Convergence Point

What Is Proprietary Trading?

“Buy low, sell high.” It’s the oldest trading mantra. But in electricity markets, executing this simple idea requires navigating non-storable commodities, hourly price volatility, gate closure deadlines, and physical delivery constraints.

Proprietary trading (prop trading) is when a firm trades for its own account—seeking direct profit rather than commissions from client services. Think hedge funds, commodity trading houses, and investment bank trading desks.

In energy markets, prop traders are the pure speculators and arbitrageurs. They have no power plants to optimize, no customers to supply. Just one goal: extract profit from price movements.

Let’s explore how they do it.

The NASDAQ Definition:

The NASDAQ website defines prop trading as:

“Principal trading in which firm seeks direct gain rather than commission dollars.”

Translation: You’re not an intermediary. You’re not a broker earning fees. You’re putting your own capital at risk to make money from the market itself.

This is fundamentally different from:

• Brokers: Connect buyers and sellers, earn commissions, take no position
• Retail suppliers: Buy for customers, earn margins, minimize risk
• Generators: Sell their own production, optimize assets

Prop traders are pure market players—their only asset is information, timing, and risk capital.

The Three Objectives: Arbitrage, Speculation, Optimization

Every prop trade falls into one of three categories:

1. Arbitrage: Risk-Free Profit (Nearly Extinct)

The Dream: Simultaneously buy and sell the same product at different prices in different markets, locking in guaranteed profit.

Classic Example:

• Germany Year-2026 futures trading at €85/MWh on EEX
• Same product trading at €86/MWh on ICE Endex
• Arbitrage trade: Buy on EEX, sell on ICE, pocket €1/MWh × volume

Why It’s Mostly Dead:

• Algorithmic trading: Computers execute these trades in microseconds
• Market integration: European markets are tightly coupled (price differences vanish instantly)
• Transaction costs: Fees, bid-ask spreads, and clearing costs eat tiny margins

Modern Reality: Pure arbitrage opportunities exist for milliseconds before high-frequency trading algorithms eliminate them. Retail human traders will never see them.

Where Arbitrage Still Exists:

• Cross-border inefficiencies: Occasional price divergence between interconnected countries
• Product mismatches: Converting monthly contracts to quarterly contracts
• Illiquid products: Exotic delivery locations or tenors

But these require deep market knowledge and fast execution.

2. Speculation: Betting on Price Movements

This is where 90% of prop trading happens: taking directional bets on future prices.

Example 1: Bullish View on Summer Prices

Analysis (April 2025):

• Summer 2025 forward trading at €65/MWh
• Weather forecasts suggest hot summer → high cooling demand
• French nuclear plants scheduled for maintenance → reduced supply
• Trader believes price will rise

Trade:

• Buy 100 MW Summer-2025 at €65/MWh
• Hold through May and June
• June: Price rises to €78/MWh due to heatwave forecasts
• Sell 100 MW at €78/MWh
• Profit: (€78 – €65) × 100 MW × 2,184 hours (summer months) = €2,839,200

Risk: What if summer is mild? Price falls to €58/MWh → Loss: €1,528,800

Example 2: Bearish View on Winter

Analysis (September 2025):

• Winter 2025-26 trading at €115/MWh (historically high)
• New LNG import capacity coming online in November
• Mild winter forecasted by long-range models
• Gas storage levels at 95% (very high)

Trade:

• Sell (short) 50 MW Winter 2025-26 at €115/MWh
• Wait for price to fall
• November: Price drops to €95/MWh as forecasts confirm mild winter
• Buy back 50 MW at €95/MWh
• Profit: (€115 – €95) × 50 MW × 2,184 hours = €2,184,000

Risk: What if winter turns brutal? Polar vortex hits, price spikes to €150/MWh → Loss: €3,822,000

3. Optimization: The Gray Zone

This involves exploiting structural inefficiencies or timing market liquidity.

Examples:

• Curve arbitrage: Year contract cheaper than sum of quarterly contracts → buy year, sell quarters
• Peak vs. Base arbitrage: Peak premium too high → sell peak, buy base, profit from spread
• Calendar spread trades: Q1 vs. Q2 price differential looks wrong historically

Why “Optimization” Not “Arbitrage”? Because there’s still risk:

• Structural relationships can shift
• Unwinding positions may face liquidity constraints
• Execution costs may exceed theoretical profit

Real Trade:

• Baseload Year-2026 at €85/MWh
• Peak Year-2026 at €105/MWh
• Historical peak premium: €15/MWh
• Current premium: €20/MWh (too high)
• Trade: Sell peak, buy 1.43× base (to match MWh), capture inflated premium
• Assumption: Premium will normalize
• Risk: What if summer heatwaves make peak even more valuable?

The Balancing Act: When to Flatten Positions

Professional prop traders don’t hold positions forever. They have strict rules for position management:

Rule 1: Stop-Loss Limits

Prevent catastrophic losses by automatically exiting losing positions:

Example:

• Bought 200 MW Year-2026 at €90/MWh
• Stop-loss set at €87/MWh (€3 loss tolerance)
• Price drops to €87 → Position automatically closed
• Realized loss: €3 × 200 MW × 8,760 hours = €5,256,000

Without stop-loss: Price could continue falling to €75, turning a €5M loss into a €26M catastrophe.

