Monthly Archives: July 2015

Opportunities and Challenges ahead for Vietnam: Lessons Learned from Thailand

Earlier this month I gave a presentation at the 13th Asia Insurance Review conference in Ho Chi Minh City, Vietnam. It was a very worthwhile event that gave good insights into this young insurance market, and it was great to be in Ho Chi Minh City—a place that immediately captured me with its charm.


Bangkok, Thailand during the 2011 floods. Photo by Petty Officer 1st Class Jennifer Villalovos.

Vietnam shares striking similarities to Thailand, both from a peril and an exposure perspective. And, for Vietnam to become more resilient, it could make sense to learn from Thailand’s recent natural catastrophe (NatCat) experiences, and understand why some of the events were particularly painful in absence of good exposure data.

NatCat and Exposure similarities between Thailand and Vietnam 

Flood profile Vietnam shows a similar flood profile as Thailand, with significant flooding every year. Vietnam’s Mekong Delta, responsible for half of the country’s rice production, is especially susceptible to flooding.
Coast line Both coastlines are similar in length[1] and are similarly exposed to storm surge and tsunami.[2]
Tsunami & Tourism Thailand and its tourism industry were severely affected by the 2004 Indian Ocean Tsunami. Vietnam’s coastline and it’s tourism hotspots (e.g. Da Nang) show similar exposure to tsunami, potentially originating from the Manila Arc.2
GDP growth Thailand’s rapid GDP growth and accompanying exposure growth in the decade prior to the 2011 floods caught many by surprise. Vietnam has been growing even faster in the last ten years[3]; and exposure data quality (completeness and accuracy) have not necessarily kept up with this development.
Industrialization and global supply chain relevance Many underestimated the significance Thailand played in the global supply chain; for example, in 2011 about a quarter of all hard disk drives were produced in Thailand. Currently, Vietnam is undergoing the same rapid industrialization. For example, Samsung opened yet another multi-billion dollar industrial facility in Vietnam, propelling the country to the forefront of mobile phone production and increasing its significance to the global supply chain.

Implications for the Insurance Industry

In light of these similarities and the strong impact that global warming will have on Vietnam[4], regulators and (re)insurers are now facing several challenges and opportunities:

Modeling of perils and technical writing of business needs to be at the forefront of every executive’s mind for any mid-to long-term business plan. While this is not something that can be implemented overnight, the first steps have been taken, and it’s just a matter of time to get there.

But to get there as quickly and efficiently as possible, another crucial step stone must be taken: to improve exposure data quality in Vietnam. Better exposure insights in Thailand would almost certainly have led to a better understanding of exposure accumulations and could have made a significant difference post floods, resulting in less financial and reputational damage to many (re)insurers.

As insurance veterans know, it’s not a question of if a large scale NatCat event will happen in Vietnam, but a question of when. And while it’s not possible to fully eliminate the element of surprise in NatCat events, the severity of these surprise can be reduced by having better exposure data and exposure management in place.

This is where the real opportunity and challenge lies for Vietnam: getting better exposure insights to be able to mitigate risks. Ultimately, any (re)insurer wants to be in a confident position when someone poses this question: “Do you understand your exposures in Vietnam?”

RMS recognizes the importance of improving the quality and management of exposure data: Over the past twelve months, RMS has released exposure data sets for Vietnam and many other territories in the Asia-Pacific. To find out more about the RMS® Asia Exposure data sets, please e-mail asia-exposure@rms.com.  

[1] Source: https://en.wikipedia.org/wiki/List_of_countries_by_length_of_coastline
[2] Please refer to the RMS® Global Tsunami Scenario Catalog and the RMS® report on Coastlines at Risk of Giant Earthquakes & Their Mega-Tsunami, 2015
[3] The World Bank: http://data.worldbank.org/country/vietnam, last accessed: 1 July 2015
[4] Vietnam ranks among the five countries to be most affected by global warming, World Bank Country Profile 2011: http://sdwebx.worldbank.org/climateportalb/doc/GFDRRCountryProfiles/wb_gfdrr_climate_change_country_profile_for_VNM.pdf

The 2015 Northwest Pacific Typhoon Season: Already a Record-Breaker

While the Atlantic hurricane season is expected to be below average this year, the North Pacific is smashing records. Fuelled by the strengthening El Niño conditions, the Accumulated Cyclone Energy (ACE)—used to determine how active a season is by measuring the number of storms, their duration and their intensity—continues to set unprecedented highs for the 2015 season.  According to Dr. Philip Klotzbach, a meteorologist at Colorado State University, the North Pacific ACE is 30% higher for this time of year than at any other time since 1971.

