Volatility is nothing new in commodity markets. But recent price shocks across oil, gas and power have highlighted a critical gap: many firms still lack the ability to accurately quantify how counterparty risk evolves under stress.
When prices move rapidly, exposure profiles change just as quickly. Trades that once looked balanced can become materially risky, and expected cashflows can shift significantly. In this environment, forward-looking metrics like PFE and XVA are essential.
Why It Matters Now
Recent market swings have reinforced a simple point - static exposure views are no longer sufficient.
- Fixed price trades can quickly move deep in or out of the money
- Floating structures can generate unexpectedly large cashflows
- Exposure becomes highly sensitive to market path and timing
To manage this, firms need to simulate how exposures evolve, not just where they stand today. As highlighted in the webinar, the ability to model and respond to this uncertainty is central to effective risk management.
PFE and XVA: Two Sides of the Same Coin
PFE and XVA are both built on simulated future exposures, but they serve different purposes:
- PFE measures potential exposure under stress and is used for credit limits and regulatory capital
- XVA captures expected losses and funding costs, directly impacting pricing and profitability
In practice, both rely on large-scale Monte Carlo simulation. The challenge is not just calculation but doing so consistently across complex commodity portfolios.
Why Commodities Are Harder to Model
Commodity derivatives introduce complexities that standard XVA frameworks are not designed to handle.
Fragmented curve construction
Commodity curves are built from futures with different tenors, units and conventions. Without a unified approach, pricing and sensitivities quickly become inconsistent.
Seasonality and scale
Energy markets are inherently seasonal. Preserving this in simulation often requires modelling dozens of individual futures contracts per curve, significantly increasing complexity.
Correlation and dimensionality
Large numbers of underlying instruments create challenges in managing correlations and maintaining stable models.
Performance constraints
Daily or hourly averaging, common in power and gas products, makes full revaluation Monte Carlo computationally intensive.
Energy Derivatives Add Another Layer
Energy trading desks face additional challenges:
- Custom delivery schedules and physical settlement
- Asian-style averaging across pricing periods
- Optionality in delivery timing, volume and location
- Multi-curve spread exposures
Even seemingly minor details, such as last trading day rules or daylight saving adjustments, can materially impact valuation and exposure.
A More Practical Approach
Most firms still calculate PFE and XVA separately. In commodities, this creates unnecessary complexity.
Both metrics rely on the same core workflow - calibration, simulation and valuation. A unified framework improves consistency, reduces duplication and better supports cross-commodity portfolios.
However, extending existing systems is rarely straightforward. As the webinar highlights, onboarding commodities into PFE/XVA frameworks is a complex undertaking, particularly given the scale and specificity of the asset class.