Engineering Answers to Property Losses
Catastrophe modelers did data even before “big data” became cool.
The vast statistical, scientific and engineering brainpower packed into their software tools allows insurance companies to churn out numbers for massive portfolios of properties and properly price their coverage, as well as allow regulators to feel confident they’re doing so.
Still, CAT models have their critics and their limitations. Some of the biggest related to engineering.
Corporate risk managers often listen to these criticisms. As they should. When it comes to the complex and unique structures in some of their catalogs of properties — think a chemical plant or a microchip fabricator — a model cannot properly paint an accurate portrait of its exposure.
“We (our brokers and us) were convinced that the traditional hurricane modeling that the industry relies upon did not accurately reflect the true measure of loss exposure we faced as a port authority.” — Danny Thompson, risk manager, Georgia Ports Authority.
And when it comes to the handful of physical plants that are driving the vast amount of a risk manager’s total cost of risk, a model will not be able to provide the loss-prevention recommendations to reduce that cost.
That’s where the engineers come in. With their notepads, tape measures and fold-up ladders — and certainly more sophisticated equipment — engineers on the ground can understand what is driving the cost of risk at a special or important physical site.
The tailored data they collect can be fed into the models to reduce uncertainty for a key property and lower the exposure numbers across an entire portfolio. The things they see can be turned into concrete suggestions for how to strengthen a property against the risk of fire, windstorm, flood, quake … you name it.
Brokers and insurers can offer these engineering services — and if you listen to some insiders, these services are nothing short of a competitive advantage — allowing them to take millions in premiums from their more model-dependent competitors.
If we talk with someone at a carrier or broker that provides risk engineering services, they will say that these services provide a distinctive advantage for everyone involved.
“It is a significant differentiator for us,” said Ryan Barber, a managing director and U.S. property practice industry specialization leader for Marsh, referring to the broker’s loss-control engineering and, in particular, its natural hazards engineering group.
As Barber explained, Marsh offers its engineering services not as an added line item to a risk manager’s bill, but as an investment. To have an engineering team visit the 10 critical locations in a risk manager’s portfolio, for instance, may cost $50,000.
But what happens if the data collected leads to a brand-new modeling output for that client — and, say, a $500,000 reduction in the average annual loss (AAL) number? A lower AAL would indicate to underwriters that they should charge that client lower rates on their property insurance.
“We have never done a [loss-control engineering] project where the cost of the project itself wasn’t covered by the reduction we saw in the AAL,” Barber said.
Better data for crucial facilities can benefit a risk manager in another way. CAT models primarily pump out two types of numbers. There’s the AAL, which tells underwriters the losses they can expect from a risk manager’s properties on average per year and how much premium needs to be charged over time to fund those expected losses.
The models also put out numbers to gauge the tail end of a risk manager’s exposure.
How much loss across the portfolio of properties can underwriters expect on a one-in-250-year return period? Or even worse, one-in-500? These numbers often tell risk managers how much property-catastrophe limits they ought to buy.
Insurance provisions in commercial loans often demand borrowers buy enough insurance to cover a 250-year return period. For this latter type of number, Barber said, clients can see a 40 to 50 percent reduction after engineers are sent in.
Before taking that step, though, risk managers should make sure they need risk engineering in the first place.
A good broker typically can advise a client on the need of site-specific data. A good broker will also crunch the analytics on a client’s portfolio and gain a sense for which few individual locations are driving the exposure, explained Charles Macaulay, chief engineer at risk engineering firm Global Risk Consultants (GRC).
Risk managers then know which critical sites to target for a boots-on-the-ground site-specific engineering study, which provides potential loss estimates similar to a model (but of course with far less uncertainty).
Such services could make Marsh stand out, as well as other brokers that have their own teams of structural engineers and property risk control specialists — or even partnerships with third-party engineering firms like GRC. But how much of a differentiator are these services?
Another way to look at that question is how much demand there is for the services. Andy Kao of AIR Worldwide, director of the modeling firm’s catastrophe risk engineering consulting arm, estimated that 90 percent of the property exposures out there are “vanilla” and do not require site-specific analysis to understand their exposure.
