​The International Society of Catastrophe Managers (ISCM) was established in March 2006. One of its  aims is to promote and develop catastrophe and exposure management  within the insurance industry. The London arm of the ISCM gathered on Friday, 28 June 2019 in the Lloyd’s Old Library to discuss the challenges presented to our community with regards to managing and modelling cyber underwriting risk.
 
The afternoon was organised in two parts:  first we heard an  underwriter's perspective on the challenges of underwriting cyber, by Helen Gemmell from Brit Insurance, followed by a panel of distinguished market experts moderated by David Clouston from Pelagius.
 
 
Part I - The Underwriter's Perspective
Helen Gemmell (Senior Underwriter, Cyber, Brit Insurance)
 
Having worked previously as a catastrophe exposure manager Helen was well-placed to address the audience and offered comments on cyber coverages and market trends as well as her views on cyber models.  The learning points from this session can be grouped into three categories: market trends, data issues and exposure management challenges.
 
Market trends
 
Coverage is offered for both first party and third party in the event of malicious and non-malicious attacks - the latter are accidental losses or system errors which can be costly and difficult to identify
 
Typically, insurance would pay for all the legal and regulatory requirements (including fines), the extortion costs and any business interruption. Should the need arise, it would also pay for IT forensics and legal defences.
 
The ever-evolving nature of the risk and the current competitive market, however, is leading to broader coverage. For example, damage to hardware (including single devices, a practice called 'bricking') which used to be excluded, has become commonly offered since NotPetya.
 
Another example is the current pressure on a standard infrastructure exclusion, which applies to all losses arising from electrical failure, water or gas leakage. Removing, or even just narrowing, this type of exclusion poses significant challenges for the industry, including pricing and aggregations across lines of business.
 
Further, undermining  the industry’s ability to build robust models is the  inconsistency  with contract wordings. Multiple  syndicates and broking firms are using their own wordings - each with different endorsements, which makes it very difficult to  capture the many variations with the terms and conditions offered.
 
Data issues
 
Data is largely incomplete. Some proposal forms are missing basic information like data volumes, patching policies or disaster recovery plans. And even when we do have data we need to question it. A client might have a business continuity plan, but this plan might itself rely on data stored on a specific device (like a laptop) which may not be accessible during a cyber event. Quite often these flaws, which seem obvious, are not picked up until a breach actually happens (e.g. zero day vulnerabilities), making it easier, ironically, to rate a company with a history of cyber incidents than one without.
 
As industry practitioners, we  need to define the minimum amount of information required to model cyber. This threshold might vary according to the size of the company. Data requirements should also vary depending on the aspect of the risk that we are modelling. Lately ransomware attacks have been the largest driver of claims, while previously it was data breaches. Ransomware attacks exploit software vulnerabilities, therefore understanding software aggregations becomes particularly important, e.g.  understanding how much a company relies on Microsoft with respect to Linux or others is critical data that is not commonly available.
 
Exposure management challenges
 
In summary: the risk is evolving, coverage is complex, wordings are inconsistent, and clients are unpredictable. Against this backdrop building a robust modelling framework looks unrealistic, but models provide guidance for underwriting and can be effective at portfolio accumulation management. We need to do more to understand and communicate uncertainty, and underwriters have a responsibility to educate exposure managers on wordings, coverages and similar.
 
In many cases, deterministic modelling does not need to be very technical, as a simple scenario can give rise to an appropriate stress test .  When looking at tail risk, that is low frequency and high severity events, probabilistic models can be complemented by scenarios and every other tool available. Clear risk management processes that are regularly reported are essential to assessing the risk as an underwriter.
 
Conclusion
 
The market is still growing significantly, both in terms of new clients seeking new policies and in terms of existing clients buying significantly more limit. However, there is a time lag. Given the complexities of the risk it might take a year to complete the proposal form and another year to underwrite, adding negotiations and sign off it can take up to three years to set up a new policy.
 
