Insights. Japan Earthquake. Insurance Industry Impact and Risk Management Lessons. Background

Transcription

1 Insights April 2011 Japan Earthquake Insurance Industry Impact and Risk Management Lessons Background On March 11 at 2:46 p.m., a magnitude 9.0 earthquake struck an area 370 kilometers (230 miles) northeast of Tokyo. The offshore quake, the fifth largest worldwide since records began, spawned tsunamis that eventually reached the entire Pacific Rim. Within an hour of the earthquake, a wall of water up to 15 meters (50 feet) high hit the Japanese coast, sweeping away towns and villages in its path. While the ground motion from the earthquake caused a manageable level of damage, the rapidity and magnitude of the tsunamis proved to be more than warning systems could handle. Current estimates of casualties from this terrible disaster are now projected to exceed 28,000. The disaster is not yet over. Four nuclear power facilities (of 17 in Japan) were damaged by the earthquake and tsunami, and were shut down. The Fukushima Daiichi facility reported explosions at three of the reactors and, as of this writing, efforts continue to reestablish cooling systems and control the threat of radiation leaks. A perimeter zone around the power facility is now evacuated, and the loss of substantial power generation capacity led to enforced blackouts across parts of Japan. The impact of earthquakes on the Japanese market is being reinforced by reports of a powerful new magnitude 7.1 aftershock on April 7 that prompted a tsunami warning and raised further concerns over Fukushima and other nuclear facilities. The earthquake, tsunami and ensuing nuclear power issues will have severe economic consequences. Initial estimates of economic losses from the Japanese government now exceed US$300 billion. The Japanese equity market experienced a sharp sell-off immediately after the earthquake and remains volatile, while the yen has strengthened against the dollar. Globally, asset markets have reacted to news of the nuclear incident with a general sell-off in risky assets, as participants look to either hedge their risk or exit positions. The government has not announced any formal fiscal policies to date, but as the extent of the earthquake and tsunami damage becomes clearer, we would expect the government to implement a significant fiscal package as it did in Forty days after the Kobe earthquake, the Japanese government announced a fiscal package amounting to trillion (US$15 billion) of supplementary spending, which was subsequently increased to 3.2 trillion (US$39 billion). Historically, large natural catastrophes in developed countries have significant short-term effects on economic activity (mostly local) and financial markets that tend to fade over the medium term. Nevertheless, further deterioration at the Fukushima nuclear facility increases the number of possible outcomes for Japan s economy, the world economy and financial markets. Even if the reactors are stabilized, Japan will still lose substantial electrical generating capacity, which will affect its ability to resume industrial production. The earthquake, tsunami and ensuing nuclear power issues will have severe economic consequences. Initial estimates of economic losses from the Japanese government now exceed $300 billion.

2 While the economic losses are substantial and are expected to rise further, Towers Watson estimates that the total insured losses from the disaster in Japan will be between US$20 billion and US$45 billion. This estimate includes damages to residential and commercial properties (including business interruption), marine, auto and life insurance, but is exclusive of nuclear losses. Any potential liabilities associated with the nuclear exposure from the event are not significantly insured by the private market. Overall, the disaster will not have a devastating impact on the Japanese insurance industry or the international reinsurance market since most of the economic damages from this event are either uninsured or covered by government insurance programs. Figure 1. Japanese earthquake insurance sharing scheme ( trillions, US$ billions) % Japanese government 50% Japanese government 50% JER 100% JER 50% non-life insurers 5% JER 5% non-life insurers $67.1 $45.3 $23.5 $13.7 $1.4 Earthquake Coverage in Japan The residential property market in Japan is covered by a combination of non-life insurers and cooperative (co-op) insurers, of which the leading provider is Zenkyoren (JA). Standard property policies offered by non-life insurers exclude earthquake risk. Residential earthquake insurance (residential EQ) coverage and extended cover to the commercial market are offered as optional coverages. For a residential EQ policy offered by non-life insurers, the amount of coverage is typically selected by the insured to be between 30% and 50% of the fire limit of coverage, with a maximum limit of 50 million for the building and 10 million for its contents. By contrast, the standard property policy offered by Zenkyoren does not exclude earthquake risk. Residential EQ coverage was first established in 1964 following a large earthquake in Niigata, in northern Japan. Until 1995 s Kobe earthquake, the penetration ratio of residential EQ on policies issued by non-life insurers had been less than 10% of households. After the Kobe event, the penetration rate increased to the current rate of 22% of households. Combined with the EQ coverage issued by the co-op insurers, roughly half of all households in Japan carry EQ coverage. However, even when covered, the insured typically participates significantly in the exposure. Residential EQ cover is managed through a combination of: Non-life insurers The Japanese Earthquake Reinsurance Company (JER), which acts as a pooling vehicle on behalf of the insurers The Japanese government All non-life insurers that underwrite residential EQ cede 100% of the coverage to JER and in turn receive a retrocession from JER. The Japanese government also participates in the mechanism (Figure 1). The commercial property market is mainly covered by local non-life insurers. Earthquake insurance is written as an extended coverage endorsement to an ordinary fire policy and is offered to selected large corporations. 2 towerswatson.com

