Home / Wiki / Stress Testing

Risk & AnalyticsReviewed 2026-07-21

Stress Testing

Repricing a fund's current positions through a hypothetical or historical adverse scenario to estimate the resulting loss — a complement to VaR for tail risk.

Definition

Stress testing reprices a fund's current book of positions under a specific hypothetical or historical adverse scenario — a defined market shock — to estimate the resulting profit or loss, without relying on any statistical distribution assumption. Where VaR asks "how much could this portfolio lose under routine, statistically normal conditions," stress testing asks a narrower and often more concrete question: "what would this portfolio lose if this specific bad thing happened again?"

Scenarios generally fall into two families: historical scenarios replay a real past event (for example, a specific week during a known market crash) applying the actual shocks observed then to today's book; hypothetical scenarios construct a plausible but not-yet-observed shock (for example, "BTC falls 40% and a major stablecoin depegs simultaneously") to probe risks the historical record may not contain. Both approaches answer the same underlying question — what does today's book actually lose in this specific scenario — using the fund's real, current positions rather than a historical return series.

A well-built stress framework resolves exactly one effective shock per position, following a defined precedence: a direct asset-level shock where one exists, falling back to a stablecoin-depeg shock for pegged assets, then an asset-class-level shock, then a beta-to-market shock for assets without a direct or class-level figure, and finally an FX shock for non-USD-denominated positions. This precedence prevents a position from being shocked twice, or shocked using the wrong assumption because two rules could technically apply.

Why it matters

VaR is built on the assumption that markets behave statistically similarly to how they have recently behaved — which is precisely the assumption that breaks down in a genuine crisis. Stress testing exists to answer the question VaR is structurally weak at answering: what happens in the tail, in a scenario that resembles the real crises a fund's asset class has actually experienced, rather than a smooth statistical extrapolation of calmer periods.

For crypto funds in particular, the historical record already contains several genuinely severe, asset-specific crises (a major stablecoin depeg, a major exchange insolvency, a broad deleveraging cascade) that a normal-distribution VaR model would treat as effectively impossible — stress testing against those actual historical shocks keeps risk assessment grounded in events that are known to be possible because they already happened.

Building a scenario result

Each position in the current book is assigned an effective shock percentage from the scenario definition, following the shock-resolution precedence above, and the position's dollar loss is its current value multiplied by that shock. Losses (and any gains, for positions the scenario benefits) are summed across the book to produce a total NAV impact in both dollar and percentage terms.

Because stress testing reprices the fund's actual current holdings rather than replaying a historical return series, it automatically reflects the fund's present positioning — a book that has recently de-risked will show a smaller loss under the same historical scenario than the same fund would have shown before de-risking, which is exactly the point: stress results should move with real changes in exposure, not stay fixed to a stale portfolio.

Repricing a book through a historical BTC shock

A fund with $10,000,000 of NAV holds $6,000,000 in BTC and $4,000,000 in cash/stablecoins. A historical scenario modeled on a real severe BTC decline applies a direct −47% shock to BTC and leaves cash unshocked at 0%.

The loss on the BTC position is $6,000,000 × 0.47 = $2,820,000; cash contributes no loss. Total loss is $2,820,000, so the post-shock NAV is $10,000,000 − $2,820,000 = $7,180,000, a NAV impact of −$2,820,000 / $10,000,000 = −28.20%. A fund holding a smaller BTC allocation, or the same BTC allocation hedged with an offsetting short, would show a materially smaller loss under the identical scenario — which is exactly how stress testing is used to compare the resilience of different portfolio constructions against the same shock.

Position	Value	Shock	Loss
BTC	$6,000,000	−47%	−$2,820,000
Cash / stablecoins	$4,000,000	0%	$0
Total (NAV impact)	$10,000,000	—	−$2,820,000 (−28.20%)

Common mistakes

Relying on stress testing alone without also monitoring VaR and day-to-day volatility — stress tests illuminate specific tail scenarios but do not describe the routine risk profile of the book on an ordinary day.
Applying a single stress scenario as if it captures all tail risk — a fund exposed mainly to a stablecoin depeg risk gains little insight from a pure BTC-crash scenario, and vice versa; a useful stress framework runs a suite of scenarios spanning the fund's actual risk factors.
Reusing stale position data in a stress test — because the exercise reprices the current book, running it against yesterday's (or last month's) positions produces a result that no longer describes the fund's actual risk.
Ignoring correlation and liquidity effects within a scenario — a real crisis often also triggers a correlation regime shift (normally weakly correlated assets moving together) and impaired liquidity (the same shock size producing more slippage on exit than in calm markets), both of which a naive per-position shock can understate.
Treating a hypothetical scenario's plausibility as settled just because it was modeled — hypothetical scenarios should be checked for basic plausibility (for example, whether the implied move is unusually large relative to the asset's own historical range) rather than accepted uncritically.

In practice

Crypto-specific stress scenarios typically go beyond a generic market-crash template to include events unique to the asset class: a major stablecoin depegging, a major exchange or lending platform becoming insolvent, or a broad, leverage-driven deleveraging cascade across correlated altcoins — each of these has real historical precedent in crypto markets in a way that has no close equivalent in traditional finance.

Nyx Fund's Stress Lab runs the fund's live, current position book through a library of historical scenarios with real shock vectors (including a May-2021-style BTC decline, a major-stablecoin-depeg scenario, and a major-exchange-insolvency scenario), plus Monte Carlo and reverse stress tests that solve for the smallest shock size that would produce a given loss — all clearly labeled as simulated analysis rather than a live NAV or trading signal.

The free Fund Health Check does not run a scenario shock, but its gross-leverage and cash-buffer checks flag the same conditions — thin cash, elevated leverage — that determine how badly a stress scenario would actually hit the book.

Try it free →

Questions, answered

What is the difference between stress testing and VaR?

VaR estimates loss under routine, statistically normal conditions using historical volatility. Stress testing reprices the current book under a specific, defined adverse scenario — historical or hypothetical — without relying on a statistical distribution assumption, and is designed specifically to probe tail events that VaR tends to understate.

What scenarios are typically used to stress test a crypto fund?

Common scenarios include historical replays of severe crypto-specific events — a major asset price crash, a stablecoin depeg, an exchange insolvency — as well as hypothetical combined shocks, alongside broader macro scenarios like a historical equity market crash applied via factor exposures.

Does a stress test use historical returns or current positions?

A stress test applies a defined shock (either from a real historical event or a hypothetical scenario) to the fund's current, actual positions — not a historical return series. This means the same scenario produces a different result today than it would have for yesterday's portfolio if positioning has changed.

Is stress testing required for a crypto hedge fund?

There is generally no strict regulatory requirement for early-stage crypto funds, but institutional allocators increasingly expect to see stress scenarios as part of operational due diligence, since VaR and volatility alone are widely viewed as insufficient given the fat-tailed history of crypto markets.