Skip to main content

Smart Curves

Create with AI, build with precision

The Forward Curve Challenge

The Challenge

Forward curves are fundamental to energy and commodity trading—essential for pricing deals, valuing portfolios, managing risk, and making strategic decisions. Yet building accurate, consistent curves remains surprisingly labor-intensive:

  • Manual Spreadsheet Workflows: Traders and analysts spend hours daily copying data from multiple sources, applying interpolation rules, handling gaps, and adjusting for market conventions. A single curve might require data from spot markets, futures exchanges, and broker quotes—each with different structures and timing.
  • Inconsistency Across Teams: Different desks build the same curve differently. One trader uses linear interpolation, another prefers cubic splines. Marketing uses yesterday's curve while trading updated theirs this morning. Portfolio valuation and risk reporting diverge because they're based on different curve versions.
  • Error-Prone Processes: Manual curve construction introduces mistakes—missed contract months, incorrect date alignments, formula errors in spreadsheets. These errors cascade through valuations, hedging decisions, and regulatory reports.
  • Time-Sensitive Operations: Markets move fast. When a major price publication hits or breaking news impacts fundamentals, curves need immediate updates. Manual processes can't keep pace.
  • Complex Market Structures: Energy markets have unique characteristics—seasonality patterns, delivery periods, locational basis differentials, calendar vs. contract month structures. Generic curve tools don't understand these nuances.

Smart Curves

OpenDataDSL's Smart Curves transforms forward curve construction from manual drudgery into intelligent automation. Smart curves in-built curve intelligence and market understanding help you build better curves:

  • Market Convention Awareness: Smart Curves understands how different energy markets work.
  • Automatic Gap Handling: Markets don't always provide complete forward coverage. Smart Curves intelligently fills gaps using configurable methods.
  • Seasonality Recognition: Energy markets exhibit strong seasonal patterns. Smart Curves automatically detects and applies.
  • Multiple Input Sources: A single curve might combine:
    • Exchange-traded futures for near-term liquidity
    • Broker quotes for illiquid tenors
    • Assessed prices from price reporting agencies
    • Proprietary models for far-forward periods

Smart Curves merges these sources intelligently, prioritizing based on liquidity, recency, and reliability.

The system builds accurate, consistent curves automatically—understanding market conventions, handling gaps intelligently, and ensuring your entire organization works from the same validated curves.

Automation Meets Expertise

Key Benefits

Transform forward curve construction from a time-consuming, error-prone manual bottleneck into intelligent, consistent, automated infrastructure—freeing your analysts from mechanical work to focus on strategic market analysis and profitable trading decisions.

Automated Curve Construction in Minutes, Not Hours

Build accurate forward curves automatically using intelligent methodologies.

What previously required 3-4 hours of manual spreadsheet work—collecting data from multiple sources, applying interpolation rules, handling gaps, validating results—now happens in minutes. Smart Curves understands market conventions, applies appropriate techniques, and delivers production-ready curves without manual intervention.

Intelligent Gap Handling and Market Convention Awareness

No more manual workarounds for missing data or market complexities.

Smart Curves automatically handles incomplete forward coverage using configurable interpolation methods, recognizes seasonal patterns, understands delivery period structures, and applies commodity-specific conventions. Power curves with hourly granularity, natural gas with calendar month structures, oil with contract vs. delivery distinctions—all handled natively and correctly.

Consistency Across Your Entire Organization

Eliminate valuation disputes caused by different curve-building approaches.

Everyone works from the same validated curves built using the same approved methodology. No more discrepancies between trading desk curves and risk management curves, or between morning valuations and afternoon reports. One source of truth for all forward pricing across all teams and systems.

Real-Time Updates with Full Version Control

Curves rebuild automatically when market data changes—with complete audit trails.

When exchange settlements publish or broker quotes update, dependent curves refresh immediately without manual intervention. Every curve update creates a versioned snapshot, providing complete transparency into what curve was used for any historical valuation, trade, or risk calculation—critical for compliance and trade reconstruction.

Flexible Methodologies: Code or Conversation

Build curves using ODSL code for precision or Fusion AI for convenience.

Sophisticated users encode proprietary curve-building logic in ODSL scripts—custom interpolation methods, multi-source blending, fundamental models. Non-technical users build curves conversationally through Fusion's Curve Assistant: "Build a Brent curve using ICE futures for 24 months, then extend using IEA forecasts through 2030." Both approaches deliver production-quality results.