Philip J. Klotzbach, Colorado State University

To date, there have been 12 named Northwest Pacific storms, of which three have strengthened to Category 5 super-typhoon status, and two have strengthened to Category 4 typhoon. Typhoon Maysak was the first of the super-typhoons to develop and is reportedly the strongest known storm to develop so early in the season—it eventually passed over the northern Philippines in late March as a tropical depression. Super-Typhoons Noul and Dolphin followed in quick succession in May, with Noul scraping the northern tip of the Philippines, and Dolphin tracking directly in-between the islands of Guam and Rota.

China is recuperating after getting hit by Typhoons Linfa and Chan-Hom only days apart. Linfa made landfall on July 9, bringing strong winds and heavy rainfall to Hong Kong and southern China’s Guangdong province. Two days later, Chan-Hom brought tropical storm-force winds and heavy rainfall to Taiwan and the Japanese Ryukyu Islands before briefly making landfall as a weak Category 2 storm over the island of Zhujiajian in the Zhejiang province. Prior to landfall, Chan-Hom was anticipated to pass over Shanghai, but swung northeast and missed China’s largest city by 95 miles. Despite this near-miss, Chan-Hom still stands as one of the strongest typhoon to have passed within 100 miles of the city in the past 35 years.

Typhoon Nangka, the first typhoon to hit Japan this season, intensified to a Category 4 storm before ultimately making landfall as a Category 1 storm over the Kochi Prefecture on Shikoku Island, Japan. Although Nangka’s strength at landfall was weaker than originally forecast, the high level of moisture within the system caused significant rainfall accumulations, leading to widespread flooding and the threat of landslides. While there was an initial fear of storm surge in Osaka Bay, there has been limited damage reported.

This record-breaking season has been strongly influenced by the strengthening El Niño conditions, which can be characterised by several physical factors including warmer sea surface temperatures, a higher number of Category 3-5 typhoons, and a greater proportion of typhoons that follow recurring or northward tracks—all of which have been evident so far this year.

With El Niño conditions expected to continue intensifying the storms to come, this season highlights the necessity for a basin-wide multi-peril model, connected through an event-based approach and correlated geographically through a basin-wide track set. These will be featured in the new Japan typhoon model, due out next year, followed by the South Korea and Taiwan typhoon models. The RMS China typhoon models currently models typhoon wind, inland flood and surge for a correlated view of risk.

As El Niño conditions continue to bolster the Northwest Pacific typhoon season, RMS will be monitoring the situation closely. In September, RMS will be releasing a white paper on ENSO in the West Pacific that will provide further insight into its affects.

2015 North Atlantic Hurricane Season: What’s in Store?

RMS recently released its 2015 North Atlantic Hurricane Season Outlook. So, what can we expect from this season, which is now underway?

2015 season could be the 10th consecutive year without a major landfalling hurricane over the United States.

The 2014 season marked the ninth consecutive year that no major hurricane (Category 3 or higher) made landfall over the United States. Although two named storms have already formed in the basin so far this year, Tropical Storm Ana and Tropical Storm Bill, 2015 looks to be no different. Forecast groups are predicting a below-average probability of a major hurricane making landfall over the U.S. and the Caribbean in the 2015 season.

The RMS 2015 North Atlantic Hurricane Season Outlook highlights 2015 seasonal forecasts and summarizes key meteorological drivers in the Atlantic Basin.

Forecasts for a below-average season can be attributed to a number of interlinked atmospheric and oceanic conditions, including El Niño and cooler sea surface temperatures.

So what factors are driving these predictions? A strong El Niño phase of the El Niño Southern Oscillation (ENSO) is a large factor, as Jeff Waters discussed previously.

Source: NOAA/ESRL Physical Sciences Division

Another key factor in the lower forecast numbers is that sea surface temperatures (SSTs) in the tropical Atlantic are quite a bit cooler than previous years. SSTs higher than 80°F (26.5°C) are required for hurricane development and for sustained hurricane activity, according to NOAA Hurricane Research Division.

Colorado State University (CSU)’s June 1st forecast is calling for 8 named storms, 3 hurricanes, and 1 major hurricane this season, with an Accumulated Cyclone Energy (ACE) index—used to express activity and destructive potential of the season—of 40. This is well below the 65- and 20-year averages, both over 100.