Rates are still largely driven by market forces but the race to the bottom has recently shown signs of stabilisation, and loss ratios are still attractive despite growing steadily. Despite all that, some insurance companies are bound to be caught unprepared by a  systematic event  such as a cloud service provider failing for a sustained period of time.  These events hold the potential for significant market aggregation.
 
 
Part II - An Expert Industry Panel
 
Moderator: David Clouston (Pelagius)
Mark Christensen (Head of Catastrophe Management @ Chubb Overseas General)
David Singh (Head of Exposure and Portfolio Management @ MS Amlin)
Domenico del Re (Director @ PWC)
Justyna Pikinska (Cyber Pricing Actuary @ Capsicum Re)
Jamie Pocock (Cyber Analytics @ Guy Carpenter)
 
The panel discussion opened with the question of human agency, and whether this specific feature of the risk allows the development of catastrophe models in the same way as natural perils do. There was some consensus on the need to learn to walk before we can run, in that cyber risk is evolving very quickly and it is still emerging (we have not seen the worst of it – will we ever?). Against this backdrop both actuarial techniques (which learn from the past) and catastrophe modelling techniques (which rely on accepted studies of the risk) have material limitations.
 
There are a few things we can learn from past experience in cyber: recent events result from human actions or operational technology failure, regulation is tightening and fines are becoming more severe . At the same time, we need to be conscious of the ever-evolving nature of cyber risk: change affects behaviours and the nature of claims. Although a few years ago the main cause of claims was loss of data, more recently it has become ransomware. Following the attack on the city of Baltimore, where the ransom was not paid, governments of cities under attack chose to pay, creating the conditions for hackers to follow with more attacks. This has clearly increased the frequency of events, which in turn can lead to a spike in annual severity.
 
But dealing with human agency is not new for our community, we see this when modelling flood for instance, and dealing with an ever-evolving peril is also not uncharted territory. For example, our hurricane models are getting better at capturing the storm surge element which in turn is becoming more severe because of climate change. Exposure managers need to think about the assets first, whether tangible or intangible, and develop ways to assess possible damage, in other words going back to basics before trying to push the envelope with probabilistic models. And that is because the potential for very severe losses is real, even more so considering insurance penetration is growing. In 2018 the global economic losses for cyber were three times the size of natural catastrophe losses.
 
On the topic of human agency and anthropomorphic risk we heard from Stephen Burr of Pool Re, in the audience. He described the process for issuing the first ILS product backed by an in-house developed model. The main feature of the Pool Re model is that it is credible, and such reliability  has allowed them to have a conversation with ILS partners. The frequency part was the most challenging to look at, requiring a partnership with academic institutions.
 
From terrorism the conversation veered towards State-sponsored attacks. There was a general question about whether the models differentiate between State-sponsored or “lone wolf” attacks, but more generally the panel agreed there is uncertainty around attribution and considerations would have to be made around targeted attacks versus collateral damage.
 
What about non-malicious losses? A coding error or other omission can cause serious damage, and policies should cover these events as well. These situations are not financially motivated and might as such be more difficult to capture within a catastrophe modelling framework. But even admitting that we are not able to capture all the nuances of the peril, the panel agreed that models have been and remain a very useful "measuring stick" to help understanding the relative risk.
 
The question of data came up next, with the observation there is no consistency in the underwriting community about rating factors. This poses a challenge in that there is no consensus about which aspects of the peril are more or less risky. It is clear that we need data standards, and maybe some level of education, to develop standard ILF curves or other rating tools. The panel went as far as suggesting there could be a market-wide initiative, an open modelling framework to agree on a common language around cyber.
 
There is, however, an immediate need to meet regulatory requirements and quantify cyber risk accurately for solvency calculation purposes, and at the moment each regulated entity is defining their own way of doing this. There are several approaches in use, from deterministic scenarios to frequency / severity modelling.
 
Such variety is evidence that the market is making an effort to understand and quantify the risk, and the panel expressed optimism that these efforts were converging towards a common understanding.
 