Rule 2: Profit Targets

Don’t get greedy—take profits when targets hit:

Example:

• Bought at €85/MWh with €95 target
• Price reaches €94.50
• Trader debates waiting for €95
• Decision: Take profit at €94.50 (close enough)
• Next day: Price crashes to €88 on unexpected news

Lesson: Pigs get fat, hogs get slaughtered.

Rule 3: Time-Based Exits

Financial players must close positions before delivery starts:

Investment Bank Timeline:

• Trading Year-2026 contracts throughout 2025
• No later than December 31, 2025: All positions must be flat
• Cannot deliver physical power (no balancing group, no generation, no customers)

What Happens If They Don’t?

• Forced liquidation by exchange or TSO
• Penalties for non-delivery
• Regulatory sanctions
• Reputational damage

Real-World Speculation Strategy: Dynamic Stop-Loss

This is where theory meets practice. Let’s walk through a complete trade cycle with risk management:

Phase 1: Entry (March 2025)

• Analysis: Year-2026 undervalued at €82/MWh
• Historical average: €88/MWh
• Fundamentals support higher prices (coal plant retirements, carbon price rising)
• Trade: Buy 150 MW Year-2026 at €82/MWh
• Stop-loss: €79/MWh (€3 risk per MWh)
• Target: €92/MWh (€10 profit per MWh)

Phase 2: Monitoring (April-May 2025)

• Price rises to €87/MWh (up €5)
• Unrealized profit: €5 × 150 MW × 8,760 hours = €6,570,000
• Risk adjustment: Move stop-loss to €84 (breakeven +€2)
• Reason: Lock in some protection while keeping upside

Phase 3: Volatility (June 2025)

• Price spikes to €93/MWh on summer heatwave news
• Exceeds target (€92)
• Decision point: Take profit or hold?
• Trader decides: Sell 50% (75 MW) at €93, let 50% run
• Realized profit on 75 MW: (€93 – €82) × 75 MW × 8,760 hours = €7,227,000

Phase 4: Reversal (July 2025)

• Price corrects to €88/MWh (profit-taking after heatwave)
• Remaining 75 MW position still profitable
• Stop-loss now at €86 (protects €4 profit)

Phase 5: Exit (August 2025)

• Price stable at €89/MWh
• Trader closes remaining 75 MW at €89
• Profit on remaining position: (€89 – €82) × 75 MW × 8,760 hours = €4,599,000
• Total profit: €7,227,000 + €4,599,000 = €11,826,000

Key Insights:

• Disciplined profit-taking (50% at target)
• Dynamic stop-loss (moved up as price rose)
• Total risk never exceeded €3/MWh on full position
• Max potential loss: €3 × 150 MW × 8,760 hours = €3,942,000
• Actual profit: €11,826,000
• Risk/reward ratio: ~3:1

Why Pure Arbitrage Is Nearly Dead

The dream of risk-free money died with technology:

1990s Arbitrage:

• Trader notices EEX price €0.50 higher than OTC equivalent
• Calls broker: “Buy 50 MW on OTC, sell on EEX”
• Execution time: 5 minutes
• Profit: Locked in

2020s Reality:

• Computer detects €0.05 difference
• Executes 1,000 MW trade in 0.003 seconds
• Arbitrage eliminated before human sees it
• Human trader’s screen shows flat prices

What Changed:

• Algorithmic trading: Machines faster than humans
• Market integration: Prices synchronized across platforms
• Narrower bid-ask spreads: Competition drove margins to zero

Survivors:

• Institutional prop desks: With superior technology and co-location
• Niche markets: Illiquid products where automation doesn’t work
• Cross-commodity plays: Gas-to-power spreads, coal-to-power, etc.

The Prop Trading Reality Check

Let’s be honest about what prop trading actually entails:

Requirements:

• Deep capital reserves: Can you survive a €20M drawdown?
• Real-time market access: Professional trading platforms ($100k+/year)
• Risk management systems: Stop-losses, position limits, VaR calculations
• Market expertise: Years of learning price patterns and drivers
• Psychological discipline: Can you cut losses without hesitation?

Compensation:

• Successful traders at banks: €500k – €5M+ per year
• Failed traders: Fired (sometimes spectacularly)
• Average career length: 5-7 years before burnout or moving to management

Failure Modes:

• Ignoring stop-losses → catastrophic losses
• Over-leveraging → bankruptcy
• Revenge trading (trying to win back losses) → worse losses
• Failure to flatten before delivery → regulatory disaster

Key Takeaways

✓ Proprietary trading = Trading for your own account, not clients
✓ Arbitrage is mostly dead: Algorithms execute faster than humans
✓ Speculation dominates: Directional bets on price movements
✓ Stop-loss discipline is non-negotiable to prevent catastrophic losses
✓ Profit targets prevent greed: Take money off the table when targets hit
✓ Financial players must flatten all positions before delivery starts
✓ Risk/reward management separates professionals from gamblers

---

Next in Series: Post 5: Retail Procurement: The Art of Buying for Thousands of Customers