SKN3X.COM

The insurance view: defining the event

The hard part is not the pricing but the question of what counts as a loss event at all — and how many policies it hits at once.

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 check whether a DeFi finding has an objectively determinable trigger.
  • You can report aggregation through shared dependencies as a risk of its own.

Check your prior knowledge

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

Core concept

Fact

Four candidates, four different triggers

A contract failure with funds leaving is unambiguously determinable on-chain: address, amount, time. A depeg is a price observation needing a threshold plus a duration, or every trading day is an event. A governance decision is a process, not a loss — the loss arises only from its effect. An oracle error is determinable only once the reference it is measured against is defined. Only the first case has a trigger that is objective without further agreement.

Risk

Aggregation is the normal case here

In classic lines individual losses are largely independent: one house fire says nothing about the next. In DeFi many risks share the same components — a stablecoin, an oracle provider, a bridge, a chain. A failure at that level produces simultaneous losses across the whole book. The load-bearing figure is therefore not loss frequency per policy but the share of the book resting on one dependency.

Uncertainty

Short history, no base rate

A calculation needs a probability of occurrence. For smart contract failures there is no reliable base rate: the population is small, young, changes with every upgrade, and assurance practice itself is still the subject of empirical research. That is no argument against cover, but an argument for carrying the assumption's uncertainty visibly in price and capacity rather than letting it disappear into a point estimate.

Interpretation

The question before the pricing

Before a risk is priced it must be settled: who determines that the case has occurred, from which source, and within what period? For on-chain determinable events the source is the chain. For everything else a named reference is required — and where none exists the risk is not uninsurable, but the cover becomes an agreement about determination rather than about loss.

Definitions

Insured event
An objectively determinable occurrence whose happening triggers the obligation to pay.
Aggregation
The accumulation of simultaneous losses from one common cause.
Capacity
The total amount a carrier is willing to write for one risk or one cause.

Model

  1. Define the event — what exactly triggers payment?

  2. Settle determination — who determines it, from what source, within what period?

  3. Check aggregation — how much of the book shares this cause?

  4. Carry the base-rate uncertainty visibly, rather than pricing it in and forgetting it

From finding to writable cover — Steps 1 and 2 decide whether anything can be written at all; step 3 decides how much.

Formulas

Aggregated exposure per cause

Exposure(cause) = Σ cover amount of all policies whose risk contains this cause
cause
stablecoin, oracle provider, bridge or chain
cover amount
the respective policy's maximum payout

Limit: The sum is as complete as the per-policy survey of dependencies. A dependency not established shows as zero and so looks like a checked absence — which is why every such listing needs a “not established” row.

Worked example

Twenty policies, one cause

Policies written
20, USD 5m cover each
Protocols covered
14 different ones
Shared oracle provider
in 11 policies
Shared stablecoin
in 16 policies

Counted by protocol: 14 risks, largest single position USD 10m. Counted by cause: stablecoin USD 80m, oracle provider USD 55m.

The same twenty policies — counted by cause, the largest simultaneous exposure is eight times what it is when counted by protocol.

Reading: The first count describes the portfolio, the second the capacity question. An underwriting guideline knowing only the first limits a quantity that does not bind when losses occur.

Retrieval

Which of the following has a trigger that is objectively determinable without further agreement?
Why is loss frequency per policy the wrong load-bearing figure here?

Exercise on real data

Read the case study as an underwriter: how many of the three positions would a failure at the shared level hit at once?

Case study: dependency chain →

Compare the guiding questions with step 3 above and record which detail you would have to survey per policy to compute the aggregation at all.

Dimension 10: dependencies →

Application

You are to draft cover against “smart contract failure”. Which four points do you settle before price is discussed?

Related case studies

Institutional reading

Insurance
Which cause currently carries the largest share of the book — and is that limit written down anywhere?
Bank
Would such cover actually transfer the risk, or merely change its addressee?
Asset management
Which part of the protocol risk would remain with the vehicle after cover?

Metrics in this lesson

Key takeaways

Evidence