Earthquakes and Our Moral Duty

Awful. Horrific. Beyond imagination. It is hard to describe fully in words the true destructive power of earthquakes. And recently we have seen them happening far more often than we are accustomed to in the past. In the past two years, we have witnessed earthquakes in Turkey, in Chile, in New Zealand, in Japan and in Haiti. Any way you look at it, earthquakes are grim. And someday Canada will have a big one. How prepared are we?

As an industry, we plan for the worst, but what happens in real life is often much worse than the worst-case scenario. For instance, in Japan they conservatively built structures for the worst case, an 8.0 quake on the Richter scale. However, the Tohoku quake in Japan registered 9.0 on the Richter scale and was followed by a huge tsunami. In Christchurch, New Zealand, a quake happened on a fault line that was previously unidentified. But it happened nonetheless.

So, as an industry, can we trust our models?

The financial crisis of 2008 amplified powerfully the lesson that over-reliance on models is dangerous. In that instance, the banking industry relied too heavily on models and failed to properly manage and mitigate the impact of a 1-in-100 year event – a downturn of housing prices in the United States. As an industry, insurers can be proud of how strongly our balance sheets withstood the financial crisis when other pillars of the financial services industry crumbled. But it is unwise for us to be smug. We too have our models and to rely on them exclusively would also be a grave error. Simulations: More than You Bargained For

At Zurich, we have turned to simulations to provide increased depth to the third party models we employ. In a recent “fire drill,” we recreated the repercussions of a large earthquake on the West Coast of Canada, complete with aftershocks, conflagration, civil disturbance by postal code, loss of water and power and sprinkler damage. This kind of drill is a tool to help us visualize the scope of these kinds of events.

Underwriters price and establish coverages, but what are the actual costs associated with settlement? In our “fire drills,” we have claims professionals settle the losses theoretically. In such an exercise, insurers may find they are covering dramatically more than expected. In our drill experience, in certain cases, the claims professionals determined settlement amounts up to double what the underwriters predicted. For example, in a recent drill exercise, we learned sprinkler damage could cause as much or more loss as fire following an earthquake. As a result of these exercises, Zurich has adjusted coverages, wordings, sub-limits and exclusions.

Managing Accumulations Better

First off, we all need to manage accumulations better. The Risk Browser Tool from Risk Management Solutions (RMS) Inc. is available to all insurers and it enables us to identify all risks in one geographic area in real time. As new business submissions are quoted, the tool maps out the accumulations of risk. This enables us to see how many risks are accumulating in particular spots. It is a disciplined and rigorous model and needs to be implemented with sensitivity to broker and market conditions. But during the catastrophes of the last few years, it has paid off for Zurich and helped us manage our accumulations effectively.

But are we managing quake risk properly in Canada?

It’s not clear to me that, as a market, we’re properly pricing quake risk in Canada, particularly when we think about it regionally. The models show the risk for quake in British Columbia is greater than Los Angeles for the 1-in-2,000 year earthquake event. But at least on the West Coast, in the commercial market, we are collecting some premium for it.

Although other climate perils in Canada, such as blizzards and ice storms, are more frequent and dominate the risk at a shorter return period, earthquakes are the primary tail risk for Canada. RMS models indicate the largest loss events could occur in Eastern Canada. However, in Quebec, competitive forces in the commercial market today have reduced the market price for earthquake coverage to effectively zero.

Are Balance Sheets Big Enough?

At Zurich, we manage our balance sheet to withstand a 1-in-2,000 year event. In Canada, all carriers are required to manage to an OSFI target that is climbing 10 years each year, towards an eventual 1-in-500 year requirement. But regardless of region, there are grounds to be concerned that the regulatory requirement from OSFI, with its current 1-in-390 year event provision, may not be fully adequate. That is because carriers are offering guaranteed replacement coverage or “extended replacement  values” that can gross up to more than 130% of the insured value. That increased exposure in the context of a post-quake demand surge means a theoretically unlimited exposure after a major event. This may not seem like a big deal now, but if the worst-case scenario occurs, it would be. I do not believe these increased exposures are currently reflected in the Dynamic Capital Adequacy Testing analysis done by all insurers, or in the individual composite risk scores OSFI allocates to each carrier.

Building Codes Matter

As we all know, the size of an earthquake does not necessarily predict the resultant damage and fatalities. Building codes matter. The recent earthquakes in Chile and Haiti offer excellent proof of this fact. Although the Chile earthquake was materially larger than the one in Haiti, a staggering 395 people died in Haiti for every one person who died in Chile. Santiago is a modern city with building codes and design requirements that properly take quake risk into account. Haiti was just not built for the event that it so tragically experienced. When buildings are retrofitted and built with earthquakes in mind, lives and assets are saved.

Closer to home, Montreal is an older city and, unlike Vancouver or Santiago, Chile, it has many buildings that are not retrofitted to modern building codes. I do not believe this risk is adequately factored into industry pricing today.

Our Moral Duty

The insurance industry has two moral obligations. The first is to deliver on our promise to pay. Insurers perform a wonderful service to humanity by providing a way for people to live day to day without ongoing fear of the unknown. When catastrophe strikes, insurers will be there to help. That is expected, and rightly so.

But we have a second duty to society – to properly price risk. The price of risk serves as a signal to consumers and corporations, enabling them to assess and trade off between risk and reward and to invest appropriately when required. By pricing risk properly, we send the right signals to society regarding the adequacy of building codes, the need for protective actions and the true size of the risks we face.

Underpricing quake coverage does not do Canada any favours. By doing so, the industry unintentionally opens the door for individual consumers and corporations to engage – improperly, and without fully realizing it – in riskier behaviours than they intend to. When pricing does not reflect the need to upgrade building codes or retrofit buildings, for example, individuals and corporations fail to take all the actions they need to protect their employees and their assets.

Our duty to society to get the price right matters. When the inevitable quake comes in the West or the East, I am confident our industry will respond with professionalism and compassion. By pricing properly now, we can ensure the tragic costs are less than they would otherwise be. Our duty is clear.