It has been called “China’s Hurricane Andrew”. Following the Sichuan earthquake in China, Helen Yates asks how future events could be better insured and whether a government-backed catastrophe pool could be the solution.
The terrible loss of life, building failures, aftershocks and amazing rescues have dominated the headlines since the magnitude 7.9 Sichuan earthquake struck on 12 May. While it is unlikely to be a major insured loss event, the catastrophe highlights China’s vulnerability to earthquakes.
At the time of writing the official death toll exceeded 65,000 and over 23,000 were still missing. Despite being a shallow earthquake, striking at a depth of less than 20km, the ground motions were felt over a very large area, its vibrations transmitted by dense rock. Most reports put the quake’s magnitude at 7.9, although Chinese seismologists claim it was magnitude 8.
The destruction was widespread across Sichuan, although densely-populated Chengdu appears to have escaped the worst of the damage. Rupturing in the mountains north and east of Chengdu, the quake was felt as far away as Hong Kong and Japan. Of the four million buildings destroyed, the greatest tragedy was the collapse of several schools, calling building codes into question.
“You can never say you are not surprised as the earthquake took place in a seismically active part of the world,” says JLT Re associate Keith Leung. “What was more surprising was the extremity of the damage we have witnessed in this event. I believe the widespread failure of buildings and structures will prompt a thorough review of the existing building codes and standards in earthquake engineering. The performance of highly engineered structures such as dams, as well as residential buildings will be subjects of much attention.”
Low insured loss
AIR Worldwide and Guy Carpenter estimate the economic losses could reach $20bn (RMB140bn). Insured losses, by contrast, are very low and unlikely to exceed $1bn (RMB7bn). Sichuan Province is largely rural and therefore insurance penetration, which is low in China anyway, is particularly low here.
Earthquake coverage is optional for both residential and commercial policies and take up in the region is minimal. “Many of the people who have lost everything in this tragedy will have no reimbursement,” says Peter Zimmerli, who heads the Swiss Re Cat Perils hub in Hong Kong. “They might get something from the government but they don’t have a vested right to receive money via insurance to rebuild their lives.”
“The insurance penetration in the Sichuan province is relatively low – around 2% to 5% depending on the line of business,” explains Milan Simic, managing director of AIR Worldwide Ltd. “The only exception to that is the CAR/EAR [construction all risks and erection all risks] because there is now more pressure in China to insure all the big construction projects that are on the ground. There, insurance penetration is in the order of 20% to 30%.” Simic adds the take up rates are much higher, 10% or greater, in China’s economic hubs – Shanghai, Beijing and Guangzhou.
History of devastation
“Neither the location nor the magnitude of the earthquake came as a surprise,” says Zimmerli. “There have been historic precedents exactly in this area.” China’s last major earthquake was 30 years ago. Occurring on 28 July 1976, the magnitude 7.8 Tangshan Earthquake flattened an entire city and caused 250,000 deaths. It sparked an immediate change in building codes as the country came to terms with the overwhelming loss of life.
The history of major earthquakes in China goes back a long way and records are extremely well documented. According to a 2006 report by Swiss Re on natural hazards in China, earthquakes “represent the largest single threat in terms of human and economic consequences” in China. In the last century, earthquakes claimed more than half a million lives in China. The world’s most deadly earthquake ever recorded occurred in 1556 in western China’s Shaanxi Province and is believed to have led to 830,000 fatalities.
The problem with earthquakes, points out Towers Perrin’s Philadelphia-based senior vice president Imelda Powers, interviewed while she was in Beijing, is that you can’t see them coming. “It’s difficult to pinpoint exactly when and where earthquakes will happen. Whereas with a hurricane you can see it coming days in advance.” Scientists are being urged to provide better quake predictions to reduce loss of life. Powers believes this has had some success in the past.
In 1975, officials ordered the evacuation of Chinese city Haicheng’s population of one million. They were prompted by earthquake foreshocks, changes in ground water level, land elevation and some odd animal behaviour. A few days after the evacuation, over 2,000 people died when a magnitude 7.3 earthquake struck on 4 February. It is estimated that the number would have been closer to 150,000 had the city not been evacuated. With none of these early warning signs, Tangshan – with a similar size population to Haicheng – was devastated just a year later.
Modelling the risk
China might have a good historical earthquake record, but catastrophe modelling is still largely at a CRESTA [Catastrophe Risk Evaluation and Standardizing Target Accumulations] level and this poses various problems. “The potential uncertainty of modelling risks at a CRESTA resolution, which is entire provinces in China is very high comparison to the modelling risks at a detailed level, as currently done in the US and in Europe,” explains AIR’s Simic.
Some Chinese provinces are larger than entire European countries, points out Zimmerli. “We have to move towards better resolution of geographical resolution in accumulation data,” he says. “I would like to see the market move to more refined exposure aggregation.” The cat models that do exist are also relatively new and accuracy is still to be seen, points out Kevin Bogardus, general Manager for Greater China at Benfield. “For most of the models you can go down to the six digit postcode level if the insurance companies are able to capture information at that level.”
The earthquake in Sichuan may provide the impetus to improve the resolution of catastrophe modelling in China. “Historically if you look at any established insurance market where large natural catastrophes have happened there has always been a boost in terms of people being more aware of risk and hence in trying to tackle that risk via scientific modelling,” explains Zimmerli. “It happened after Hurricane Andrew in 1992, after winter storms Lothar/Martin in 1999 and it certainly happened after Katrina.”
