Which trading desk are you?
The big question:
“We know algos are the future, but where do we even start?”
The reality:
Every successful automation journey begins with understanding exactly where you are today; and having clarity about what’s holding you back.
Today’s mission:
Meet three real archetypes from the market, understand their unique challenges, and see which path forward makes sense for you.
Introduction: same destination, different starting points
In our conversations with trading desks across the UK power market, we’ve seen that while everyone wants to capture the benefits of algorithmic trading, the path to get there varies dramatically based on who you are, what resources you have, and the ways algo trading can help you operationally.
Today, we’re pulling back the curtain on three distinct personas we see repeatedly in the market. These aren’t theoretical personas; they’re based on real conversations with real traders who are navigating this transformation right now.
Persona 1: the independent asset owner
Who they are
You own 2-3 wind farms. Your assets are performing well, but you’re stuck in a trading arrangement that made sense five years ago but feels increasingly outdated. You’re likely paying a utility or service provider to handle your imbalance trading. They charge a fixed fee, place trades on your behalf, and aim to keep you “within a pound of market average.”
Current reality: When your wind forecast changes at 2am, someone else is making the trading decision. You pay the same fee whether they capture value or miss opportunities, often leaving money on the table.
The frustration points
“We’re leaving money on the table” Your service provider doesn’t have the same incentive you do to excel. They’ll meet their contractual obligation to stay within a range of the market average price, but if there’s an opportunity to make you an extra £5/MWh? Why would they bother? They get paid the same either way.
“We have no visibility or control” You see the trades after they happen, not the decision-making process. When market conditions create opportunities, you’re dependent on someone else’s overnight shift trader, who might be managing multiple clients simultaneously.
“The economics don’t work for 24/7 coverage” You’ve done the maths. Hiring even a small shift team would eat up more than your entire trading margin. But paying fixed fees to leave money on the table doesn’t feel great either.
The opportunity
Start with simple rules-based trading for overnight periods. No shift team required.
Your path forward:
- Start with backtesting to understand what’s possible with your specific assets
- Deploy simple RBT rules for overnight and daytime imbalance management
- Gradually expand rule complexity as you build confidence
- Keep your service provider for complex situations initially, then phase them out
Why independent asset owners win with automation
→ Immediate cost savings from reduced outsourcing fees
→ Direct P&L ownership means you capture all the upside
→ Start simple with just overnight hours, expand gradually
→ No hiring required — the rules run themselves
Persona 2: the established utility trading desk
Who they are
You have hundreds of megawatts under management, a 24/7 trading desk, and, critically, a room full of very clever data scientists who joined your company to do cutting-edge work but spend their days building Excel reports.
Current reality: You know automation is the future. Your competitors are already moving. But every conversation about algos triggers alarm bells in your compliance department.
The frustration points
“Compliance thinks we’re building Skynet” Your compliance team’s first question about any automation: “Can we add a button that asks for human confirmation before every trade?” They genuinely worry that algorithms will “end the world” – or at least end their careers if something goes wrong.
“We have variance across our shift team” Your 12-person shift team includes stars who consistently beat market benchmarks and others who… don’t. But their decisions aren’t recorded in detail; you just see the trades at the end. This variance is costing you real money.
“Our data scientists are getting restless” You hired brilliant quantitative minds from top universities. They came to work on interesting problems. Instead, they’re manually updating dashboards. How long before they leave for a hedge fund that lets them actually build algos?
The opportunity
Use backtesting as your compliance security blanket, then progress methodically through the automation stages.
Your path forward:
- Lead with backtesting; show compliance bounded risk across thousands of scenarios
- Start RBT with tight limits and extensive monitoring
- Document everything; create an audit trail that makes manual trading look risky by comparison
- Graduate to full algos only after proving success at each stage
The key insight: Your compliance team fears the unknown. Backtesting makes the unknown known. When you can show them exactly how an algo would have performed across the 2018 ‘beast from the east’, the Texas freeze, and every other black swan event, their fear transforms into cautious optimism.
Why established utility trading desks win with automation
→ Reduce variance; one well-tested algo outperforms your bottom 50% of traders
→ Create defensible decisions; every trade has a documented rationale
→ Unlock your talent; let data scientists build value, not reports
→ Scale without hiring; manage 10x the volume with the same headcount
The Sainsbury’s principle: As Chris Regan puts it: “You don’t eliminate humans – you have one person supervising ten automated checkouts instead of ten people scanning groceries. That’s how you keep compliance comfortable while scaling operations.”
Persona 3: the quantitative powerhouse
Who they are
You’re a hedge fund or prop shop with serious technical firepower. Your team speaks fluent Python, has already built successful algos in other markets, and sees UK and other European power markets as inefficient and ripe for the picking. You want to quickly spin up new strategies and test new these without needing to manage bulky backtesting infrastructure or having to manage clunky exchange data feeds.
Current reality: You could build a sophisticated trading system tomorrow. But building the robust backtesting infrastructure and staying up to date with the new data structures the exchange asks you for in order to be able to execute algorithmic trades? That’s what’s slowing you down.
The frustration points
“Exchange connectivity is more complex than the algos” Getting the technical setup right to easily connect to and maintain connectivity to the exchange stops you from spending more time on iterating your algos.
“We need to prove the opportunity before going all-in” Your investment committee wants to see returns before approving full market entry. But how do you generate returns without first building out the complex backtesting environment you’d need their approval to get set up in the first place? Classic chicken-and-egg.
“Speed to market matters” Every month you spend navigating bureaucracy is a month your competitors are capturing alpha. In a market this volatile, six months of delays could mean millions in missed opportunities.
The opportunity
Start simple to prove value, then scale rapidly with your technical advantages.
Your path forward:
- Use a proven backtesting environment initially to start proving your algos quickly
- Begin with simple arbitrage; day-ahead to intraday spreads
- Prove the business case with 3-6 months of returns
- Then ramp up with IC approval and unleash more of your quants
Progressive complexity:
- Month 1-3: Basic arbitrage with rule-based execution
- Month 4-6: Add ML-powered price prediction
- Month 7+: Full algo deployment with sub-second execution
Why quantitative trading houses win with automation
→ Leverage existing expertise; your Python/ML skills transfer directly
→ Exploit market inefficiencies; spend time on building out your IP, not the adjacent infrastructure
→ Scale rapidly; go from zero to major player in 12 months
→ No legacy constraints; build the optimal system from day one
Start where you are
The beauty of the backtesting → RBT → full algo progression is that it works for everyone:
- Asset owners can start with simple overnight rules and build from there
- Utilities can use backtesting to satisfy compliance before risking capital
- Quant funds can prove the opportunity to then get the resource to scale up
Your next steps
- Identify which persona most closely matches your situation
- Be honest about your specific blockers (we’ll tackle these in Episode 3)
- Start backtesting now; whatever your situation, testing costs nothing and proves everything
Remember: The question isn’t whether to automate, it’s how to do it in a way that fits your unique situation. Every company trading power successfully with algos started exactly where you are now. They just took the first step.
Coming in Episode 3: the blockers
Now that you know which persona fits, we’ll dive deep into the specific obstacles you’ll face – and more importantly, how firms just like yours have overcome them. From compliance concerns to technical barriers to market access challenges, we’ll show you the playbook for breaking through.
Spoiler alert: The biggest blocker isn’t technology, regulation, or even compliance. It’s something much simpler – and entirely within your control.
Follow Brady on LinkedIn to catch Episode 3 the moment it drops.