3 Given the high natural catastrophe risk present in this part of the world, solvency regulation of Japanese non-life insurance companies includes a solvency margin requirement specific to catastrophes. Currently, the required solvency margin for catastrophe risk is defined as the larger of the loss arising from a repeat of (1) the Kanto/ Tokyo earthquake of 1923, which is deemed equivalent to a one-in-200-year event, and (2) the Isewan Typhoon of 1959, which is deemed equivalent to a one-in-70-year event. Japanese insurers have a long track record of providing a high level of security to their policyholders. For example, we note that the available combined solvency margins for the three largest non-life insurance groups in Japan, Tokio Marine Group, NKSJ Group and MS&AD Group, exceeded US$120 billion as of March The total is equivalent to four times the required risk solvency margin of these carriers, inclusive of catastrophe risks. Zenkyoren has a solvency margin of US$90 billion, representing almost five times its required solvency margin. Based on the large market share of these three non-life groups (90% of the market share of non-life premiums, excluding the co-ops) and some of the largest co-ops (Zenkyoren in particular), Towers Watson believes that Japanese industry s capitalization will be sufficient to withstand our estimates of insured losses from this recent disaster. Insurance Loss Estimates As noted earlier, Towers Watson s estimate of insured losses is in the range of US$20 billion to US$45 billion, exclusive of any potential nuclear exposure (Figure 2). We provide more detail below on the various exposures that comprise this estimate. One of the key underlying assumptions in our assessment is the Japanese government estimate of US$300 billion of economic damages. Residential Our estimate of insured losses to residential properties from the combination of the earthquake and the ensuing tsunami is US$9.5 billion to US$21.9 billion, based on the assumption that approximately half of the population affected by these events currently purchases earthquake insurance. Using publicly available information and based on numerous discussions with industry executives, Towers Watson arrived at the breakout of loss estimates in the accompanying figures. It is important to note that the co-ops, including Zenkyoren, follow a different set of regulations from those that apply to insurance companies in Japan. In particular, they do not share their EQ exposures with the JER. We understand that Zenkyoren has purchased an excess of loss program from the international reinsurance markets that is said to be one of the largest programs in the world. It is believed that Zenkyoren s catastrophe reinsurance coverage is approximately US$8 billion excess of US$3.3 billion, bringing our ground-up loss estimate for Zenkyoren to between US$3.5 billion and US$8.5 billion. The loss estimates assume Zenkyoren holds a 60% market share of the total co-op market. Commercial (domestic) Our estimate of the insured losses to commercial properties from the combination of the earthquake and the ensuing tsunami is US$4.7 billion to US$11 billion. This estimate is based on the assumption that the industry losses will be split two-thirds to residential and one-third to commercial properties. Figure 2. Japan earthquake insured loss estimate (US$ billions) Low High Residential $9.5 $21.9 Commercial Life Marine Auto International insurance Total $20.0 $45.0 towerswatson.com 3

