Defining events: the 2010-11 Australia and New Zealand disasters, 10 years on

Mother Nature firing from all corners

In just a few months from the end of 2010 to early 2011, a series of extraordinary natural catastrophes in Australia and New Zealand cost many lives, brought hardship to many more and inflicted heavy financial losses. Major floods in Australia and earthquakes in New Zealand became the countries' costliest natural catastrophes, the insurance industry's costliest events in the region, and among the largest insured events worldwide. The 2010-11 catastrophe season is still unsurpassed in terms of large-scale devastation and insured claims paid by the insurance industry. These events also brought lessons for the communities and insurance industry.

Ten years on, we remember those who lost their lives, pay tribute to the massive reconstruction efforts and reflect on how these events changed the insurance industry. While Australians had previously coped with bushfires, large hail and other storms, few in New Zealand had experience of large-scale disastrous events prior to the Christchurch earthquakes. There is much we can do to lessen the impact of future events by enhancing the regions' resilience.

This is vital today following the latest devastating floods on Australia’s east coast in March 2021, which caused widespread damage as well as loss of life. Australia also suffered a costly season in 2019-20, with high damage bills from bushfires, hailstorms and floods, despite a severe drought. We believe this frequency of natural disasters, coupled with socioeconomic factors such as accumulation of human and physical assets, short-term climate drivers and long-term climate trends, will result in rising exposure to catastrophe losses. There is an urgent need for climate action and investment in strengthening economic and insurance resilience against future shocks.

Figure 1. Timeline of events

Catastrophes that changed the insurance industry

How did the 2010-11 catastrophes occur and what change did they bring about in the insurance industry?

Earthquakes in New Zealand

In September 2010 and February 2011, New Zealand’s third-largest city, Christchurch, was rocked by two destructive earthquakes and a long and complex series of major aftershocks lasting until the end of 2011. Both shocks triggered widespread liquefaction at levels never seen before due to the nature of the surface (deep alluvial sediments and high water-tables). This destroyed buildings, infrastructure and land, and caused the Christchurch CBD to be cordoned off for almost 29 months, causing severe disruption. While the 2011 earthquake claimed 185 lives, of which 115 died in one building with various design deficiencies (according to the comprehensive Royal Commission), there were no deaths in the 2010 earthquake, testament to the country's investment in strengthening the seismic structural resilience of buildings and critical infrastructure.

The 22 February 2011 Christchurch earthquake was, from a seismological point of view, an aftershock of the earlier earthquake of 4 September 2010. Although weaker, the aftershock had a far more devastating impact on the city than the earlier large seismological event, mainly due to proximity to the CBD, a shallow depth of epicentre, and site amplification effects that increased ground accelerations. The second earthquake also created higher losses for the insurance industry. Though it was well known that multiple aftershocks follow large earthquakes, and this often persists for years, the New Zealand events tested the industry‘s assumption about the scale of consecutive disasters in the same region. A subsequent event on 13 June 2011 added to the overall damage.

The earthquakes damaged about 90% of greater Christchurch's housing stock, approximately 167 000 homes.1Re/insurers paid roughly 80% of all damages, as New Zealand has among the world's highest insurance penetration rates for seismic risk. This is due to high risk awareness and the New Zealand Earthquake Commission (EQC), which provides compulsory earthquake insurance for residential properties whenever a fire policy is purchased by the owner. The February 2011 seismic shock is the third-largest insured loss from this peril worldwide.

Lessons for the insurance industry:

  • Hazard beyond expectation: Prior to the earthquakes, the industry’s seismic risk models did not predict the extent of liquefaction-induced ground deformation, nor how it affected the seismic performance of several large modern buildings. Many property owners were not only faced with having to repair or rebuild their homes, but also with restoring the land itself.
  • Losses beyond expectation: Insured claims far exceeded the model-based prediction, taking the insurance industry by surprise and exposing the industry to loss escalation. This also created potential issues for the sustainability of re/insurance offerings. Several factors contributed to the higher-than-expected claims, including difficulties processing the large number of claims; challenges in allocating losses to the long series of individual earthquake events; costly upgrades resulting from strengthened building codes; and special features of policy wordings such as unlimited replacement values. Uncapped earthquake policies are now largely removed from the market. The Canterbury series also revealed important shortcomings in earthquake models. These included progressive weakening of structures not captured by damage functions due to earthquakes occurring in quick succession, as well as highly uncertain cost apportionment due to under-insurance2.
  • Building codes save lives, not buildings: The Canterbury series confirmed inherent issues related to building codes that ended up inflating insured claims beyond expectation. In New Zealand, the demolition rate of reinforced concrete buildings was around 60% following the events in 2010 and 2011.3 While the buildings met engineering expectations with respect to saving lives, they were demolished because they were beyond repair.

