-series data engineering with pandas at scale (multi-index frames, frequency alignment, currency/tenor handling). pandas is the entire data plane here.
• Snowflake / data-warehouse experience: writing SQL against curated vendor tables, EAV vs wide-table schemas. This is how all market data actually arrives — not via Bloomberg/Reuters terminals directly.

Skills:

• Strong Python 3.11+: dataclasses, type hints, pydantic v2 — not just "knowledge of Python." Set a clear bar (mid/senior).
• Parser / AST / interpreter design — this is a DSL. Comfort with tokenizing, expression trees, immutable AST nodes, and a parse→validate→execute pipeline is the single most repo‑specific skill.
• pandas + numpy/scipy proficiency (statsmodels, scikit‑learn for the analytics extra).
• GitLab CI/CD specifically (pipelines, not just "use gitlab"): the repo runs lint→test→build→publish stages.
• Testing discipline: pytest, property‑based testing (Hypothesis), fixture/regression suites. Given ~230 test files and hermetic catalog‑regression, this is essential — I'd make it a first‑class skill, not implied.
• Linting/quality tooling: ruff, type checking (pyright/basedpyright).
• Optimization libraries: scipy (SLSQP), optionally cvxpy/numba — for the weight‑optimization functions.
• API/contract discipline: designing stable interfaces since a separate service consumes this library.
• Knowledge of how to use gitlab
• Knowledge of Software architecture principles
• Team player, supporting front‑end, middle‑tier teams with API implementations
• AI‑assisted / agent‑driven development — the repo is heavily tooled with Cursor skills and agent workflows (mapping helpers, changelog automation, gap analysis). A candidate comfortable working alongside these tools will ramp much faster.

Knowledge:

• Understanding of how to compute portfolio statistics (returns, volatility, correlations, ratios, greeks) and interpret these numbers to make sure results make sense. Using edge cases to stress test correctness of analytical computations
• Understanding of leverage in portfolio construction
• Portfolio optimization (not just statistics): mean‑variance, max‑Sharpe/Sortino, risk parity, efficient frontier, constraints
• Portfolio simulation / rebalancing cost modeling — path‑dependent portfolio simulation with rebalancing cost/turnover, not a classic event‑driven backtester.
• Numerical/matrix methods: PSD covariance repair (nearest‑correlation/Higham), Ledoit–Wolf shrinkage, regression. Strong numerical‑stability instincts matter here.
• FX conversion & multi‑currency return handling (hedged/rebased returns).
• Liability‑driven / goals‑based analytics (liability paths, success rates) — present in the repo, likely valued by your team.
• Financial data quality & metadata semantics: understanding frequency, tenor, measurement type, and enforcing correctness across a large function catalog. The repo has a strict metadata contract enforced in CI.
• Portfolio to benchmark attribution, portfolio return bias based on risk factors
• Market Data
• Understanding on how to deal with missing data for some of portfolio assets, also when computing asset correlations and covariances
• Knowledge of Technical Analysis methods
• Knowledge of Fundamental analysis ratios
• Portfolio Back‑testing and Screening
• Understanding how back‑testing works, with focus on leverage drift for rebalancing.
• Order Management
• Understanding of trade order management basics, VWAP, etc. trade order execution strategies