However, all it takes is one significant event to cause significant loss.

Landfalls are difficult to predict more than a few weeks in advance, as complex factors control the development and steering of storms. Despite the below-average number of storms expected in the 2015 season, it only takes one landfalling event to cause significant loss. Even if the activity and destructive energy of the entire season is lower than previous years, factors such as location and storm surge can increase losses.

For example, Hurricane Andrew made landfall as a Category 5 storm over Florida in 1992, a strong El Niño year. Steering currents and lower-than-expected wind shear directed Andrew towards the coastline of Florida, making it the fourth most intense landfalling U.S. hurricane recorded. Hurricane Andrew also holds the record for the fourth costliest U.S. Atlantic hurricane, with an economic loss of $27 billion USD (1992).

Sometimes, a storm doesn’t even need to be classified as a hurricane at landfall to cause damage and loss. Though Superstorm Sandy had Category 1 hurricane force winds when it made landfall in the U.S., it was no longer officially a hurricane, having transitioned to an extratropical storm.  However, the strong offshore hurricane force winds from Sandy generated a large storm surge, which accounted for 65 percent of the $20 billion insured losses.

While seasonal forecasts estimate activity in the Atlantic Basin and help us understand the potential conditions that drive tropical cyclone activity, a degree of uncertainty still surrounds the exact number and paths of storms that will form throughout the season. For this reason, RMS recommends treating seasonal hurricane activity forecasts with a level of caution and to always be prepared for a hurricane to occur.

For clients, RMS has released new resources to prepare for the 2015 hurricane season available on the Event Response area of RMS Owl.

The Curious Story of the “Epicenter”

The word epicenter was coined in the mid-19th century to mean the point at the surface above the source of an earthquake. After discarding explanations, such as “thunderstorms in caverns” or “electrical discharges,” earthquakes were thought to be underground chemical explosions.

Source: USGS

Two historical earthquakes—1891 in Japan and 1906 in California—made it clear that a sudden movement along a fault caused earthquakes. The fault that broke in 1906 was almost 300 miles long. It made no sense to consider the source of the earthquake as a single location. The word epicenter should have gone the way of other words attached to redundant scientific theories like “phlogiston” or the “aether.”

But instead the term epicenter underwent a strange resurrection.

With the development of seismic recorders at the start of the 20th century, seismologists focused on identifying the time of arrival of the first seismic waves from an earthquake. By running time backwards from the array of recorders they could pinpoint where the earthquake initiated. The point at the surface above where the fault started to break was termed the “epicenter.” For small earthquakes, the fault will not have broken far from the epicenter, but for big earthquakes, the rupture can extend hundreds of kilometres. The vibrations radiate from all along the fault rupture.

In the early 20th century, seismologists developed direct contacts with the press and radio to provide information on earthquakes. Savvy journalists asked for the location of the “epicenter”—because that was the only location seismologists could give. The term “epicenter” entered everyday language: outbreaks of disease or civil disorder could all have “epicenters.” Graphics departments in newspapers and TV news now map the location of the earthquake epicenter and run rings around it—like ripples from a stone thrown into a pond—as if the earthquake originates from a point, exactly as in the chemical explosion theory 150 years ago.

The bigger the earthquake, the more misleading this becomes. The epicenter of the 2008 Wenchuan earthquake in China was at the southwest end of a fault rupture almost 250km long. In the 1995 Kobe, Japan earthquake, the epicenter was far to the southwest even though the fault rupture ran right through the city. In the great Mw9 2011 Japan earthquake, the fault rupture extended for around 400km. In each case TV news showed a point with rings around it.

In the Kathmandu earthquake in April 2015, television news showed the epicenter as situated 100km to the west of the city, but in fact the rupture had passed right underneath Kathmandu. The practice is not only misleading, but potentially dangerous. In Nepal the biggest aftershocks were occurring 200km away from the epicenter, at the eastern end of the rupture close to Mt Everest.

How can we get news media to stop asking for the epicenter and start demanding a map of the fault rupture? The term “epicenter” has an important technical meaning in seismology; it defines where the fault starts to break. For the last century it was a convenient way for seismologists to pacify journalists by giving them the easily calculated location of the epicenter. Today, within a few hours, seismologists can deliver a reasonable map of the fault rupture. More than a century after the discovery that a fault rupture causes earthquakes, it is time this is recognized and communicated by the news.