Platform Integration for Automated Workflows

Curves power your entire operation seamlessly.

AI Agents rebuild curves on schedule and monitor for significant moves. Portfolio valuation tools use curves automatically for mark-to-market calculations. Risk systems access curves via API for real-time exposure analysis. Custom analytics reference curves with simple syntax. External systems receive curve updates through automated integrations—all without manual exports or data transfers.

Curve Creation

Using ODSL code

Example timespread function

		function timespread(input)
			// Bootstrap the input curve to create 
			//an arbitrage-free monthly curve
			bootstrapped = bootstrapCurve(input)
			 							
			// Create a new curve to store the spread values
			spread = Curve(input.ondate)
			 							
			// Get all contracts from the bootstrapped curve
			contracts = bootstrapped.contracts
			 							
			// Get the total number of contracts
			contractSize = contracts.size()
			 
			// Iterate through contracts
			for i = 2 to contractSize
				// Get the current contract
				current = contracts[i - 1]
				 
				// Get the previous contract
				previous = contracts[i - 2]
				 
				// Calculate the spread value as the difference 
				// between consecutive contracts
				spreadValue = current.value - previous.value
				 
				// Create the spread tenor label (e.g., "M01-M02")
				spreadTenor = previous.tenor + "-" + current.tenor
				 
				// Add the spread contract to the curve
				spread.add(spreadTenor, spreadValue)
			next
			 
			// Return the spread curve
			timespread = spread
		end

Using Fusion AI

Can you create an odsl function that performs a timespread, e.g. M01-M02
Answer from assistant: Code

Perfect! ✅ The script has been validated successfully. Here's the timeSpread function:

timeSpread Function

function timeSpread(input)
bootstrapped = bootstrapCurve(input)
spread = Curve(input.ondate)
 
contracts = bootstrapped.contracts
contractSize = contracts.size()
 
for i = 2 to contractSize
current = contracts[i - 1]
future = contracts[i - 2]
 
spreadValue = future.value - current.value
spreadTenor = current.tenor + "-" + future.tenor
 
spread.add(spreadTenor, spreadValue)
next
 
timeSpread = spread
end

What it does:

Calculates month spreads (month-over-month spreads) from a forward curve. It compares prices that are 1 month apart (e.g., M01 vs M02, M02 vs M03, etc.).


Function Signature:

timeSpread(input)

Parameters:

  • input (Curve) - A forward curve to calculate spreads from

Example Usage:

// Load a crude oil curve
crude = ${data:"ICE.IFEU.B:SETTLE:2024-01-15"}
 
// Calculate spreads
spreads = timeSpread(crude)
 
// Print results
print spreads
// Output: Curve with M01-M102: 2.50, M02-M03: 1.75, M03-M04: 0.50, etc.

Integration with Platform Capabilities

Smart Curves don't exist in isolation—they're integral to OpenDataDSL's complete platform

Data Catalogue

Curves automatically pull from catalogued market data. When you build a curve, you reference standardized datasets, not raw vendor feeds.

Smart Loaders

Curves rebuild automatically when underlying data updates. A new futures settlement triggers immediate curve recalculation.

AI Agents

Automate curve construction workflows—daily updates, exception monitoring, publication to downstream systems.

From Manual to Intelligent

Smart Curves transforms forward curve construction from a repetitive, error-prone manual task into an intelligent, automated process. Whether you prefer the precision of ODSL code or the convenience of Fusion's conversational AI, OpenDataDSL provides the tools to build accurate curves efficiently.

Your team spends less time on mechanical curve construction and more time on what matters—understanding market dynamics, identifying opportunities, and executing profitable trades.

Ready to Transform Your Curve Operations?

Join energy and commodity professionals who are already leveraging OpenDataDSL and Fusion to work smarter, faster, and more profitably.

Get Started TodayODSL Language  

How about a demo?

Our team is here to find the right solution for you, contact us to see this in action.

info@opendatadsl.com

+44 1245 555053

Sign up to our newsletters!

Fill out your details below and somebody will be in contact with you very shortly.

We'll never share your email address
Enter information about your area of interest, include your telephone number if you would like us to call you.