The discussion concluded with an analogy with the evolution of catastrophe models. Considering catastrophe modelling as a discipline was born in the mid-1980s, each panellist was asked to determine where cyber models are today. The consensus was that we are around the mid-1990s, but  on also suggested that the evolution of cyber models is going to be a lot quicker than natural hazards models, with one year in the cyber space equivalent to maybe 7 years in the natural hazards space.
 
At this pace, it will be only a few years before cyber catastrophe models could be as developed as current natural hazards ones – and rightly so.

End-of-year renewals behind us, January seemed the perfect time for a casual networking get-together of the Zurich ISCM crowd. This time we met in a location along the local Limmat river. But obviously, it was much too cold to indulge in pretty views outside, we stayed firmly inside and huddled together around the bar. A few (unanticipated) company events took place the same evening, thus the turn-out was somewhat lower as in prior years. Nevertheless, as ISCM members greeted old colleagues and made new acquaintances, the place soon teemed with lively discussions. The big topics this year…? Two ISCM members came straight from their very last working day, after their company had restructured (away) the cat unit. So clearly, the question of what all the recent M&A activity and continuous financial industry cost-cutting would mean for the cat risk management community in Zurich was high on the discussion agenda. Of course, renewals observations and the cat market developments in general were on people’s mind as well. In summary, the annual networking event once again proved a good opportunity to bring the ISCM community in Zurich together and bring alive the value of our membership. A small crowd of devoted members stayed all the way to the end, when we got into a disappointing argument with the bar owner over the terms of the agreement we made, and thus over the final bill. Well, no matter who was right or wrong, we all swore that we’d never set foot into that bar again!

Special thanks for organizing this event go to Amaryllis and Lysiane, as well as to Annemarie (by now our unofficial event photographer…).
 

The ISCM along, with the Institute and Faculty of Actuaries (IFoA) and Lloyd’s of London, recently hosted a panel discussion on the impact of hurricane clustering on the (re)insurance industry. A packed Lloyd’s Library listened intently as industry representatives, academia, researchers and catastrophe modeling company experts brought the audience up to date on the latest thinking on this important and topical subject. The panellists, Dimitris Papachristou from the Prudential Regulation Authority (PRA) of the Bank of England, Junaid Seria (SCOR), Nick Barter (Chaucer), Susanne Kolwinksi-Ward (AIR), Ivan Kuhnel (CoreLogic) and Steve Jewson (RMS), with frequent interaction from an engaged audience, discussed and debated for two hours if anything has changed in our understanding of the impacts of clustering, in light of the events of 2017 and 2018.
 
The topic for the session, “Hurricane Clustering in the North Atlantic: A Discussion”, was introduced and expertly moderated by Emma Watkins, from Lloyd’s. The panellists were given a number of questions to consider including:

• Does clustering exist?
• Is it linked to specific climate factors or cycles?
• Is the impact of clustering material?
• Should we consider including it explicitly on catastrophe models?

 
Each of the panellists approached the questions from a different angle, as one would expect given their different roles in the catastrophe management world: some focussed on hazard, some on loss impacts while others on whether the hazard presented by clustering is changing in a significant way. This enhanced the quality of the presentations, allowing the presenters to go “deep” on their questions, encouraging further discussions by panellists and audience alike, adding to the overall experience! The research conducted by the PRA indicates that the frequency of hurricanes in the past two years does represent a significant shift from clustering we would expect to see in the basin over time. AIR also indicated that the change in hazard was indeed statistically significant, based on their research, on a U.S. country-wide basis, for major (category 3-5 on the Saffir-Simpson scale) hurricanes. RMS stated that, considering wind only, also on a U.S.-wide basis, the impact of clustering on loss was not material, and was not needed for consideration of reinstatements.
 
