Derivative Quantitative Research Trading Team

You sit in a team that builds and tests the mathematical models used to price and hedge derivatives - financial contracts (options, swaps, futures) whose value depends on something else, like a share price or an interest rate. The work is mostly writing code, analysing market data and checking whether a model's assumptions hold up in real trading, and the "customers" are the traders on the desk who take positions using those models, plus the risk and control functions who need to know what the positions could lose. In practice the job title lumps together several different jobs - quantitative research, quantitative trading and quant developer work - and which one you actually do depends heavily on the employer.

Approximate graduate salary

Quantitative graduate roles are at the higher end of the UK graduate market. Starting packages are typically somewhere around GBP 50,000-80,000 or more, and this varies a great deal - by employer type (banks versus proprietary trading firms, which tend to pay more), by whether you hold a PhD, and by how much of the package is bonus rather than base salary. Treat any single figure with caution; the spread is genuinely wide and London roles dominate the market.

What you'd actually do

Writing and debugging code, usually Python for research and analysis, sometimes C++ or another compiled language for anything that has to run fast in production - this is genuinely most of the week for many people in these teams.
Pulling apart market data: cleaning tick data or end-of-day prices, spotting bad records, building the historical data sets that any model or backtest depends on.
Backtesting - running a pricing model or trading idea over historical data to see how it would have behaved, then arguing about whether the result is real or an artefact of how the test was set up.
Calibrating pricing models: taking observed market prices for liquid instruments and fitting model parameters so the model reproduces them, then using it to price something less liquid.
Answering ad hoc questions from traders, often quickly and under pressure - 'why does the model say this option is worth that?', 'what's my exposure if volatility jumps?' - which means being able to explain maths to someone who wants a one-line answer.
Producing and checking risk numbers, particularly the 'Greeks' (sensitivities of a position's value to moves in price, volatility, time and rates), and investigating when the profit and loss explained by the model doesn't match what actually happened.
Sitting in on or near the trading desk, in meetings with traders, risk managers and sometimes model validation teams who independently review the models before they can be used.

How graduates get in

Structured graduate schemes at investment banks, run under headings like 'quantitative analytics', 'markets quant' or 'strats'. This is the most common single route. Applications typically open in the autumn for the following summer and are often filled early, so applying late in the cycle is a real disadvantage.
Converting from a summer internship. At many banks and larger trading firms the internship is effectively the main hiring channel for graduate places, so applying to penultimate-year internships matters more than most students realise.
Direct entry at proprietary trading firms and market makers, which usually run their own hiring rounds rather than a formal 'graduate scheme'. These tend to test maths, probability and mental arithmetic very hard and care less about your CV narrative.
From a PhD or research master's in maths, physics, statistics, computer science or engineering. This is common for the more research-heavy 'quant research' seats and, at some employers, is close to the default. A PhD is not required for most graduate-level roles, though.
Hedge funds and asset managers hire quantitative graduates too, but usually in smaller numbers and often without a publicised scheme - these roles are found through targeted applications, academic supervisors' contacts or specialist recruiters.
Moving sideways after a couple of years in a related role - risk, model validation, technology or a data-heavy analytics job - is a realistic but slower route in, and more common than people assume.

What employers ask for

A strongly quantitative degree: maths, physics, statistics, computer science, engineering, or economics with a heavy maths component. The subject genuinely matters here, unlike in most graduate finance roles - a non-quantitative degree is a real obstacle.
A high degree classification. A first or high 2:1 is the usual expectation, and at the more selective firms a first (or a postgraduate degree) is effectively the norm. Some employers still ask for strong A-level results, particularly in maths.
Demonstrable programming ability, usually Python and often C++ as well. Employers vary in how much they expect on day one: some teach it, others test you on it in the interview and expect working fluency.
Comfort with the specific maths that underpins derivatives: probability, stochastic calculus, linear algebra, numerical methods, statistics. You will not be expected to know derivatives pricing in detail before you start, but you will be expected to pick it up fast.
Postgraduate study - an MSc in financial mathematics, computational finance, statistics or similar, or a PhD - is common but not universally required. It helps most for research-heavy seats and matters less at firms hiring on raw problem-solving tests.
No professional exam is required to do the job. Some people take the CQF (Certificate in Quantitative Finance) or CFA later, but neither is a gatekeeper in the way that, say, ACA is in accountancy.

