You own the software a trading desk runs on. Prices come in, orders go out, and in between sit the positions, the profit and loss (P&L) and the record that proves it all afterwards. You lead the engineers and quants who build that, you decide what to build and what to buy, and when a number is wrong at 07:30 the head of trading calls you. It is a trading job that happens to be done in software, not an IT job that happens to be near traders.
Why it is in demand right now
Trading is moving closer to real time. European intraday power volumes were up 39% year on year in August 2026 (EEX). That is more decisions per hour than people can make by hand, so desks are building algorithmic platforms and need someone accountable for them. ENGIE advertised a Head of Trading for algorithmic energy markets in London in August 2026 with the brief to build "a best-in-class algorithmic trading platform".
Batteries and flexible assets. Every battery needs software that trades day-ahead, intraday, balancing and ancillary services at the same time. Kraken, the Octopus technology arm, was spun out as a company valued at $8.65bn and bought a Danish trading-automation start-up rather than build from scratch.
Old systems are being replaced. Shell created a Market Data Integration Lead inside its trading business in August 2026. Senior architect and director roles for ETRM systems (the trade capture and risk software a desk runs on) are open in London, and specialist ETRM analysts are being offered £100k to £140k, or £1,000 a day.
What it pays (London, 2026)
| Role | Pay |
|---|---|
| Permanent, base | £130k to £200k |
| With bonus | £180k to £300k |
| Contract, per day | £900 to £1,300 |
Indicative. Anchored on August 2026 postings: ETRM analysts at £100k to £140k or £1,000 a day; a Head of IT for a commodities trading firm at €120k to €180k. Oil majors and trading houses pay at the top; utilities below.
What the interview really tests
- Can you walk a trade from the click to the settlement invoice, and say where it usually breaks?
- Can you tell that a P&L number is wrong before the trader does?
- Have you shipped something a desk relied on? What happened the day it failed?
- Build or buy an ETRM: can you argue both sides with numbers?
Where it goes from here, honestly
Net positive, but more selective than it looks. The tailwinds are real: intraday volumes, batteries, and power purchase agreements with floors and ceilings that need proper pricing and risk systems. The headwinds are just as real. Battery revenues are already compressing as capital piles in, and when margins fall, technology budgets get questioned. Consolidation favours the large houses, so the boutique that hired three developers last year may not exist in three. AI tooling means one senior engineer now ships what a small team did, so junior developer jobs are thinning out. And REMIT II, the updated European rulebook for wholesale energy trading, turns part of the job into compliance tooling.
Put together: the role is safe for the person who is accountable for the numbers and can run a platform like a product. It is not safe for the person who writes the boilerplate. Everything below is aimed at the first person.
The roadmap
Four stages, in order. Each stage names the guides and projects to work through in the Energy Trading & ETRM Skills library, roughly how long it takes, and the thing you will have at the end that you can show someone. Guides are textbook-style PDFs; projects are code you build.
| Stage | Time |
|---|---|
| 1. Understand the desk | about 20 hours |
| 2. Run the plumbing | about 20 hours |
| 3. Build it | about 40 hours |
| 4. Lead it | about 15 hours |
Around 95 hours in total, or eight weeks at twelve hours a week.
Stage 1: Understand the desk
Energy Trading Fundamentals course. What a position is, what a curve is, what P&L means and what limits are for. You cannot run the systems until you can read the numbers they produce.
- Trading Fundamentals
- Position Aggregation and Delta Exposure
- Forward Curves and Market Data
- Mark-to-Market P&L
- Risk Management Fundamentals
What you leave behind: a one-page explanation, in your own words, of how a six-trade book turns into a P&L number.
Stage 2: Run the plumbing
Practitioner Track. The pipeline a trade goes through and the nightly jobs that keep the desk honest. This is where most senior candidates are weakest, and where the interview questions come from.
- Trade Lifecycle and Settlement
- Market Data Operations
- Market Microstructure and Order Books
- Day-Ahead and Intraday Power Markets
What you leave behind: a drawn map of the ten stages from trade capture to reconciliation, and a written diagnosis of the guide's EUR 420k "P&L moved with no new trades" incident.
Stage 3: Build it
Python for Commodities and the Systems Builder Track. Working code, with tests, on GitHub. Each project reuses the same reference portfolio as the guides, so the numbers tie out.
- Writing Professional Python
- Production Python for Trading Systems, with the
settlerepository - Project 2: Position and P&L Engine
- Project 3: A Toy Power Exchange
What you leave behind: two public repositories: a position and P&L engine that reproduces the guides' figures, and an exchange that clears a day-ahead auction and matches intraday orders.
Stage 4: Lead it
The parts of the Practitioner Track that a head of function is judged on: what the regulator requires of an automated desk, how strategies are chosen, and how a trading bot is actually run.
- Energy Market Regulation and Compliance, with the REMIT II algorithmic trading obligations
- Algorithmic Trading in Power Markets, with the
cadencebot - Power Trading Strategies
What you leave behind: a one-page build-or-buy note for an ETRM, and a running bot with its control and audit log, which is the thing a hiring manager will actually open.
If you already work on a desk: skip the guides in Stage 1 you can teach, but do not skip the projects. The projects are the evidence.
The first thirty days
The roadmap is eight weeks. The first four decide whether you finish. Twelve hours a week, split roughly two thirds reading and one third building.
| Week | Do | Done means |
|---|---|---|
| 1 | Trading Fundamentals. Position Aggregation and Delta Exposure. Start Project 2 (phases 1 to 3). | You can add up a book by hand and your code agrees with you. |
| 2 | Forward Curves. Mark-to-Market P&L. Risk Management. Finish Project 2. | Repository public, tests green, README shows the reference portfolio's P&L. |
| 3 | Trade Lifecycle and Settlement. Market Data Operations. Production Python (rebuild the nightly settle job). | Your ten-stage map is drawn. The 420k incident write-up is one page. |
| 4 | Market Microstructure. Day-Ahead and Intraday. Project 3, phases 1 to 4. Regulation and Compliance, the REMIT II chapter. | The toy exchange clears an auction. You can explain what REMIT II asks of an algo desk in three sentences. |
Weeks five to eight: finish the exchange, run the cadence bot, read Strategies, and write the build-or-buy note.
Prove it: what to put in front of a hiring manager
Not a certificate. Five things they can open.
- A trade's journey, drawn from memory. Capture to reconciliation, one page, with the three places it breaks marked.
- The position and P&L engine on GitHub. Tests green. README states the reference portfolio and the P&L it produces, and the figures match the guides.
- One incident, written up. The night the curve moved the P&L by EUR 420k with no new trades: what you would check first, second and third.
- A build-or-buy note on an ETRM. Under one page. Both sides, with numbers, and a recommendation.
- A running bot with its audit log. The
cadencebot, your own controls around it, and the log that shows what it did and why.
Take those five to an interview and the conversation changes. You are no longer asked whether you understand the desk.
What a trading technology team learns on the courses, and where to start.