There was a lot of discussion of the application of clustering in an insurance industry context. All the modelers implement clustering, allowing it to flow through to their financial models – AIR and CoreLogic use the negative binomial distribution to represent clustering in their models; RMS uses the poisson distribution for the non-clustered view, and a combination of negative binomials for the clustered view. Data availability and reliability was a common theme throughout the session – all agreed we simply need more data, with the resulting uncertainty not having changed much since clustering was first seriously considered in 2004 and 2005. The industry representatives, from SCOR and Chaucer, both stated that collaboration between researchers, modelers and business was critical to progress our understanding. Further research is also needed to understand whether any changes in clustering are “real” and would have a material impact on losses in the future.
 
The session also featured an update on the partnership between the ISCM and iCAS (Casualty Actuarial Society) to develop a joint program of credentialization for people working in the catastrophe management industry. As we wrapped the discussion up in the Library, a group of attendees adjourned to continue the discussion at One Under Lime, outside on a chilly and blustery night in London. All agreed that it was extremely valuable to be brought up to date with the latest perspectives on clustering, and that we had not heard the last of the clustering debate!

Academia should learn more about what the needs of the insurance/re-insurance industry are so that we can do a better job serving you. I look forward to other opportunities to learn more and maybe also show some of what we have been working on that could be of interest.  If offered again next year, I will definitely plan to attend it again
- Paolo Gardoni, Professor - University of Illinois at Urbana-Champaign

Thanks again for organizing the conference last month. Overall, I thought it was very good.  Dr. Gensini’s presentation was the highlight, and I really liked the approach of the morning presentations breaking the models into the parts of hazard vs. vulnerability vs. financial modeling. I thought all of those were good presentations as well. The presentation on the ISCM/iCAS credential generated a fair amount of discussion between me and my two staff members that attended the conference, so that update was useful.  Well done on the topics that were chosen and I hope you consider organizing another Chicago event next year!
- Derek Berget, ERM Modeling Director - American Family Insurance

I am glad to hear the ISCM is considering Chicago again.  I thought all sessions were good, especially the keynote by Dr. Gensini.  I also enjoyed having the round table discussion and wish it could have been longer – There really are some big topics we need to discuss about the changes of the Insurance Industry and in CAT modeling.  I would like to see a more detailed presentation on what others do in pre-event and post-event projections and reserving.  I am looking for new ways we can increase our effectiveness with the tools at hand.
- Zach Antle, Senior Reinsurance Catastrophe Risk Analyst - EMC Insurance Companies

 I wanted to tell you all how impressed I was with the NYC educational seminar, the venue, and the people.  Over 40 years of attending conferences to date the RAA cat school is in my mind the best and as such the benchmark to which all others should be compared.  My favorite RAA seminars were the very early ones because the size was more intimate.  The smaller numbers gave everyone a chance to interact with each other as well as the presenters.  The early events also lacked the pressures of client meetings, other distractions, and typically people learned more.  This event brought back memories of what the RAA Conference first set out to be.  Even as an old man I found each session to be more open, educational, interactive than any other seminar that I can recall.  This is not to distract from the RAA cat school which is on another level with a different and more robust agenda, it is just to say how impressive this event was.  Personally I like the intimate settings better than the gigantic settings since it feels more comfortable to engage.  I would encourage the ISCM to do more of these and see your niche as one that no one else occupies!  You really taught others!
- Andy Castaldi, ISCM Past President

I very much enjoyed the NY Education Event last month.  Thank you to the ISCM for the work you do for seminars like this.  With me being new in the field, it is great to hear from people with such expertise and experience.  I can look to them and see all possibilities in Cat Modeling.  Yet it also is great that I get to meet people around my age, in similar points of their careers, and share experiences.  I thought the seminar covered a variety of topics.  My favorite part was the roundtable session.  It was informative to have an open session with a back-and-forth style conversation.  I think it would be beneficial to have more sessions like that.  Specifically, an information-type session, where people in the earlier part of their career are able to talk with people who have been in the field for years.  We would be able to ask questions related to career advice, how to become more involved and the future of the field.
- Megan Royek, Reinsurance Analyst - The Toa Reinsurance Company of America