“If there is any upside [to the earthquake] it would hopefully be better modelling and risk control for the future,” says Bogardus. “We’re working very closely with the insurance companies to help them understand their risks and exposures.” CRESTA have put forward a zoning suggestion for China to improve the granularity of data in the models, but as with everything this will take time, predicts Simic. “It takes a minimum of two to three years to get things changed,” he explains. “People and organisations such as CRESTA say that this needs to be done but people on the ground need to be persuaded that it’s to their advantage.”
Encouraging take up
One reason for China’s low insurance penetration is that the population does not yet see the value in buying insurance. “People think of earthquake risk as more remote than it really is,” explains Powers. She believes the government needs to educate the wider public on the value of insurance. Zimmerli thinks the Sichuan earthquake could be a major motivator. “It could be a wake-up call for China as a society to further drive insurance as a means of mitigating the financial aspects of such a disaster.”
With China’s economy growing at a staggering pace, the country’s growing and increasingly affluent middle class could now be more easily encouraged to purchase insurance. This already seems to be happening. Premium growth rates have actually outstripped economic growth rates. According to Benfield, total premiums increased from $26bn to $61bn between 2001 and 2005, representing a growth rate of 25% per annum. Life insurance has led this trend rising from 41% of the total insurance market in 1996 to comprise an incredible 71% of the market in 2005.
As the market has opened up in line with China’s WTO accession commitments and the removal of state-run China Re’s 15% cession rates, international reinsurers have flooded into the market. Lloyd’s, Swiss Re, Munich Re, Gen Re and SCOR among others have entered in the hope of gaining first-mover advantage and reaping the reward in years to come as China develops into an economic powerhouse.
As recently as 4 May, according to a story in the South China Morning Post, Lloyd’s was keen to expand its scope in China to cover natural disasters. “Reinsurance companies can help provide stability to the insurance sector and play an important role when catastrophes like snowstorms, floods and earthquakes hit China,” said Lloyd’s chairman Lord Levene.
The bid for market share by international players has seen prices soften considerably in the competitive environment. Property per risk and property catastrophe rates fell between 10% and 20% at the 1 January renewals. Despite the efforts of some players, reinsurance prices are still around 50% below the global average. While there will be some reinsurance losses as a result of the earthquake, these are not expected to be significant and are unlikely to harden up rates. “It’s very difficult to predict how the market is going to react,” says Zimmerli. “There will certainly be reinsurance companies contributing to the insured loss.”
“China will remain a very attractive market in terms of its enormous growth potential,” adds Zimmerli. “If you take a longer-term perspective, that will definitely remain.” While the earthquake will serve as a stark reminder that these risks are out there, he doesn’t think it will put any new entrants off the market. “Key for anyone who wants to do business in China will be that they have a sound technical approach in terms of risk assessment and pricing of the business.”
The combination of increasing wealth and a major catastrophe could boost insurance take up, believes Simic. “It could potentially lead to a major rethinking of how disaster mitigation is carried out,” he predicts. “One solution is, with government backing, to encourage greater uptake of insurance and you can do that through a number of ways.” People in many parts of the world still view insurance as a form of taxation without any major benefits, adds Simic. “There has to be a culture built over many years and decades to help people see the benefits.”
Another solution could be the creation of a government-led catastrophe pool, similar to those set up in Taiwan and Japan. Such a public/private initiative has been discussed in the past and as recently as January, following China’s worst snowstorms in 50 years, it was firmly back on the agenda. Should such a pool be set up, it could create new opportunities for reinsurance companies and even the capital markets. The Chinese government, which budgets to cover the cost of floods and typhoons every year, would also benefit. “The government would not have to sacrifice or postpone other plans in the national budget in order to fund emergency response from catastrophes covered by such arrangements,” explains Leung.
“In 2006, both the government and the insurance regulator issued statements announcing China’s intent to establish a natural hazards insurance scheme on a national level,” explains Zimmerli. He believes indigenous insurers will be encouraged to provide earthquake cover should such a pooled solution be set up, because it will offer economies of scale, while the government involvement will also be reassuring. “Should such a pooled solution now be put on the fast track then insurance take up on the earthquake side would increase dramatically.”
While the scale may not currently exist for insurance-linked securities to be an option in China, “sooner or later the capital markets will become very interested in these options,” says Zimmerli. Powers points out that the Taiwan Residential Earthquake Insurance Pool successfully launched a catastrophe bond in 2003. She says the challenge is not in setting up a catastrophe pool, but in encouraging take up once it has been set up. “In all these situations there are some ways for people to opt out and not participate.”
The Sichuan earthquake, January snowstorms and continued threat of typhoons is now, more than ever before, likely to persuade the government of the value of a long-term solution. The near misses are also convincing. Last year, Super Typhoon Wipha led to mass evacuation as the storm moved worryingly close to Shanghai. There is no escaping China’s vulnerability to natural hazards.
Once the government decides to act, it will all happen very quickly, predicts Simic. “It’s almost a blessing that China has a very strong government and if they want something they will can implement it relatively easily.”
Helen Yates is editor of Global Reinsurance.