Building scenarios instead of quoting metrics

Four components without which a stress scenario carries no decision — and why the reverse question is the most useful one.

This lesson has had no expert review. It was written for this platform and against the evidence it cites; nobody has gone through it independently.

Learning objectives

You can build a scenario from trigger, path, duration and outcome measure.
You can formulate a reverse scenario: what would have to happen for the limit to break?

Check your prior knowledge

Answer these for yourself before reading on. Wherever you hesitate is where this lesson pays off.

What separates a scenario from a metric?
Why is a historical crash not automatically the right scenario?
Which number does a committee need in order to decide?

Core concept

Fact

Four components, otherwise it is just a number

A usable scenario names the trigger, the transmission path, the duration and the outcome measure. Without the path it is an assertion; without the duration liquidity cannot be judged; without the outcome measure nobody knows what to respond to. “Minus 30 percent” alone satisfies none of the four.

Interpretation

Historical and hypothetical answer different questions

A historical scenario replays an observed path; it is evidenced and usually too mild, because the market has grown and has different dependencies since. A hypothetical scenario is built on the path that exists today; it is plausible, not evidenced. Both are legitimate — they just have to stay separately labeled, so nobody attributes one's evidential status to the other.

Interpretation

The reverse question is the most useful one

Rather than asking “what happens at minus 30 percent”, one can ask: which price decline, over what duration, would break the internal limit? The answer is a single figure, it is checkable, and it is immediately usable for a committee — because it asserts no probability but names a threshold.

Definitions

Transmission path: The sequence by which a trigger reaches your own position.
Reverse stress test: Asking for the event that would just break a defined limit.

Model

Trigger — what happens first?
Path — how does it reach your position?
Duration — over what period?
Outcome measure — capital affected, not an estimated loss

A scenario in four parts — The historical-or-hypothetical label belongs on the scenario, not in a footnote.

Formulas

Threshold from the reverse question

required decline = 1 − (limit / current position value)

limit: internal value below which action is required

Limit: Computed without exit price impact and without second-round effects. Both shrink the permissible decline — the figure is therefore an upper bound, not a reserve.

Worked example

The same position, two questions

Question A: “What happens at −30 %?”
Answer A: a loss figure with unknown probability
Question B: “Which decline breaks the limit?”
Answer B: −18 % over five days, including exit price impact

A yields a figure that cannot be placed without a probability. B yields a threshold that can be held against observed volatility and against earlier moves.

The same position, but only the second answer can be checked.

Reading: B requires no forecast. It moves the discussion from “how likely is that” to “has a decline of this size already happened in this market” — and that is an answerable question.

Interactive model

The worked example, with movable figures. Estimate first what happens — then check.

Health factor and liquidation price

Move the collateral price and watch the distance to the HF = 1 threshold.

Estimate first, then check

Before you move the slider: answer the example's question B for this position — what price decline breaks the HF = 1 boundary? Then check whether a decline of that size has occurred in a market like ETH before.

The starting values are the worked example's own figures — change one input at a time.

Collateral price change

Collateral units

Starting price per unit (USD)

Loan incl. interest (USD)

Liquidation threshold

What the model does not show: The boundary here is one position's liquidation threshold. A portfolio boundary built from loss tolerance and exit costs comes earlier, and price impact at exit is absent from this model altogether.

Retrieval

A probability of occurrence.A checkable threshold instead of a loss figure with no reference.Less computation.

Because historical data is imprecise.Because the market has grown since, with different dependencies than at the time of the observed episode.Because it contains no duration.

Exercise on real data

Find a market and check how its sigma sits against the threshold from the reverse question. Record what a sigma from a short history does not say about it.

Sigma as a yardstick for a threshold →

Application

Formulate a reverse stress test for a position in a lending market.

Related case studies

Institutional reading

Bank: Is the threshold from the reverse test written down anywhere as a trigger for action?
Insurance: Which path, not which figure, carries the calculation?
Asset management: Who watches whether the market is approaching the threshold?

Metrics in this lesson

Key takeaways

Trigger, path, duration, outcome measure — without all four it is a number, not a scenario.
Historical is evidenced and usually too mild; hypothetical is plausible and unevidenced.
The reverse question yields a checkable threshold with no claim about probability.