The quality of the speakers and their presentations were very good in the sessions I attended.  the presenters effectively broke down complex topics into ways a newbie could easily digest.  I liked that I saw students in attendance.  When I attended my first seminar in 2016, I believe I was the only student. Everyone that I met there was already employed.  This is a great networking opportunity and was happy to see more students there.  Attending these seminars are beneficial.  Overall, I think the experience was totally worthwhile. I look forward to attending future events and learning more about CAT Modeling and Re/Insurance. I’m especially interested in the ISCM / iCAS designation that will be coming in the near future!  I’ll be taking advantage of that opportunity ASAP.
- Albert Betancourt - American Family Insurance

This year's ISCM conference was all about tech innovation and how it can be applied to catastrophe management. Hosted by SwissRe in the brand new SwissReNext building, the conference had the futuristic setting that fit the occasion.

Peter Zimmerli, the head behind ISCM Switzerland, gave a warm welcome speech to the about 60 Catastrophe Specialists of various companies. The first topic of this afternoon was “Blockchain” and its (potential) applications in catastrophe insurance. Tobias Noack (Etherisc) presented examples where the technology is already used, e.g. in a flight delay insurance product. Scott Beckermayer (Allianz) explained the ecosystem that a blockchain can build and how this could make captive insurance administration more efficient. Michael Stahel (LGT Capital Partners) introduced his ideas on how the whole insurance value chain could be simplified from the client to the reinsurer.

The second part was dedicated to “Data Analytics and Machine Learning”. Jason Futers (Insuredata) kicked off the topic by introducing his company’s exposure enhancement capabilities. David Fox (Geospatial Insights) presented possibilities around post-disaster event response using drones or satellites. Grazia Frontoso (Google) showed interesting research around machine learning on damaged car images. As last presenter of the afternoon, Loris Foresti from MeteoSwiss demonstrated machine learning approaches used for example in precipitation forecasting.

In panel discussions following each of these blocks the Zurich ISCM members actively participated with their questions, doubts or ideas around the topics presented. Zurich showed itself from its best side when the last item on the agenda was ticked off: a marvellous sun set provided the backdrop as the 2018 ISCM conference came to an end with the sound of clicking glasses on the roof terrace of SwissReNext.
​
A great “Thank you” goes to our speakers and their commitment to support our event. Furthermore, we received greatly valued support in event organisation, catering and panel moderation from this year’s host Swiss Re. And last but certainly not least: the event would not have been possible without a truly fantastic ISCM Zurich organisation team.
 

On 31st May over 100 cat modelling / exposure management professionals came together for a networking event in London. The venue was the Slug and Lettuce on St Mary Axe, opposite the world famous ‘Gherkin’ at the heart of the global insurance industry.
​
Amongst those in attendance were personnel from underwriters, brokers, reinsurers, vendor modellers, investment funds, regulators and the Lloyd’s market. Those present ranged from experienced individuals, to new market entrants enjoying their first opportunity to get to know colleagues in the wider insurance market.   Brian Owens - a current ISCM Board Member from RMS. was present to join in the discussion.

Discussion topics were diverse, from conference debriefs to the upcoming Hurricane season, new cat models releases, emerging risks and even the return period of an England victory at the 2018 football World Cup (history suggests a 1/20, with tournaments held every 4 years we could see an England victory this century).
​
The event was sponsored by the ISCM alongside Ariel Re, Neon UW, Sirius IMA and Lloyd’s, and it was made possible by the organisational efforts of Giovanni Maccioni, David Ryan and Natalie Van Eck.
​
Everyone agreed that an event of this type was long overdue in London, arguably the city with the largest community of catastrophe modellers in the world. We intend not to wait so long for the next get together, with plans underway to hold another informal event in London at the end of September.

- N. Van Eck / D Ryan / G. Maccioni

​On the 10th of April 2018, a good crowd of Zurich region ISCM members met at the bar "Le Philosophe", cozily located on a small courtyard in Zurich's historic old town.  So any particular topic we were philosophizing about…?

