The moment of truth
You’ve identified your persona. You understand your blockers. You know your competitors are already moving.
Now comes the part where most people stumble: actually starting.
Playbook 1: The Independent Asset Owner
Your 8-Week Path to Direct Trading Control
Current state: Paying a service provider, frustrated with lack of control.
End goal: Automated overnight imbalance management, direct P&L ownership.
Week 1-2: Backtest & Design
Document your pain points and write down 3-5 simple rules. Your first strategy should be embarrassingly simple:
IF wind forecast changes by >20% at gate closure -4 hours
THEN trade 75% of the expected shortfall/surplus
LIMIT: Never exceed 50% of installed capacity
Backtest this against 12 months of data. Calculate savings vs current service provider fees.
Common mistake: Over-complicating. If you can’t explain it in one sentence, simplify.
Week 3-4: Get Buy-In
Your 4-slide deck:
- The Problem: £X in annual service fees, no control
- The Opportunity: Backtesting shows £Y savings possible
- Risk Management: Overnight-only, position limits, service provider backup
- Timeline: Live in 4 weeks, evaluate after 3 months
Focus on cost savings with management, bounded risk with legal, better tools with operations.
Week 5-6: Go Live
Configure 2-3 rules in PowerDesk Edge with strict limits (30-50% capacity max). Start Monday night, 23:00-06:00 only. Keep service provider for daytime initially. SMS alert for every trade at first.
Check every trade the next morning: Did it work as expected? Respect limits? Beat the alternative?
Week 7-8: Expand
Add daytime rules. Gradually increase position limits. After 3 months of stable trading, transition fully from service provider (keep them on standby for 6 months).
First quarter goal: 15-25% cost reduction, zero limit breaches, full audit trail.
Playbook 2: The Established Utility Trading Desk
Your 12-Week Path to Institutional-Grade Automation
Current state: Brilliant people, compliance concerns, trader variance.
End goal: Backtested strategies live, documented decisions, data scientists building value.
Week 1-3: Win Over Compliance
Week 1 – Education session (90 minutes): Show compliance what algo trading actually means, how peers are doing it, REMIT II compliance, and audit trail advantages.
Key message: “We’re making trading more compliant and auditable, not less.”
Week 2 – Risk demonstration: Present backtesting across extreme scenarios (Beast from the East, COVID, price spikes). Show how algorithms would have performed with risk limits in every case.
Week 3 – Pilot proposal:
- Start with backtesting only
- Then 30-day pilot, one strategy, 10% normal volume
- Ex-post review (not ex-ante approval)
- Kill switch for compliance
Get them to say yes to: “Can we backtest one strategy and review together?”
Week 4-7: Build & Backtest
Choose ONE strategy:
Option A – The Variance Reducer: Identify your most predictable decision (e.g., day-ahead to intraday flattening). Document how your best traders handle it. Code as rules. Backtest against what all 12 traders actually did.
Metric: “Would have beaten 10 of 12 traders and reduced variance by X%.”
Option B – The Night Shift: Focus on 00:00-06:00, simple mean reversion, low volume.
Metric: “Would have generated £X in overnight spreads we currently miss.”
Use minimum 2 years data. Stress-test across extreme events. Document every parameter choice.
Week 8: The Review
Present to compliance, head of trading, and risk. Show strategy explanation, backtesting results vs manual trading, risk controls, and proposed pilot parameters.
Ask for: “Approve 30-day pilot with £X position limits and full monitoring.”
Week 9-11: Supervised Pilot
Week 9 (Shadow mode): Algo runs but doesn’t trade. Compare decisions to actual trader actions.
Week 10-11 (Live with training wheels): Real trades within strict limits. Daily reviews with traders, weekly with compliance.
Week 12: Scale or Kill
Decision criteria:
- Outperform manual baseline (or match with lower variance)?
- Zero major incidents?
- Compliance comfortable?
- Traders see value?
If GO: Expand strategies, increase limits, add trading hours.
Long-term: Add 2-3 strategies every 2-3 months. Each expansion uses the same rigorous process. Trust builds through consistency.
Playbook 3: The Quantitative Powerhouse
Your 16-Week Path from Zero to Trading
Current state: Strong quant capabilities, no market access.
