logo image

This blog was originally published on InsurTech Gateway by Hambro Perks, click here for the original blog.


It is a fascinating time to work in the risk analytics business.

Traditional risks are changing, with much of this change being driven by technology. From the challenges posed by autonomous vehicles to the rapid digitization of the “smart home”, with automatic detection of threats such as fire and theft, systems are getting smarter and risks are changing.

Other types of traditional risk however still offer tremendous opportunities — last year’s storms in the U.S. have shown that even in one of the world’s most established insurance markets, uninsured losses are still a major problem. Barely half of the losses from Harvey, Irma and Maria were insured — the rest lies with the uninsured victims of the disaster, or if they are lucky, with the federal government who will help them rebuild.

This “protection gap” between those who have adequate insurance and those who do not, represents both a huge societal challenge, and a massive opportunity for the insurance market.

For many retail consumers, insurers have been progressively moving away from what has historically been a customer experience characterized by infrequent and adversarial interaction, toward becoming trusted advisors; tailoring products to individual customer needs and offering advice and resources to avoid damage. Insurers are moving from a traditional role of indemnification to one of shared responsibility.

Technology is also having an influence on how insurers respond to consumers. This can easily be seen around the way that insurance is bought. You are no longer asked whether your front door lock complies with a specific lock specification or standard (for readers in the U.K, the BS3621, for instance), you just select your door lock from one of the images shown. It is all about the customer now, and making their experience better, letting consumers see directly how much their premium might fall if they fit a better lock. It benefits everyone to increase resilience.

SB lock

Will insurers instruct locksmiths to fit better locks ahead of any burglary? Image: Pixabay

So too in the broader property and casualty (P&C) industry, insurers and reinsurers are looking for ways to deliver more valuable products. The notion that insurers settle claims, but also help reduce the potential for claims, is a powerful one. At RMS, we provide the analytics that insurers use to price risk. We are increasingly seeing our analytics used to understand the benefits that can be had by taking action to reduce risk.

Reducing the likelihood of claims is of fundamental benefit to everyone, and we often talk about the concept of the “resilience dividend” — the conceptual reduction in premium that can result from taking a measure that reduces risk.

This is a hot topic in the public sector right now — governments, both national and local, are striving to allocate budgets most effectively, getting maximum “bang for their buck.” The aim is to reduce risk whilst offering financial protection against disaster, therefore providing a dual benefit to insurance protection. We know for example that in low-middle income countries, that for every US$1 spent in pre-disaster financing (insurance), at least US$4 in disaster relief funding is saved. But more than this — that US$1 insurance spend could actually be channeled into resilience financing — providing both risk reduction and post-event recovery funding.

If your insurance protection can also be designed to reduce risk in the first place, surely that is best for everyone. Just as the “little black box” in the car encourages safer driving to reduce premium, so too the resilience and risk reduction can be structured into huge public financing decisions. If we retrofit buildings against earthquake, we reduce risk, and ultimately premium should fall. The virtuous circle is completed in that these savings can then go towards the retrofitting expense.

Within this approach, InsureTech firms are helping to drive change. Analytics is moving from two-stage risk assessment — beforehand to price, afterwards to settle claims — to a full risk life cycle assessment. This involves responding to changes in risk as preventative measures are taken and the situation on the ground evolves. For example, a parametric approach to flood insurance as offered by FloodFlash enables a consumer to choose exactly where their flood insurance protections kick in — whether that is at a low-level flood that disrupts business, or a higher level of water that causes more physical damage. That protection need is also likely to vary as insureds take mitigation measures, which could range from physical infrastructure such as sand bags or flood barriers, or to moving the contents to the second floor when it starts to look bad, through to fitting electrical points higher up a wall to avoid flood damage.

The analytics needed to support these sorts of ideas in the property catastrophe world are highly sophisticated, to understand the impact of an individual flood risk protection measure, we need detailed, highly dependable analytics. Modeling U.S. or Europe flood requires very high-resolution data, across domains of thousands of miles. The range of possibilities also requires incredibly dense stochastic event sets to understand tail risk. The result is a computational demand that can only be met by big data solutions, deployed at scale in the cloud.

Accessibility to these analytics is on the rise, and quickly. New start-ups can get the benefits of incredibly sophisticated risk analytics with a web browser or a few API calls.

These improvements in efficiency help drive the affordability of insurance, spurring competition, and ultimately driving down cost for the insured. Insurance is an industry where there is ample room for improvement — in today’s London insurance market for example, some 40 percent of the cost of insurance goes to covering business costs and administrative burden, rather than covering risk.

And so this symbiosis continues — insurance products (and channels) evolve, supported by evermore sophisticated analytics, but also driving the demand for these analytics, and pushing the boundaries at every step.

Will insurers one day send someone round to fit new locks before a break-in rather than after?  Maybe. Will locksmiths provide bundled insurance, disrupting the insurance value chain delivering protection alongside physical product? Potentially. Will insurers embrace the opportunity and partner with the experts in their field to figure out the most effective risk reduction measures and deliver this value to consumers? Definitely — and it is already happening.

Consumers and the insurance market as a whole, are demanding higher value products, greater convenience, and greater resilience. Insurance buyers want more impact from their purchasing power. Technology, and especially the InsureTech sector, is pushing the boundaries of what is possible. It really is a fascinating time to work in the risk analytics business.