The answer: Yes and No.  First and foremost the event offered a casual way for ISCM members to meet, chat and network.  It was great to see a couple of new faces among the crowd of many well-known colleagues.  The small bar was packed and judging from the noise level, there were myriads of things to discuss.  Apart from people catching up on personal matters, one could overhear a number of (cat) industry specific topics: mergers & acquisitions, general cost pressure, the 2017 cat events, ILS and cat bond capital, machine learning for risk assessment, … who knows, it may well be that one or the other of these topics will land on the agenda of the upcoming ISCM Zurich half-day conference in September 2018.

The casual get-together for drinks would not have been possible without the voluntary commitment of the organizing team (Annemarie Buettner, Michael Gloor, Lysiane Mayoraz, Simona Esposito, Peter Zimmerli) as well as the ISCM Board support – many thanks!  We will continue to try to foster exchange and support the development of the cat risk management community in Zurich and beyond.

Peter Zimmerli, Swiss Re
​

I think this is a good, well-written letter focused on objectively looking at the NFIP issues that really need to be addressed at a legislative level, particularly the impact on property owners, the role of government and industry, and the local economy in many parts of the US.  

Bob Miccolis, FCAS, MAAA, FCA
Chair, Leadership Advisory Council
The CAS Institute (iCAS)
thecasinstitute.org

Personal contact info:
Mobile/Cell: +1 610.745.5126
RMiccolis@thecasinstitute.org  

​Do you sometimes struggle with the vast number of options in today's natural catastrophe models? Do you feel that that they are getting ever more complex? If so, you're not alone, as our second ISCM conference in Zurich "Are we over-engineering nat cat models?" showed.

This year, the event was hosted by Partner Re at their premises on Zurich's lake shore. A crowd of close to 100 re/insurance professionals from the greater Zurich area attended in anticipation of controversial discussions between all relevant stakeholders: insurers, reinsurers (including non-traditional capital markets players), brokers, modelling companies and associated organisations all shared their views on this hot topic.

Erik Ruettener and Manuel Prechtl from conference host Partner Re launched the afternoon event with their welcome address and opening remarks. Then the heat turned up perceptibly, as Beach Associate's Rick Thomas– not known for holding back with clear statements – took the stage to present his view on why cat models (or at least parts of them) were indeed over-engineered. He was followed by an equally adept presentation from RMS's Robert Muir-Wood, outlining thoughts on where cat models do have leeway for further improvements.

These opening statements successfully laid the grounds for the subsequent panel discussion. The ISCM Zurich event organising committee had secured a diverse set of senior leaders with a total cat risk experience  of well over a 100 years: Milan Simic from AIR, Joerg Steffensen from Hannover Re, Matthew Eagle from Guy Carpenter, Sibylle Steimen from Allianz and Eduard Held from PERILS. Paul Della-Marta brilliantly mastered the tough job of keeping the discussion on track, as well as ensuring lively participation from the audience.  And what was the overall sentiment on the question of over-engineered nat cat models at the end of this afternoon? Well, it was a tie… the spontaneous poll taken after the panel finished showed an almost equal split into a pro and a con camp. Closing remarks by Kirsten Mitchell-Wallace of Scor and Peter Zimmerli of Swiss Re rounded off an interesting exchange amongst nat cat professionals.

As well as enjoying the content of the conference, participants appreciated the opportunity to network. The coffee break in the middle of the afternoon was eagerly utilised by all to refresh old connections and build new ones. And even after the event closed and evening started to set in, a large bunch of cat experts continued to mingle on the premises of Partner Re. And a dare-devil remainder of the group went on to test food & wine quality in a near-by restaurant, but this will have to remain a story for another day!

The organisation of such an event, as always, would not be possible without the help of a number of people contributing with great dedication. The core organising committee consisting of Peter Zimmerli, Manuel Prechtl and Kirsten Mitchell-Wallace, is grateful for the dedication shown by Lysiane Mayoraz and Stefan Rimkus from Scor, Paul Della-Marta, Anne-Sophie Scheidegger and Irene Staedeli from Partner Re, Dominik Renggli from Swiss Re and Maria Frontoso from RMS. A great thank you also goes to all speakers that accepted our invitation and especially those who had to travel some distance to Zurich.
​
Last but not least: After Scor in 2016 and Partner Re in 2017, calendars can already be marked for 2018 – Swiss Re committed to host next year's ISCM Zurich event in its newly inaugurated building across the lake on the 20th of September 2018. Stay tuned for more on this!

