Expected Potential Exposure (EPE)

Expected Potential Exposure (EPE) is the core quantity used to measure and manage counterparty credit exposure in derivatives. It represents the expected positive mark-to-market (MTM) of a portfolio over time, capturing both the magnitude of potential exposure and the probability of that exposure occurring.

At any future time t, exposure is defined as the positive part of the MTM:

    Exposure(t) = max(V(t), 0)

    EPE is the expectation of this quantity:

   EPE(t) = Expected value of max(V(t), 0)

This can be understood as a probability-weighted average of all future positive MTMs, where each outcome is weighted by its likelihood. Negative MTMs are excluded, as they represent exposure of the counterparty to you.

Economic Interpretation

EPE is not just a statistical construct — it has a direct economic meaning.

It is equivalent to the forward value of an option to unwind the position at zero cost, i.e. the premium associated with being exposed only when the position is in-the-money.

This gives EPE its characteristic properties:

• Exposure is one-sided
• It is convex, increasing with volatility and time
• It can be material even when current MTM is zero

Why EPE is the Correct Measure for CVA

Credit Valuation Adjustment (CVA) reflects the expected loss arising from counterparty default. Loss only occurs when:

• The counterparty defaults
• The exposure is positive
• EPE directly captures the second condition. CVA is therefore driven by the interaction of:
• EPE (exposure profile)
• Probability of default (PD)
• Loss given default (LGD)

Conceptually:

CVA is approximately the time integral of:
     EPE(t) × PD(t) × LGD

EPE is the correct exposure profile because it reflects the full distribution of future MTMs, not a point estimate, and aligns with how exposure actually materialises at default.

EPE as a Hedging Quantity

EPE is not only the basis for expected loss — it is also the quantity that determines the size and evolution of the CVA hedge.

CVA can be viewed as the cost of insuring future positive MTM against counterparty default. The exposure to be hedged is therefore the future positive exposure profile, i.e. EPE.

In practice, this is hedged using credit protection (e.g. CDS), with:

CDS notional aligned to the EPE profile (adjusted for discounting and recovery)

If the counterparty defaults at time t, the loss is proportional to the positive MTM at that time, while the CDS payoff is proportional to its notional and loss given default. Matching CDS notional to EPE aligns the hedge with expected exposure at default.

Dynamic Rebalancing and Self-Funding

EPE evolves continuously as market conditions change:

Rates, FX, volatility alter future MTM distributions

Portfolio composition changes over time

Collateral terms reshape exposure

As a result, the CDS hedge must be dynamically rebalanced.

Crucially, the same market moves that change EPE also generate offsetting P&L in the portfolio or its risk hedges. This creates a self-financing dynamic:

If exposure increases, EPE rises and CDS notional increases

• This increase is funded by the market move driving the exposure change

If exposure decreases, CDS can be reduced

• The reduction releases value back to the portfolio

Ignoring transaction costs, this means:

• The CDS hedge can be rebalanced over time so that it continuously matches the EPE profile without introducing structural funding drag.

Link to Exposure at Default

At default, the realised loss depends on the spot positive MTM of the portfolio.

By dynamically aligning CDS notional with EPE over time:

The hedge tracks the expected exposure at each point in time

By extension, it aligns with the distribution of realised exposure at default

In this sense:

• EPE is not just an average — it is the correct hedge ratio for the exposure that will exist at default.

Distinction from Peak Exposure Metrics

It is important to distinguish EPE from Peak Exposure (or Peak Potential Exposure), which is often cited as a stressed or worst-case MTM measure.

Peak exposure metrics are typically:

• Based on high-percentile scenarios (e.g. 95th or 99th percentile)
• Designed for risk limits, stress testing, and regulatory capital
• Conservative by construction

They are appropriate in contexts where exposures are not actively managed or hedged, such as buy-and-hold portfolios or institutions without an active CVA hedging function.

By contrast:

• EPE is an expectation, not a stress measure
• It reflects the average exposure conditional on probability, not tail outcomes
• It is the correct input for valuation and hedging, not for setting conservative risk limits
• Peak (or pseudo-peak) measures are often embedded in capital calculations, but they are conceptually distinct from the exposure relevant for CVA pricing and hedging.

Confusing the two leads to systematic misinterpretation of both risk and value.

Impact of Collateral and FVA

Collateralisation fundamentally alters the EPE profile:

• No CSA: full EPE profile → material CVA and FVA
• Two-way CSA: exposure reset via variation margin → EPE collapses to gap risk → CVA negligible, FVA minimal
• One-way CSA: asymmetric exposure persists → EPE remains material → both CVA and FVA are structurally significant

FVA depends on whether exposure must be funded.

Under no CSA, funding is symmetric — both positive and negative MTMs matter. The relevant profile is therefore not EPE alone, but the expected MTM profile, i.e.:

Expected value of V(t)

This can be viewed as the combination of:

• EPE (positive exposure), and
• ENE (expected negative exposure)

In this case, the option-like convexity effects largely offset, and the relevant quantity becomes the forward MTM rather than a one-sided exposure measure.

Under one-way or asymmetric CSAs:

Only one side funds exposure

The symmetry breaks

FVA becomes directly linked to the EPE profile and inherits its asymmetry and convexity

Summary

EPE is the central quantity in counterparty risk because it unifies:

• Valuation (expected loss via CVA)
• Hedging (CDS notional and rebalancing)
• Economics (option-like exposure to positive MTM)
• It captures the expected, one-sided nature of exposure, evolves consistently with market dynamics, and defines the correct hedge required to match exposure through time.
• It is distinct from peak exposure measures used in risk and capital frameworks and must not be conflated with them.
• EPE is therefore not just a risk metric — it is the bridge between pricing, funding, and hedging in derivatives portfolios.