End goal: Live trading with proven strategies, business case for full infrastructure.
Week 1-4: Parallel Workstreams
Workstream A (Operations):
- Identify intermediation partners
- Understand credit requirements
- Map regulatory requirements
- Begin conversations immediately
Workstream B (Quants): Access PowerDesk Edge, load historical data, test three simple strategies:
- Day-Ahead to Intraday Arbitrage – Test predictable spread patterns
- Intraday Mean Reversion – Simple statistical arbitrage, liquid hours
- Imbalance Price Prediction – Basic features before adding complexity
Week 4 milestone: Show IC enough promise to approve full market entry.
Week 5-8: Refine & Secure Approval
Deep dive on most promising strategy. Optimize parameters, calculate realistic costs, estimate slippage, stress-test across regimes.
Week 8 IC presentation:
- The Opportunity: £X annual revenue (3 years backtesting)
- Entry Strategy: Intermediation for 6 months, then direct
- Risk Controls: £Y daily loss, £Z monthly triggers shutdown
- Cost: £A setup, £B monthly, breaks even at £C P&L
- Timeline: Live in 8 weeks, evaluate after 6 months
Week 9-12: Build & Deploy
Week 9-10: Finalize intermediation, set up PowerDesk Edge live trading, build dashboards, configure risk limits.
Week 11: Paper trading. Validate feeds, test controls.
Week 12: Go live at 10-20% volume with strict risk limits:
- £50k-100k max daily position
- £10k daily loss stop
- Auto-shutdown after 3 losing days
Week 13-16: Scale & Prove
Progressive scaling: 50% volume (Week 13), add second strategy (Week 14), 75% volume (Week 15), full deployment (Week 16).
Month 4 decision: If profitable 3 of 4 months and live performance within 20% of backtesting, request IC approval for direct market access and expanded team.
Months 5-16: Build Your Edge
Now add complexity: ML-enhanced prediction (Months 5-8), multi-market arbitrage (Months 9-12), additional European markets (Months 13-16).
Key insight: Success in power isn’t about sophistication. It’s about starting simple, proving value quickly, and scaling methodically.
Universal Best Practices: The Supervision Framework
Level 1 (Real-time): Live dashboard, automatic alerts, kill switch, full logging
Level 2 (Daily): 15-minute stand-up reviewing previous day, flagging unexpected behavior
Level 3 (Weekly): Compare to backtesting, analyze degradation, update limits
Level 4 (Monthly): Full performance review, compliance audit, strategy approvals
The winning model: One experienced trader supervises 5-10 automated strategies, focusing on the 10% of decisions requiring human judgment while algos handle 90% of routine execution.
Common Pitfalls to Avoid
- Skipping backtesting – What works manually doesn’t always work algorithmically
- Over-engineering first strategy – Simple beats sophisticated for getting started
- Inadequate risk controls – Hard limits aren’t for normal times; they’re for 3am disasters
- Fighting the culture war – Bring traders along, let them design strategies
- Giving up too early – Every successful algo trader has losing strategies
Your Next 48 Hours
If you’re an Asset Owner:
- Book a PowerDesk Edge information demo
- Gather 12 months of forecast and generation data
- Write down 3 simple trading rules
- Calculate annual service provider costs
If you’re a Utility:
- Book a PowerDesk Edge information demo
- Schedule 90-minute compliance education session
- Identify most predictable trading decision to automate
- Calculate variance across your trading team
- Assign data scientist to lead backtesting
If you’re a Quant Fund:
- Book a PowerDesk Edge information demo
- List 3 intermediation partner candidates
- Define 3 simple strategies to backtest
- Draft IC proposal outline
Don’t Get Left Behind
Every day you wait, your competitors don’t just maintain their lead – they extend it.
- Month 1: They learn what works
- Month 3: They refine strategies
- Month 6: They deploy at scale
- Month 12: They have proprietary insights you’ll never catch up to
The firms pulling ahead in European power trading in 2027 will be the ones who started this in 2025.
The question isn’t whether you’ll automate. It’s whether you’ll lead or follow.
Ready to begin? Book a PowerDesk Edge demo and start backtesting today. It costs nothing except time, and it proves everything.
The Brady Technologies team has helped firms at every stage of this journey – from first backtest to scaled automation.