The Zurich ISCM events have always relied on the infrastructure provided by a hosting company – this year it was Partner Re. Key faces behind the flawless organisation within Partner Re were (from left to right) Paul Della-Marta, Erik Ruettener and Manuel Prechtl.

​People are wondering where the actual versus modeled lossfor Maria may eventually wind up. The answer to that question depends on the often ignored and missing fifth box of modeling – what I refer to as Modeling Chaos. Cat models are a mix of science, engineering, experience and judgement. Modeling building coverages have a strong foundations in science, engineering and experience. As we move further away from the traditional building coverages the models gradually move along with it introducing more professional judgement and with it more loss uncertainty in the analysis. 

Models do an excellent job of evaluating wind, earthquake, or flood risk when considered as a standalone structure insulated from other surrounding influences. However catastrophe events do not exclude other influences. Natural phenomena typically impact large areas introducing many new hazard elements and competing recovery needs which are invisible to most of today's models. The interaction between the modeled risk(s) and it's surrounding influences can be quite extreme, variable, and chaotic. As such it is difficult to calculate and model. Some may be included by a general loss amplification adjustment but the majority of the surrounding influences are too variable by event or location and not considered. These should be within a missing and not fully comprehended fifth box of models that accounts for the chaos surrounding an event. A fifth important box that needs to be included with the hazard, vulnerability, encoding and financial modeling boxes of today. 

Models are learning models and with each new event we begin to act on things that we suspected were deficient in our prior thinking. Three major chaotic components that are not fully considered when modeling extreme events and need to be are how the human, infrastructure, interplay between mutiple hazards, and the competing elements influence the eventual insured and economic losses. 

Human elements included preparedness, reaction, decision priorities, socio-economic resiliency, psychological, and other responses that can decrease or increase recovery costs. This would also include the potential for any post event political (regulatory and statutory) decisions. 

The quality and availability of a sound supporting infrastructure. This would include how extensive the supporting infrastructure can be, how well built it is, are there multiple means for moving resources in and out of the impacted area, how quickly can the infrastructure be repaired and most importantly how geographically accessible is the affected community by other means? Islands such as Puerto Rico are more isolated than other areas of the US and this makes it more difficult to undertake widespread infrastructure recovery. The lack of working infrastructure impedes our ability to rebuild the infrastructure the areas it serves. Examples of other the major areas that may be difficult to rebuild quickly due to geographical challenges include Long Island and NYC, Miami (peninsula), and San Francisco. Longer recovery times mean larger losses. 

Competing Elements.  This includes during and after the event. During an event the debris and building components of one structure could become damaging projectiles to other surrounding structures that otherwise would not have been damaged. Too many times an up to code building was damaged by the debris from a neighboring building. If a model does not have information about these surrounding structures then how can a model account for the additional exposure?.. it can't   After an event affected entities compete against each other for resources, material and labor, pushing up rebuilding costs and times. The greater and more congested, hard to get to, and conflicting the number of competeing entities seeking resources are then the more expensive and time consuming it will be to rebuild.

As an industry our first concern is to help those impacted by hurricane Maria. We hope that people can get back to their normal lives as soon as possible. As someone who has looked at natural catastrophes for decades I also see Maria as another lesson to all of us just how important the influence that this fifth box has on the actual versus modeled losses. It could be significant or less than anticipated by some. Nevertheless the fifth box, Modeling Chaos, needs to be accounted for when considering capacity,modeling, and underwriting results. To ignore this box challenges our understanding of the true economic and insured risk that we face. I am eager to see how the human, infrastructure, and competing elements of the fifth box differed in the industry's modeling of Maria.