Cyclone Yasi

Severe tropical cyclone Yasi slammed into the North Queensland coast near Mission Beach in the early hours of 3 February 2011. A category five storm at landfall, it was one of the most powerful tropical cyclones to hit Queensland since records began. The Bureau of Meteorology estimated peak wind gusts of up to 285km/h near the eye wall. 

As bad as Yasi was for those who experienced its fury (and paid its claims), it could have been far worse had it not veered 140km from the city of Cairns. Thankfully the hit or miss nature of Queensland cyclone risk (a function of discrete population and asset concentrations along the coast) has been characterised mostly by "misses" over the past couple of decades, with cyclones making landfall predominantly in less exposed areas.

According to the Insurance Council of Australia, Yasi generated AUD 1.4bn of insured losses at the time, far eclipsing the AUD 540m insurance payment from tropical cyclone Larry, which made landfall in a similar area five years earlier.
The industry's early estimates of Yasi's losses were closer to AUD 1bn, but claims continued to develop, partly due to the resource constraints that led to post-event inflation in the wake of the Brisbane floods only a few weeks prior.

Lessons for the insurance industry:

  • History alone cannot fully inform us about future risk potential: Swiss Re has reconsidered as-if loss scenarios for major historical events – such as 1992's Hurricane Andrew directly hitting Miami, Florida4. Our modelling suggests that if Hurricane Andrew were to directly hit Miami, rather than following its historical path a mere 20 miles to the south, the as-if losses would be as much as three times as high. These types of studies highlight the loss sensitivity of cyclone paths relative to the geographical spread of (re)insured assets Similarly, had Yasi tracked directly over Cairns, instead of turning away and crossing the coast to the south, we would expect the losses to be multiples of the actual Yasi event loss.
  • Incorporating the ‘what if’ into prediction models:  Scenario analyses and probabilistic catastrophe modelling are both important methods to assess risk scenarios and inform underwriters about the correct price for risk. This is why Swiss Re is currently completing a full rebuild of our Australian Cyclone model, incorporating lessons learned from historical events such as Yasi and Debbie.  We also validate model output with realistic disaster scenarios to ensure that our models are correctly informing underwriters about the full range of loss outcomes.

Floods in Australia

Prolonged, extensive rainfall over Queensland from December 2010 to January 2011 led to flooding of large areas, including the state capital Brisbane. Rains spread southward to inundate parts of NSW and Victoria. Thirty-three people died in these floods; three remain missing. More than 78% of Queensland was declared a disaster zone and more than 2.5 million people were affected. Some 29 000 homes and businesses suffered some form of inundation.

2010 was Australia’s third-wettest calendar year and the wettest on record for Queensland. By December, the land and river catchments were already saturated to the point that the additional monsoon rains at the end of the year led to catastrophic water run-off and damage. The 2010-11 La Niña5 was one of the most intense on record, following years of severe drought in many parts of the country. As such, it brought relief to many Australian farmers, but devastation to others. In addition, La Niña coincided with a negative phase of the Indian Ocean Dipole, an irregular oscillation of sea-surface temperatures in the Indian Ocean, which tends to amplify the effects of La Niña. 

La Niña plays a significant role in driving Australia's weather variability, cooling atmosphere and increasing precipitation. Warmer ocean temperatures under La Niña conditions mean that the east coast of Australia experiences twice as many floods during La Niña years than in El Niño6 years. These conditions are also known to increase the average number of cyclones in a season. Swiss Re's analysis of historical insured losses in Australia under different natural climate variability modes shows that the flood losses under La Niña are 1.7 times the average loss, and for cyclone events, 1.6 times the average loss.

