Using a Financial Impact Analysis to Define Risk Tolerance


Many of the world’s leading corporations do not have an explicit and agreed definition of their risk tolerance or appetite. We facilitate and guide many clients in this area. An important tool for us to start a conversation about tolerance and appetite is the financial impact analysis (FIA). An illustration is shown below.

The approach is simple. We use the FIA model we have developed to demonstrate how growing levels of unbudgeted financial losses would affect key financial measures for the client.

In our work with one recent client, executives indicated at the time that a $ 100 million negative event in a given year would be the limit of their appetite. We then worked with the client to design insurance programs and risk management activities to ensure that a loss of $100 million or worse is a sufficiently remote possibility.

In the client FIA example below, the financial metrics turn red (become unacceptable) when an unbudgeted loss exceeds $25 million in a given year. In this case, the company decided that a 5% decline in key metrics would be unacceptable, and this is the appetite we have used. The parameters and the loss scenarios can be changed in real time so the client can test sensitivities; this assists conversations among executives and risk managers about what the tolerances are and, once tolerances are set, helps define what executives’ risk appetite is.


Financial Impact Analysis chart We have developed a proprietary model that defines exactly what set of insurance program structures (per event and aggregate deductibles, premium, limits) yields the lowest risk for a given cost of risk for a company given its specific history and risk appetite.

Again, risk is defined in terms of ‘intolerable downsides’ to a company’s most important financial measure or measures—for example, cash flow, EBITDA, or ROIC. Using the example of a company that wants to avoid a $100 million impact to cash flow in a given year, our modelling can tell what insurance program structures are most efficient in protecting them against that downside. In this way, we demonstrate the value of insurance programs as a hedge to key financial measures.

The cost of risk is expressed as the premium plus the expected losses incurred for a given insurance strategy (set of insurance programs). The efficient frontier is the set of these insurance strategies (program alternatives), with which you cannot decrease risk without increasing costs, and for which you cannot decrease cost without increasing risk. A screen shot of our model is shown below.

Cost-Benefit Analysis of Risk Strategies

Efficient Frontier of Risk Strategies  Each red dot in the cloud above represents a unique set of insurance programs generated by our model. The yellow point represents the client’s current program. As you can see, their current program costs about $ 163 million in terms of premiums and the expected losses for these programs in a given year. Reading across to the y axis, their cash flow can be expected to be reduced by $100 million in insurable losses at a 0.5% probability (1 in 200). Let’s walk through a recent client example:

Path to Optimization

In the example above, the company’s current insurance program (the red square) costs $47 million and the downside to its most important measure, EBITDA, is $90 million in insurable losses at a one-in-200-year interval, given their current program. We modelled the value added from each layer of each line of insurance and then demonstrated how ‘switching off’ specific layers—that is, removing them—would add value either by reducing cost without increasing risk, or decreasing risk without increasing cost, or reducing both cost and risk. We demonstrated the following:

  • Primary layers for WC, GL and AL were not contributing to protecting EBITDA at the 0.5% probability—the 1 in 200 chance of occurrence. In the diagram, you can see that exiting these layers reduced costs by nearly $3 million annually without increasing risk
  • By adding an excess layer of D&O, risk was reduced greatly (by about $ 30 million at the 0.5% probability level) and the cost of risk actually declined by buying this insurance. The reason is that the layer is priced inefficiently in the market—expected losses actually exceed the premium.
  • Exiting a number of other liability layers reduced cost by another $2.5 million annually while only marginally increasing risk.

At the end of the work, we demonstrated nearly $ 7 million in expected cash flow savings plus a significant reduction in risk to EBITDA, from $90 million to $60 million when looking at the 0.5% probability case.

About Phil Ellis

Phil is Global Head of Strategic Risk Consulting for Willis' Risk & Analytics. He works with large and complex …
Categories: Analytics | Tags: , , , ,

Leave a Reply

Your email address will not be published. Required fields are marked *