Skills that matter

Programming that other people can use

Almost all output is code - a pricing library, an analysis script, a backtest - and code that only you can read or that breaks under real data is worth very little on a desk.

Probability and stochastic calculus

Derivatives pricing rests on modelling uncertain future prices as random processes, so you need to be able to reason about distributions, expectations and how randomness compounds over time rather than just apply formulas.

Numerical methods

Most realistic pricing problems have no clean closed-form answer, so you solve them with Monte Carlo simulation, finite difference methods or optimisation routines, and you need to know when the numbers are converging and when they're quietly wrong.

Data scepticism

Market data is full of stale prices, bad ticks and survivorship effects, and a huge share of impressive-looking backtest results are just data problems that nobody caught.

Explaining a model to a non-mathematician in one or two sentences

Traders and risk managers make decisions on your numbers without reading your derivation, so being able to say what the model assumes and where it breaks is what makes you trusted rather than tolerated.

Knowing when a model is wrong enough to matter

Every pricing model is a simplification, and the judgement that separates good quants from merely capable ones is spotting which simplifications are harmless and which will cost money in a stressed market.

Where it leads

First one to three years: learning the products, the firm's existing code base and the desk's conventions. Most of your work is small, well-defined pieces of a bigger model or analysis, reviewed by someone senior. Progression here is fairly structured at large banks and much more informal at smaller trading firms.
Next: owning a model, an asset class or a piece of infrastructure end to end - you become the person the desk asks about, say, interest rate volatility or a particular exotic product. Timelines vary widely; this can be three years in or considerably longer.
A fork in the road. Some people move towards research seniority (designing models, leading a small quant team), some move onto the trading desk itself and start taking risk, and some move into quantitative development and build the systems. Which doors are open depends a lot on the firm's structure.
Senior routes include heading a quant team, becoming a portfolio manager or senior trader, or moving into model risk and validation - the function that independently checks models, which is closely tied to regulatory expectations for banks.
Exits outside finance are genuinely common and worth knowing about: technology, data science, energy and commodities trading, insurance modelling and academia all hire people with this background. The skills travel better than the job title suggests.

What people get wrong

“It's a trading job - you'll be taking positions and watching profit and loss from week one.”

In most of these teams, especially at banks, the quant researcher does not take risk at all. Traders take the positions; you build and maintain the tools they use. Some proprietary trading firms do blend the two roles, but assuming that is the norm leads people to apply for the wrong jobs.

“You need a PhD to get in.”

A PhD is common and is close to expected for some pure-research seats, but plenty of people enter from an undergraduate or master's degree, particularly at trading firms that hire on problem-solving tests rather than research track record. The stronger universal requirement is high-level maths plus real coding ability.

“It's mostly elegant mathematics, and software engineering is somebody else's job.”

A large share of the week is engineering and data work - cleaning inputs, maintaining libraries, chasing down why a number changed. People who dislike the unglamorous parts of coding tend to be unhappy in these roles regardless of how good their maths is.

“All 'quant' jobs are basically the same.”

The umbrella covers quite different jobs: front-office desk quants who sit with traders and answer questions in minutes, model validation quants who independently review and challenge models, quant developers who mainly write production software, and systematic researchers who look for trading signals. The daily experience, hours and culture differ substantially, so it's worth reading a job description carefully rather than the title.

Where this varies

The biggest variation is employer type. At large investment banks the role is usually formalised, sits alongside risk and control functions, is subject to model governance and regulatory review, and comes with a structured graduate programme, rotations and clear titles. At proprietary trading firms and market makers, teams are smaller and flatter, hiring is based heavily on timed maths, probability and coding tests, and the line between researcher and trader is blurrier. Hedge funds and systematic asset managers sit somewhere in between and often hire in ones and twos rather than cohorts. Geographically, this is overwhelmingly a London role, with smaller clusters elsewhere - some energy and commodities quant work sits outside London, and a few firms have technology or research offices in other UK cities. One further limit worth being explicit about: because these teams are secretive about their methods, published detail on what any particular firm's researchers do day to day is thin, so judge individual roles by the job description and the interview process rather than by the team's name.

General guidance about the role across the UK market, not about any specific employer. Entry routes and requirements vary — always check the individual job advert.