Reverse Stress Test

A reverse stress test is an analytical method that works backwards: rather than choosing a scenario and calculating the loss, you start with an unacceptable loss (or insolvency) and identify the scenarios that would cause it. This reveals hidden vulnerabilities and concentration risks that forward-looking stress tests might miss.

How reverse stress testing works

Step 1: Define the unacceptable outcome. This is usually specified by regulators or management. Examples:

• Portfolio loss exceeds 20% in one month.
• Trading losses exceed $500 million.
• Equity capital is reduced to zero.
• The institution becomes insolvent.

Step 2: Identify the vulnerabilities. What exposures, if they move adversely, would cause the outcome? Work through the portfolio:

• Which positions lose the most if rates move?
• Which lose the most if spreads widen?
• Which are concentrated (few counterparties)?
• Which are illiquid?

Step 3: Determine the scenarios causing the loss. Work backwards. “To lose $500M, what needs to happen?” Example:

• Stock portfolio loses 30% = -$150M (if 50% of capital is stocks).
• Bond portfolio loses $200M = rates rise 200+ bps or credit spreads spike 200+ bps.
• Derivative positions lose $150M = volatility spikes 50 points.
• Total: -$500M.

Step 4: Assess plausibility. Is this scenario plausible? A 30% stock loss has happened (1987, 2008, 2020). A 200 bps spread widening has happened (2008). So yes, plausible.

Step 5: Risk management response. If the scenario is plausible and would destroy the firm, reduce exposures, hedge, or increase capital.

Why reverse stress testing is valuable

Forward stress testing asks: “Given this scenario, what is the loss?”

Reverse stress testing asks: “What scenario would kill us?”

The second question forces focus on tail risks and vulnerabilities that might be buried in a large portfolio.

Example: A bank’s forward stress test says “In a 50% stock market decline, we lose $2 billion.” The bank thinks this is acceptable. But a reverse stress test asks: “What if a major counterparty defaults, triggering a $5 billion loss in derivatives, while spreads widen 300 bps, causing another $3 billion bond loss?” Total: $8 billion, wiping out capital. The reverse test reveals a catastrophic but plausible scenario the forward test missed.

Regulatory use

The US Federal Reserve and other regulators now mandate reverse stress testing for systemically important banks. The requirement is to demonstrate that the bank has identified plausible scenarios that would cause significant losses and has plans to manage them.

Banks must report: “Here are the top 5 scenarios that would cause unacceptable losses. Here is what we are doing to reduce exposure or hedge them.”

This forces explicit thinking about tail risks and concentration.

Constructing reverse stress tests

Start with the unacceptable outcome and trace backwards:

Institution: Large global bank with $50B capital. Unacceptable outcome: Loss of $10B (20% of capital). Vulnerabilities identified:

• Trading positions in credit derivatives: $2B unrealised gains.
• Exposure to emerging market bonds: $8B notional.
• Counterparty exposure to large hedge fund: $5B.
• Liquidity buffer: $5B in liquid securities.

Scenarios causing the loss:

1. Credit crisis: EM bond spreads spike 300 bps → $2.4B loss. Credit derivative position loses 50% → $1B loss. Hedge fund counterparty defaults → $3B loss. Total: $6.4B.

2. Systemic shock: Credit spreads spike 500 bps → $4B loss. Counterparty defaults → $3B loss. Liquidity crisis; asset sales at fire-sale prices → $3B loss. Total: $10B.

3. Contagion: Major hedge fund fails → counterparty loss of $5B. Panic in markets; correlated losses across desk → $5B. Total: $10B.

Assessment: All three scenarios are plausible (they occurred, or similar, in the past). The bank is vulnerable.

Risk management response:

• Reduce EM exposure to $4B.
• Reduce counterparty exposure to $2B.
• Reduce leverage on trading desk.
• Increase liquid buffers to $10B.

Limitations and challenges

Defining “unacceptable.” Who decides what loss is unacceptable? Regulators say 20% for banks; private firms might tolerate 10%. There is subjectivity.

Plausibility assessment. Determining whether a scenario is plausible is inherently subjective. A black swan is, by definition, not foreseeable.

Scenario completeness. Even with reverse stress testing, there is a risk of missing scenarios. The 2008 crisis combined multiple failures (credit, liquidity, systemic) in ways few anticipated.

Model dependency. Calculating losses in scenarios still requires models, which have model-risk.

Despite these, reverse stress testing is a powerful tool because it forces explicit vulnerability assessment.

Reverse stress testing in portfolio management

Individual investors can use simplified reverse stress testing:

• Unacceptable outcome: “I cannot afford to lose more than 20% of my portfolio in one year.”
• Scenarios: “What would cause a 20% loss?” Stock market down 30%, bonds down 5%, alternatives down 10%.
• Vulnerabilities: “I am 70% stocks. A 30% stock decline causes 21% portfolio loss. That is unacceptable.”
• Response: “Reduce stocks to 50%, increase bonds to 50%.”

This is simpler than bank-level reverse testing, but the logic is identical.

See also