“It will always be fairly uncommon for them to pursue this because of the nature of what is at risk,” he said of risk managers. “For 90 percent of the stuff out there, the models do a great job.”
Macaulay estimated that as many as one in five of GRC’s clients ask the firm to collect data on a site-specific basis to feed into their models. The rest of their clientele come for the loss-control aspects, whereby engineers make suggestions that would reduce the damage if an event were to happen.
The knowledge can strengthen properties and risk managers’ resolve to reduce their amount of CAT coverage and limits.
“We (our brokers and us) were convinced that the traditional hurricane modeling that the industry relies upon did not accurately reflect the true measure of loss exposure we faced as a port authority,” said Danny Thompson, risk manager at the Georgia Ports Authority.
“The site-specific engineering study verified this. Additionally, we wanted to identify any loss control measures we could implement in order to reduce the magnitude of our wind and flood exposures. There seemed to be a greater comfort level for the underwriters based on having the site-specific engineering study to rely upon.”
Some carriers appreciate this “greater comfort level” so much that they keep risk engineering teams in-house — as is the case with FM Global, AIG, Berkshire Hathaway Specialty and Allianz.
Are these carriers pulling in millions of premiums from competitors because of their engineers? We could look at their premium underwritten numbers and guess.
But what we can say with certainty that an insurer like FM Global doesn’t offer risk engineering services just because they think it sounds attractive to a prospective client. It’s woven into the core of what FM Global does.
“From the FM Global perspective, whatever is driving [interest in site-specific engineering] hasn’t changed,” said Brion Callori, senior vice president of engineering and research.
That is, what has always been driving it is the desire to prevent losses. Better data for underwriting and modeling is a by-product, Callori said.
FM Global’s site-specific engineering can deliver insights all the way up to an insured’s C-suite to help those executives know how to better allocate capital for risk improvement and understand how risk impacts the bottom line.
Not every client is ready for the heavy-duty site-specific engineering reports at a lot of locations, Callori added. FM Global meets with clients to understand how much engineering they want across different locations and exposures.
“The level of commitment and implementation varies per company,” agreed Maarten van der Zwaag, global head of property risk consulting for Allianz Global Corporate & Specialty. “This is a key item to address when starting discussions and setting up any site-specific engineering loss control program.”
FM Global, for its part, fosters that commitment and understanding through tech-enabled tools to communicate the value of engineering at any level.
The Rhode Island-based insurer, for instance, offers a benchmarking and analytics tool called RiskMark, which scores individual locations by their comparative risk management quality. Insureds’ executives can receive the scores, and other engineering reports, through a cloud-based and now mobile portal called MyRisk.
Callori estimated that FM Global participates in the property insurance programs of one-third of the Fortune 1000. In other words, loss control to the point of site-specific engineering resonates with risk managers no matter the market conditions.
Van der Zwaag also stressed that when both an insurer and client drive the process, the collaboration tends to lead to stronger long-term bonds. That’s a bonus for carriers: client retention.
“Effective risk management programs will require commitment and investments of both client and insurer.” — Maarten van der Zwaag, global head of property risk consulting, Allianz Global Corporate & Specialty
“Effective risk management programs will require commitment and investments of both client and insurer. This will increase client retention and at the same time generate steady risk improvements year on year, resulting in a reduction of losses. This is a win-win situation.”
If it presents that “win-win” situation consistently and effectively, the Munich-based executive said, an insurance company can gain the interest of new clients and win business.
But millions of dollars in new client business? Instead of guessing, it might again be better to look at it in terms of demand, and the timing of that demand. For risk managers, the power of site-specific engineering comes to the fore when they need to prove how good their risks are during hard markets, said AIR’s Kao.
During soft markets, risk managers can shop their property programs. It’s the insurers that have to sell themselves. To win business in a competitive market, they may offer up value-added engineering services, as well as the modeling and big data reporting bells and whistles to go