Haiti earthquake black swan

[E]

This morning on NPR, I heard Haiti President Preval comment that the entire country has changed.

The Haiti earthquake is as good example as any of a black swan event. Curious if any actuaries or ERM pros are already incorporating elements of it into catastrophe models?

Any other comments relating to ERM and the Haiti earthquake are welcome.

[campbell]

What are you considering your definition of "black swan event"?

Some catastrophic occurrence that people think is really rare, but on hindsight, isn't?

If that's your definition, I agree.

[E]

In addition to what you said. . . I think of it as a "game changer" event. . . something that irrevocably and significantly alters the path of things.

[JMO]

I read something about this in my latest issue of The Economist. Developing countries are under-insured for catastrophic events. Whatever coverage they have is through various donors. Awareness campaign needed?

[E]

I would think actuaries could play a key role in developing catastrophe models for developing nations. Anyone have any examples of this already happening?

[Jack]

The Haitian earthquake was so catostophic becuase the tend to build with unreinforced masonry. They don't need models the need rebar and quality cement.

[EddieC]

The Haitian earthquake was so catostophic becuase the tend to build with unreinforced masonry. They don't need models the need rebar and quality cement.

Maybe the developing countries need more CERAs.

[campbell]

Yeah, I'm not sure what one would insure [or who would insure] areas that have very low construction standards. I would think infrastructure improvement would be cheaper than cat coverage [and the cat coverage would be more affordable once there's good infrastructure]

[JMO]

http://ccrif.org/
Carribean Catastrophe Risk Insurance Facility. Governments of the area band together, apparently.

[E]

Catastrophe Risk Models Show January 12 Haiti Quake Presages Future Caribbean Earthquakes (http://insurancetech.com/showArticle.jhtml?articleID=222400067&cid=nl_ins_daily)


"The reason that people are surprised is that these events don't happen very frequently," Blakeman comments. "There have been devastating earthquakes in the past and we're pretty sure they'll happen in the future."

A reminder of that eventuality came with a 5.8 tremor that struck Grand Cayman, nearly 600 miles west of Haiti, on Jan. 19. The event was likely related to the Haiti quake and while it resulted in negligible damage, it demonstrates the potential for greater insurance losses, should more prosperous Caribbean locations be hit, Blakeman says.

Nearby locations with greater insurance exposure and tectonic features similar to Haiti include the Dominican Republic and Jamaica, according to Don Windeler, director of model management, Risk Management Solutions (Newark, Calif.), who notes that major earthquakes struck the Jamaican capital Kingston in 1692 and 1907. "The northern half of Hispaniola is dominated by the Septentrional fault system, an active fault with the potential to generate events in excess of M7.2. 'The Dominican city of Santiago is particularly exposed to this fault system," he says. "Just to the south, the Enriquillo-Plantain Garden fault system — the source of the 2010 Haiti earthquake and multiple others over the past 400 years — continues to the west where it eventually intersects Jamaica in the capital city of Kingston."

As infrequent as major such events may be, the Haiti earthquake fits in, more or less with risk modelers' appraisal of seismic risk, according to Windeler. "It occurred on a previously identified fault system that had generated major events in the past, creating a "hot spot" on hazard maps of the area."

The insurance industry continues to improve its picture of seismic risk with help from the USGS, and other organizations, according to Paul Earle, a USGS seismologist. Earle reports that the international Global Seismic Hazard Assessment Program (GSHAP), launched in 1992, is currently being updated to include the dimension of risk. Another program that has generated interest in the insurance industry is the USGS' PAGER (Prompt Assessment of Global Earthquakes for Response), which provides estimates of the number of people and names of cities exposed to severe shaing following significant quakes worldwide. "USGS is taking that effort into rapid predictions of loss of life and potential monetary loss, but that's still in the research phase," says Earle.



Related:

Insurers Must Calculate Risks of Tsunamis That Typically Follow Earthquakes (http://insurancetech.com/claims/showArticle.jhtml;jsessionid=TDNWLS3YZ2RFLQE1GHPSK HWATMY32JVN?articleID=222400072&_requestid=198735)

[whisper]

The Haiti earthquake is as good example as any of a black swan event. Curious if any actuaries or ERM pros are already incorporating elements of it into catastrophe models?


