Potential Future Exposure
Potential future exposure estimates how much a derivatives counterparty could owe you at some point down the road, not just what it owes you today.
Prerequisites: Counterparty Credit Risk, Initial Margin vs Variation Margin
If a bank sells you an interest-rate swap today, the current mark-to-market might be zero — a fair deal at inception. But five years from now, rates could have moved so far that the swap is worth $40 million in the bank's favor. If you default at that point, the bank loses whatever it's owed then, not what it was owed on day one. Potential future exposure (PFE) is the risk measure built to answer: how bad could that number get?
PFE is a forward-looking, high-percentile estimate of what a counterparty could owe you at a future date — not what they owe you now. It exists because derivatives exposure isn't fixed at trade inception; it drifts with the market.
Simulating a moving target
Unlike a loan, where the exposure is roughly the outstanding balance, a derivative's value swings with the underlying market. To estimate PFE, risk desks simulate thousands of possible future paths for the relevant market factors (rates, FX, equity prices), revalue the trade along each path, and look at the exposure — the value if positive, zero if negative, since you only lose money when the counterparty owes you — at each future date. PFE is usually reported as a high percentile (often the 95th or 99th) of that simulated exposure distribution, at a chosen horizon such as one year.
Drag the simulation and watch how the spread of possible paths widens the further out you look. Exposure behaves the same way: near-term, a swap's value can't have drifted far from today's mark, so exposure is narrow. Years out, the market has had more time to move, so the distribution of possible values — and the high percentile that defines PFE — is much wider. This is why PFE profiles for long-dated swaps typically rise, peak somewhere in the middle of the trade's life, and decline again as it approaches maturity and there's less time left for the market to move further.
Worked example
A 5-year interest-rate swap has a current value of zero. A Monte Carlo simulation of 10,000 rate paths revalues the swap at the 1-year mark under each path. Sorting the resulting values from worst (for the bank) to best, the 95th percentile of exposure — the level exceeded only 5% of the time — comes out to $12 million. That $12 million is the 1-year, 95% PFE: the bank should be prepared for the possibility that this trade owes it that much in a year, even though it's worth nothing today.
What this means in practice
PFE feeds directly into how much capital a bank must hold against a counterparty, how much collateral it demands, and whether a trade is even worth doing given the credit line available. A hedge fund that looks fine on a current-exposure basis can still be a large PFE user if its trades have long maturities or volatile underlyings — swaps and long-dated options generate far more PFE than short-dated FX forwards of similar notional, because there's simply more time for the market to move.
PFE assumes exposure moves independently of the counterparty's own credit quality. When the two are correlated — a counterparty more likely to default exactly when it owes you more — PFE understates the true risk. That specific failure mode has its own name: Wrong-Way Risk.
Related concepts
Practice in interviews
Further reading
- Gregory, The xVA Challenge (ch. 6)