Lessons for the insurance industry:

  • The role of definitions: In the year following the floods and after significant discussion with the Federal Government, a standard definition of flood was introduced. This provided clarity to the consumer on what constitutes flood, regardless of insurer. 
  • The danger of the insurance protection gap: insurance penetration for flood risk in Australia was lower than that for earthquakes in New Zealand. The insured loss for that Queensland summer – estimated by the Insurance Council of Australia to be in the order of AUD 2.3bn (indexed to 2017 values) – was only a fraction of the total AUD 5.6bn flood damage bill, and the Australian federal government was forced to levy a new tax to fund the remainder. This highlighted the large flood insurance protection gap in Queensland on both private and government assets, and the opportunity to use risk transfer mechanisms more widely to pre-fund losses rather than post-fund them through community taxes.
  • The value of data to achieve insurance penetration: In the years following these floods, Australia's insurance industry responded by rolling out greater flood cover for private homes. Developments in flood risk assessment, such as the National Flood Information Database and other bespoke methods, meant that insurers offered this new product at prices based on the risk level at the individual property. As more data have become available, flood insurance availability in household policies has grown from 3% in 2006 to 93% of policies sold today, according to the Insurance Council of Australia. However, risk-based pricing also means that those with the highest risk levels are faced with insurance premiums that may become unaffordable, highlighting the importance of flood mitigation and appropriate planning to reduce and minimise risk, and assist in keeping insurance affordable.

2020-21 – 10 years later

Rebuilding after such large-scale natural disasters has been complex and lengthy, even in a country advanced in seismic engineering building like New Zealand. However, insurance plays a central role in building resilience, allowing for faster and more efficient reconstruction. New Zealand’s high insurance coverage ensured that the re/insurance industry funded a high proportion of the recovery costs. In Australia, the Queensland floods translated into higher uptake of flood insurance and we saw risk mitigation measures such as flood levees prove effective, helping improve insurance affordability as with the Queensland Roma case study. 

In 2019-20, Australia experienced a series of weather events including extended bushfires, severe convective storms and floods that resulted in high insurance claims. The 2019-20 bushfire season was the longest ever recorded, in the country’s hottest and driest summer on record, burning more land area and houses than ever before. Insurers faced claims of USD 1.6bn, the highest ever generated in one year by the bushfire peril in Australia. Fire seasons are typically more severe during El Niño conditions, however, the 2019-20 fires occurred in a neutral phase of the El Niño Southern Oscillation, where the combination of a strong and long-lived positive Indian Ocean Dipole, and strong negative phase of the Southern Annular Mode, fueled the fires. 
We expect extreme weather events to intensify in coming decades. Australia has warmed by 1.44°C since 1910, with 2019 the hottest year on record.7  The frequency of severe bushfire conditions in southern and eastern Australia is likely to increase in future climate scenarios, due to projected warming and drying. 

Drawing on the lessons of the past decade, we are acutely aware that for Australia and New Zealand, a resilient future will require a combination of mitigation, smart planning, and financial risk-management.

1.  Whole of Government Report: Lessons from the Canterbury earthquake sequence, Department of the Prime Minister and Cabinet (dpmc.govt.nz), 31 July 2017.
2.  www.actuaries.digital/2017/01/12/aftershock-can-model-change-impact-the-real-economy-in-new-zealand/
3.  J. H. Kim, “Quantitative Analysis of Factors Influencing Post-Earthquake Decisions on Concrete Buildings in Christchurch, New Zealand”, University of British Columbia, 2015.
4.  https://www.swissre.com/Library/hurricane-andrew-the-20-miles-that-saved-miami0.html
5.  http://www.bom.gov.au/climate/updates/articles/a020.shtml
6.  What is El Niño and what might it mean for Australia? (bom.gov.au)
7.  State of the Climate 2020, Commonwealth Scientific and Industrial Research Organisation (CSIRO), 2020. https://www.csiro.au/en/Showcase/state-of-the-climate

Author

​Climate risk - how can we be more resilient? Focus on Asia Pacific region