Gordon Thompson, AmeRisk
12 May 2020Actuarial & underwriting

How to identify the next ‘unprecedented’ event

A highly contagious virus, and the resulting instability it brings, is spreading throughout the world. Scientists race to understand the virus and are desperate to find a cure. As the death toll mounts, societies crumble and the world panics. This is either the year 2020, or the plot of the movie “ Contagion”, released in 2011.

On the TED Talks stage in 2015, Bill Gates declared a highly infectious virus to be the current greatest risk of global catastrophe. As he outlined the ways in which we are woefully unprepared, and how our global health community should be preparing to fight this type of virus, it’s as if he were reading the headlines from earlier this year.

Was the pandemic unpredictable?

As we learn more information about COVID-19, and watch the effects on humanity, the phrase about hindsight being 2020 holds double meaning. This catastrophic event and our lack of preparation weren’t the result of an unknown risk, but the result of ignored warnings.

“The Delphi method provides a valuable recognition of potential future parameter considerations, based on experts familiar with the event under consideration.”

The predictions made by the writers of “Contagion” and Bill Gates are just two examples of a message from the past about our current world situation.

What other warnings are being unheeded by businesses? How can you identify the low frequency, high cost risks that can upend your business? More importantly, how can you mitigate/insure those risks?

This pandemic has caused supply chain disruption, loss of key employees and business interruption due to the actions of civil authorities, just to name a few. Did your captive protect you in this situation? How can you insure your assets better for the next “unprecedented event”?

Assessing high frequency, low cost risks

An enterprise risk assessment with your risk manager is a great place to start. After the risks are identified, actuaries often work with the client to model the cost associated with the identified risks.

Actuaries make use of models to forecast the cost of losses associated with economic, financial or hazard risks for events that may happen in the future. These risk models are often calibrated using past observations and attempt to capture the potential behaviour of companies and individuals.

When the risk phenomenon being observed occurs frequently, and each occurrence represents a small or limited size of loss, it is called a high frequency, low cost risk. Typically insured by commercial insurance, the cost of this risk, using the law of large numbers, is forecast using standardised actuarial models based on historical information.

Identifying and assessing low frequency, high cost risks

But what about a situation like the current pandemic? COVID-19 is a prime example of a low frequency, high cost risk situation—a 1-in-100-year event that can shut down the global economy and change the trajectory of many businesses well into the future.

One technique to estimate future model parameters is the Delphi method. This technique hinges on the use of expert opinions. While any individual expert must overcome bias, averaging several expert expectations can lead to reasonable projections and forecasts.

Actuaries were using the Delphi method as early as the 1940s to develop the 1941 annuity tables. At the time mortality rates were rapidly improving, due to advances in medical research, vaccinations, and expanded medical services. These changes needed to be taken into account. Consulting with experts, using the Delphi method, they included the impact of future medical improvements on future mortality levels for these projected tables.

The Delphi method is useful when data is immature, unavailable, or when there are recognised changes in frequency/severity due to economic, financial, or behavioural trends. While all projection methods retain a level of uncertainty, the Delphi method provides a valuable recognition of potential future parameter considerations, based on experts familiar with the event under consideration.

The origins of the Delphi method

The Delphi method has a fascinating history. It gets its name from the island of Delphi, the site of the ancient Greek oracle where Apollo was able to foresaeethe future. We no longer have mythical oracles, and the Delphi method takes the human approach, and asks experts in a variety of fields to assess what they predict will happen in the future.

The method was established at the beginning of the Cold War in the 1950s, and was designed to forecast how technology could impact warfare. The world was just emerging from two world wars that had seen the advent of flight for war, and the first use of an atom bomb. The method was used to try to predict how technology could continue to advance war, as well as how to protect our nation and arm ourselves with the future game-changing technology first.

How it works

The Delphi method is a qualitative data analysis process which relies on a panel of experts. These experts are asked for their forecasts or opinions on a topic in an open-ended format. Then, a facilitator provides an anonymised summary of the forecasts of the group and the reasons behind their opinions.

The experts are asked to consider the other expert opinions and are given the chance to review and revise their opinion on the same topic. This distillation process is repeated several times in an attempt to achieve a consensus. The process concludes when a consensus has been reached.

Current applications

The Delphi method is applied in places other than warfare, especially in areas related to public policy, economic trends, healthcare, education, and business forecasting.

Philip Tetlock, a University of Pennsylvania psychologist, used the Delphi model when he brought together a team of ‘super forecasters’ to find ways to accurately predict major events in world affairs. He found the following four situations to be key when setting up successful forecasters, which are also four key components of the Delphi method.

  1. They change their mind, frequently and in little bits: being open to change and changing your mind with new information makes you more likely to be an accurate forecaster.
  2. They work in teams: the team ecosystem, with equal rights and responsibilities for every person on the team, builds an environment where teams learn from each other, leading to more accurate predictions.
  3. They make actual predictions: forcing answers with firm predictions, and assigning probability values, results in predictions that can be proven true or false. With that comes a sense of responsibility.
  4. They are at least a little knowledgeable: to make predictions, as opposed to guesses, you need to know at least a little in the topic area.

Setting up your own Delphi team

Keeping in mind the four keys to success as outlined by Tetlock, how can you use the Delphi method to identify and prepare your captive for future risks?

At AmeRisk we regularly use our version of the Delphi method to quantify risks in new and emerging markets, specifically in situations for which no historical data exists, or when proxy data doesn’t give an accurate representation of the potential risk.

We work to identify potential risks, and assess the current coverage. The team comprises the business owner, the captive manager and other professionals working with the captive. Consulting with experts in fields outside your captive team is also recommended: this helps your team gain outside perspective and adds to the knowledge pool of the group.

It’s uncomfortable to think about all the ways in which your business could be harmed, but identifying the market risk for your business or product is necessary to insure that risk.

For example, one of our captive insureds made energy drinks. The business grew dramatically as part of the energy drink boom in the early 2000s.

Energy drinks flooded the market and became very popular with teenagers around the world. Health and government officials, however, became increasingly concerned. They were worried about the potential effects the drinks might be having on citizens.

In 2014 European countries began market-wide bans on energy drinks to alleviate these concerns. A market that was booming, bust overnight.

In hindsight this situation seems obvious. However, when the policy for this captive was written, the possibility that an entire country (or the entirety of the EU) might ban their product was unexpected.

Taking into consideration the new product in an emerging market, the captive had coverages for product recalls and government actions. When the product was banned they were able to make claims and access capital from their captive.

The insured received $4 million in loss payments for product recall and government actions. This allowed the company to recover from the shock, cut their losses, and re-focus their strategic efforts.

Applying the lessons to your own business

Start by reviewing your risk exposures and current coverage with your risk manager. Find the gaps in your coverage and devise a plan to close the gaps. Remember to insure your risk with broad, rather than narrow, coverage endorsements. Think outside the box: consider how various risk scenarios could affect your business, based on the opinions of experts.

What are some changes in the environment that may cause business losses that could be covered? For example, climate change. You may not want to hire a climatologist, but you can use the work of those experts to research the effects a changing climate may have on your business or your supply chain.

Discuss these predictions from the experts with your team and continue to work as a team to extrapolate the cost if the predicted event happens.

Preparing for an uncertain future for your business is something best done well in advance. Whether that future contains hurricanes, terror attacks, zombies or the next pandemic, a thorough risk analysis can identify your blind spots and help you secure the future of your business.

Gordon Thompson is a consultant at AmeRisk Consulting. He can be contacted at: gthompson@ameriskconsulting.com