Honestly, if a Cat model didn't have this type of event built in on a known fault line that has produced major quakes in the past, they'd have fallen out of the market by now.

[tommie frazier]

has the term "black swan" jumped the shark?

as whisper points out, how was this not in the 1 in 250 range for that fault line? is that really that unforseen for a natural disaster? esp when that threshold is the kind that many places claim to prepare for. i'd think black swan events go WAY beyond the simple/existing modeling.

[Dr. John Zoidberg]

We don't write business that has a 1 in 250 of 38 dollars worth of damage after the 20 cent deductible.

[E]

Going beyond how we think of P&C insurance in the US, how could actuaries help developing nations better assess risks that could significantly alter the path of their civilization?

I'm thinking about this more from a general risk management perspective and less of a P&C perspective. (I'm not a P&C guy.)

[tommie frazier]

DJZ-yeah, P&C insurance wise this was nearly a loss free event. despite it being a human disaster.

E-i just can't imagine that preparations for a 7 on a fault line should be out of disaster planning for an entity on a fault line. if it was a 9, then I doubt anyone would be that ready. but what can we do to help them prepare? that geologists and seizmologists have not been able to do? i don't know. est damage assessments and dead anad services disrupted, sure.

[E]

DJZ-yeah, P&C insurance wise this was nearly a loss free event. despite it being a human disaster.

E-i just can't imagine that preparations for a 7 on a fault line should be out of disaster planning for an entity on a fault line. if it was a 9, then I doubt anyone would be that ready. but what can we do to help them prepare? that geologists and seizmologists have not been able to do? i don't know. est damage assessments and dead anad services disrupted, sure.

I wonder if there would be value in adding a higher level planning layer to all of it. Seismologists focus on the geological angle and actuaries have focused on the insurance angle. Each area probably overlaps to some degree, but it's still probably silo-oriented (I know the word silo is has probably jumped the shark as much as BS, which has an interesting acronym, but I digress).

I don't know a lot about the US Homeland Security department--other than the general negative press it received post-Katrina--but I suspect that it attempts to fuse all these silos together. It may be utopian to envision the same type of thing working for all nations. (Not even sure if ours works yet.)

With the Haiti earthquake, there is the obvious infrastructure damage element that, as actuaries, we probably focus on more naturally. But the Haiti earthquake's aftermath goes well beyond the loss of infrastucture. There is a significant anthropological element. There civilization is now on a different path. Families are split up, knowledge has been lost, productivity will be hampered for a long time. These are the aspects that may not be on the radar during risk assessment but are fully realized when the disaster plays out.

[tommie frazier]

i agree that a lot of our focus tends to be on what we know best. but the matter at the center of this is from the seizmologist/geologist/cat model people: what is the lilihood of an event of a certain magnitude? if realistic enough of a threat, there should be contingency plans in place. the geology/cat model is not a silo-it is the source feuling the planning of all the others.

[E]

i agree that a lot of our focus tends to be on what we know best. but the matter at the center of this is from the seizmologist/geologist/cat model people: what is the lilihood of an event of a certain magnitude? if realistic enough of a threat, there should be contingency plans in place. the geology/cat model is not a silo-it is the source feuling the planning of all the others.

I agree--the seismology event is the catalyst. Maybe then the focus should be on scoping out all the ripple effects to make a more robust model.

[campbell]

I'm curious if the construction code is such that buildings stand up well even to hurricanes, as they do blow through that area with some frequency. I would be interested in knowing if there is much variation between the various Caribbean islands in this respect.

[E]

I'm curious if the construction code is such that buildings stand up well even to hurricanes, as they do blow through that area with some frequency. I would be interested in knowing if there is much variation between the various Caribbean islands in this respect.

That's an interesting point. The Caribbean is definitely an area susceptible to natural disasters, in general, on a fairly frequent basis. It's probably a valuable area to use as a case study on risk management for developing nations.

[JMO]

I'm curious if the construction code is such that buildings stand up well even to hurricanes, as they do blow through that area with some frequency. I would be interested in knowing if there is much variation between the various Caribbean islands in this respect.

