Reverse Stress Testing

Study guide: Exam 6C

Reverse Stress Testing starts from failure and works backwards: instead of asking "what happens if X occurs?", it asks "what would have to occur for this insurer to become non-viable?" It is a required element of FCT and a standard tool in ORSA, and its value is that it finds vulnerabilities that forward scenarios — anchored on the risks management already has in mind — systematically miss.

• Non-viability, not merely a bad year. The endpoint is the point at which the business model fails: capital below the minimum, an inability to meet obligations, or a loss of market access from which the insurer cannot recover.
• Why it finds different things. Forward stress tests test the risks management chose to test. A reverse test starts from an outcome and searches over all combinations that could produce it, so it surfaces the exposure nobody was watching — a single reinsurer, a single distribution channel, a single jurisdiction's regulatory decision.
• The output is a description of failure modes, and it is qualitative as much as quantitative. Naming the combination is often more valuable than pricing its probability, because the response — a limit, a diversification requirement, a contingency plan — follows from the naming.
• Plausibility is assessed afterwards. The exercise deliberately does not begin with a probability constraint; a scenario is first identified as a failure mode and then judged for how plausible it is. Filtering for plausibility first defeats the purpose.
• It feeds the internal target and the recovery plan. If the identified failure mode is credible, the insurer must either hold capital against it, reduce the exposure that creates it, or have a documented plan for responding to it.
• The technique is used across financial regulation — banking supervisors and Solvency II both require it — and the Canadian application is through FCT's requirement that the actuary consider what would render the insurer non-viable.

Working Backwards From Failure An insurer with a 185\% MCT ratio, capital available of \$370 million and a base solvency buffer of \$200 million asks what would make it non-viable.

Conduct the reverse stress test.

Answer The target. Non-viability requires the ratio below 100\%, so capital available below \$200 million — a loss of \$170 million after tax, or roughly \$233 million pre-tax. (The buffer would also move, generally upward in stress, so this is conservative in the insurer's favour.)

What combinations reach \$233 million?

• Catastrophe plus reinsurer failure. A 1-in-250 event costing \$310 million gross, with \$270 million reinsured — of which 60\% sits with one reinsurer who fails. Retained: \$40M plus \$162M uncollected = \$202M, plus reinstatement costs. Close to sufficient on its own.
• Reserve deterioration plus rate suppression. A 12\% deficiency on \$900 million of liabilities is \$108 million, and three years of suppressed rates at a 107\% combined ratio on \$500 million of premium adds roughly \$105 million. Together, sufficient.
• Investment collapse alone. With a bond-dominated portfolio, reaching \$233 million requires a market event more severe than 2008. Not sufficient alone — which is itself a useful finding.

What the exercise reveals. The failure modes are reinsurer concentration and the combination of reserve error with rate suppression — neither of which is the risk an investment committee spends its time on, and the first of which the MCT does not charge for at all.

The actions that follow:

• Cap any single reinsurer's share of the catastrophe programme, and require collateral or registered status.
• Set the internal target above the level implied by the second combination, not merely above the supervisory target.
• Build the first combination into the FCT adverse scenarios so it is tested annually rather than once.

The general point: the reverse test converted an abstract 185\% ratio into two named, addressable exposures. That translation is what makes it worth doing.