Ben Brookes

You May Also Like
Ben Brookes BIA award
July 15, 2019
Spread the Word on Building Resilience for Vulnerable Communities
February 24, 2016
Clearing the Path for Catastrophe Bond Issuance

Cat bond efficiency has come a long way in the last decade. The premature grey hair and portly reflection that peers back at me in the mirror serves as a reminder of a time when even the simplest deals seemed to take months of work.  A whole thriving food delivery industry grew up in the City of London just to keep us fed and watered back when success was measured on capacity to work a 120-hour week, as much as on quantitative ability. Much has changed since then. Of course, complex ground-breaking deals still take a monumental amount of effort to place successfully—just ask anyone who’s been involved with Metrocat, PennUnion or Bosphorus, and they’ll tell you it’s a very intensive process. But there’s little doubt that deal issuance has streamlined remarkably. It is now feasible to get a simple deal done in a matter of a few short weeks, and the market knows what to expect in the way of portfolio disclosure and risk analysis information. Indeed, collateralized reinsurance trades have pushed things further, removing some of the more complex structural obstacles to get risk into insurance linked securities (ILS) portfolios efficiently. This week, I was on a panel at the Securities Industry and Financial Markets Association (SIFMA) Insurance and Risk Linked Securities Conference, discussing the ways in which the efficiency of the cat bond risk analysis could be further streamlined. This topic comes up a lot—a risk analysis can be one of the largest costs associated with a transaction (behind the structuring fees!), and certainly a major component of the time and effort involved. If there’s one aspect we can all agree on, I suspect it’s the importance of understanding the risk in a deal, and how that deal might behave in different catastrophic scenarios. Commoditizing the risk analysis into a cookie-cutter view of a few well-known metrics is not the way to go—every portfolio is unique, and requires detailed, bespoke understanding if you’re to include it in a well performing ILS portfolio. Going further, it is often suggested that the risk analysis could be removed from cat bonds—indeed, there’s no other asset class out there where the deal documents themselves contain an expertized risk analysis. Investors are increasingly sophisticated—many can now consume reinsurance submissions and have the infrastructure to analyze these in-house. The argument goes, why not let the investors do the risk analysis, and take it out of the deal—that way the deal can be issued more efficiently. One deal—Compass re II—has tested this hypothesis via the Rewire platform, and successfully placed with a tight spread. Compass was parametric—this meant that disclosure was complete. The index was fully described, so investors (or their chosen modeling consultancy) could easily generate a view of risk for the deal.  This would not have been so straightforward for an indemnity deal—here, as an investor, you’d probably want to know the detailed contents of the portfolio in order to run catastrophe models appropriately. Aggregates won’t cut it if you don’t have a risk analysis.  So, for this to work with indemnity deals, disclosure would have to increase significantly. An indemnity deal with no risk analysis would also open up the question of interpretation—even if all the detailed data were to be shared, how should the inuring reinsurance structures be interpreted? This can be one of the most time consuming elements of even the simplest indemnity deals.  Passing this task on to the market rather than providing the risk analysis in the deal would inevitably lead to a change in the dynamic of deal marketing—suddenly investors would be competing more and more on the speed of their internal quoting process, and be required to develop large modeling infrastructure, far larger than most ILS funds currently have access to today.  Inevitably this would take longer and lead to a more uncertain marketing process.  Inevitably it must load cost into the system, which might well be passed back to issuers by way of spread or to end investors by way of management fees. Or both. Suddenly the cost saving in the bond structure doesn’t look as attractive. I believe there’s a better alternative—and it’s already starting to happen. Increasingly, we are being engaged by potential deal sponsors much earlier in their planning process, often before they’ve even contemplated potential cat bond structures in detail. In this paradigm, the risk analysis can be largely done and dusted before the bond issuance process begins—of course, it’s fine-tuned throughout the discussions relating to bond structures, layers and triggers etc. But the bulk of the work is done, and the deal can happen efficiently, knowing precisely how the underlying risk will look as the deal comes together. This leads to much more effective bond execution, but doesn’t open up the many challenges associated with risk analysis removal. Detailed understanding of risk, delivered in the bond documentation, but with analysis performed ahead of the deal timeline. Perhaps the catastrophe bond analysts of the future won’t have to suffer the ignominy of receiving Grecian 2000 for their 30th birthdays. Ben and the RMS capital markets team will be talking more about innovation in the ILS market at Exceedance 2016– sign up today to join us in Miami…

Ben Brookes
Ben Brookes
Managing Director, Capital and Resilience Solutions, RMS

As managing director of the RMS capital and resilience solutions group, Ben oversees the advisory function for insurance-linked solutions (ILS) transactions, resilient financing and corporate catastrophe risk modeling. Ben also manages RMS product solutions for portfolio modeling and underwriting in the ILS space.

During his 15 years at RMS, Ben has worked on more than 30 catastrophe bond projects, including the design and development of indices to securitize new perils, and the continued work to improve and streamline transaction efficiency and disclosure. Recent efforts have been focused on driving the use of catastrophe analytics beyond the insurance market, into mainstream asset classes, the public sector, and corporate spaces.

Ben holds a master's in engineering mathematics from the University of Bristol.

cta image

Need Help Managing Your Portfolio?

close button
Overlay Image
Video Title

Thank You

You’ll be contacted by an RMS specialist shortly.

RMS.com uses cookies to improve your experience and analyze site usage. Read Cookie Policy or click I understand.