That's an interesting point. The Caribbean is definitely an area susceptible to natural disasters, in general, on a fairly frequent basis. It's probably a valuable area to use as a case study on risk management for developing nations.

Myh understanding is that Haiti had NO building code. (Or if there was one, it wasn't enforced.) But that is not the underlying problem. The underlying problem is/was that Haiti is a "failed state" and there is nobody taking any responsibility for leadership, let alone for disaster planning. So long as the top guys are only interested in getting power for themselves, infrastucture and other governmental responsibilities will come in second place. Also, to the extent that the UN and others were attempting to help, the really bad news was that the location of the quake knocked down whatever structure (physical and organizational) they had established.

Admittedly I have no plan to solve all this. And it is worth noting that Haiti is not the only place in the world with such a leadership problem.

[E]

Myh understanding is that Haiti had NO building code. (Or if there was one, it wasn't enforced.) But that is not the underlying problem. The underlying problem is/was that Haiti is a "failed state" and there is nobody taking any responsibility for leadership, let alone for disaster planning. So long as the top guys are only interested in getting power for themselves, infrastucture and other governmental responsibilities will come in second place. Also, to the extent that the UN and others were attempting to help, the really bad news was that the location of the quake knocked down whatever structure (physical and organizational) they had established.

Admittedly I have no plan to solve all this. And it is worth noting that Haiti is not the only place in the world with such a leadership problem.

These are good points. How is "failed leadership risk" accounted for in a typical enterprise's risk framework? I suspect that would be an interesting conversation with senior management.

[JMO]

These are good points. How is "failed leadership risk" accounted for in a typical enterprise's risk framework? I suspect that would be an interesting conversation with senior management.
A friend of mine who invested in a few small companies used what he called the S/S ratio. Smart/Stupid. The higher the proportion of smart people, the more likely he was to invest. Of course, his was a fairly subjective view of who was smart and who was stupid. But I do know that he classed actuaries among the smarts.

[whisper]

These are good points. How is "failed leadership risk" accounted for in a typical enterprise's risk framework? I suspect that would be an interesting conversation with senior management.

Depends on the issue.

1.) If its the leadership hasn't engaged in this type of analysis - and it should - chances are this is a conversation that is not happening in the 1st place.

2.) If its something where leadership has taken a position - and the wrong position - you see that they've taken a wrong position, articulate what is wrong with it, and then get their ear (or pocketbook - depending on the relationship). It may be easy to see that they've taken a wrong position - it may be very hard to articulate why.

[Runner]

I'm curious if the construction code is such that buildings stand up well even to hurricanes, as they do blow through that area with some frequency. I would be interested in knowing if there is much variation between the various Caribbean islands in this respect.
I don't know what the code is, but I would bet that unreinforced masonry holds up relatively well to huricanes, since it's heavy. Gravity is on your side there. It doesn't hold up well to earthquakes. That's a fact.

Whether builders follow building codes is not the point. The point is they know hurricanes are coming, and they didn't know an earthquake was coming. They build stuff the same way the built the building next door. The ones standing (prior to the recent earthquake) were the ones that could stand up to hurricanes.

[E]

I don't know what the code is, but I would bet that unreinforced masonry holds up relatively well to huricanes, since it's heavy. Gravity is on your side there. It doesn't hold up well to earthquakes. That's a fact.

Whether builders follow building codes is not the point. The point is they know hurricanes are coming, and they didn't know an earthquake was coming. They build stuff the same way the built the building next door. The ones standing (prior to the recent earthquake) were the ones that could stand up to hurricanes.

I think that's one problem all entities (and individuals) face: We plan for the last disaster. Interestingly, according to this source (http://www.telegraph.co.uk/news/worldnews/centralamericaandthecaribbean/haiti/6978919/Haiti-earthquake-history-of-natural-disasters-to-hit-the-country.html), the last major earthquake was in 1946--not that long ago in the context of building construction. The 1946 earthquake was massive too.

[the BS method]

I don't understand how you could insure people that can barely afford to eat. And turning to the government isn't going to yield any answers because they surely have a hand in such widespread poverty.

If insurance is to be required the population might disperse as things are rebuilt.