4 Life Because of the tremendous loss of life, life insurers face increased death, accident and health claims. Life insurance coverage in Japan is relatively straightforward, with one exception. An accidental death benefit rider, typically double the death benefit, is excluded by contract in the event of an earthquake. However, in accordance with local customs and concern for insureds families, the Japanese life insurance industry has announced that the earthquake exclusion will not be enforced for this event and that payments arising from the accidental death benefit coverage will be made. Our current estimate of the losses to life insurers is US$3.0 billion to US$4.9 billion. We note that our estimate does not contemplate any scenario where there is a significant release of radioactive material from the stricken Fukushima power plant, which could cause significant health and mortality issues. The level of uncertainty surrounding the earthquake s impact on the international reinsurance market is disproportionately influenced by time element coverage obligations. Marine The ensuing tsunamis tossed many large ships aground along with many smaller commercial and personal vessels. Losses to port facilities were also extensive. Our current estimate of the marine losses related to the event is US$1.1 billion to US$1.5 billion. Auto Many thousands of cars were flooded due to the tsunamis. Our current estimate of the losses to auto insurers is US$0.2 billion to US$0.7 billion. International Our current estimate of the losses to international insurers is US$1.5 billion to US$5.0 billion. One of the most uncertain elements of this catastrophe is the insured loss associated with large international (non-japanese) corporations. These companies will be primarily insured by major international insurers. Insurance protections often include time element coverages such as business interruption, extra expense and contingent business interruption (CBI) cover when suppliers can t supply needed production materials to customers. The CBI element is particularly difficult to estimate because the financial effects of supply-chain interruption will not be known for some time. The policy terms and conditions for time element coverages also vary considerably from company to company. While our estimate does include some level of losses from these types of coverages, we believe the range of potential outcomes for this portion of the insured loss is very wide. But we also believe that time element coverages will likely account for a significant portion of the impact on the international reinsurance market. This business is more heavily reinsured in both the treaty and facultative markets, which operate, respectively, by contract and the prerogative of the reinsurer s appetite for risk. The level of uncertainty surrounding the earthquake s impact on the international reinsurance market is disproportionately influenced by time element coverage obligations. 4 towerswatson.com

5 Insurance Industry Implications In contrast to other major catastrophe losses that have occurred in developed countries, a relatively small proportion of Japan disaster losses will affect the international reinsurance market. Comparing Disasters Hurricane Katrina Approximately US$65 billion of the roughly US$150 billion in economic loss (43%) caused by Hurricane Katrina was insured, and most of that amount was reinsured in the wider international market. Due to Japan s low take-up rates for earthquake coverage, we believe only approximately 10% to 12% of the estimated US$300 billion economic loss from the Japan earthquake will be insured. Of that US$20 billion to US$35 billion insured loss, we estimate that only US$12 billion to US$15 billion will be reinsured internationally, approximately 30% to 40% of the insured loss, or 4% to 5% of the overall economic loss. Consequently, the impact on the world reinsurance market, although significant, is unlikely to approach the level of Katrina. Christchurch Earthquake By contrast, insured losses from the February 2011 earthquake in Christchurch, New Zealand, are currently estimated between US$10 billion and US$12 billion. As with Katrina, the vast majority of this loss will be borne by the world reinsurance market. For some reinsurers, the New Zealand loss may equal or surpass that from Japan. Unlike other major catastrophes in the developed world, the low take-up rates for earthquake insurance and the propensity of the Japanese market to insure domestically means that the vast majority of the Japan loss will be absorbed within the country. The difference may be framed, in part, by a comparison of the two economies. Japan has a strong economy capable of rebounding, whereas New Zealand has a smaller economy and may have been better off transferring that risk. However, with a ratio of debt to GDP of approximately two to one, there can be no doubt that the impact of the earthquake on the Japanese economy will be unwelcome, at best. Overview So we do not believe that Japan s catastrophe will result in the global price hardening that occurred following Katrina. However, the cumulative impact of worldwide catastrophe losses in 2010 and 2011, combined with the release of the catastrophe model RMS Version 11.0, will likely result in some firming of catastrophe rates that will vary on a regional basis. After Katrina, catastrophe rates increased 50% to 100% for peak-zone exposures in the United States. However, several factors affecting rates in 2005 do not appear to be present now, at least to the same extent. In the wake of the three major hurricanes (Katrina, Rita, Wilma), there was the two-pronged recalibration of catastrophe risk by the major modeling firms and substantial increases in required capital for reinsurers promulgated by the rating agencies. Also, while most multiline reinsurers absorbed the catastrophe losses reasonably well, certain catastrophe specialists either exited the market or experienced financial distress. These factors caused a simultaneous increase in demand for catastrophe reinsurance with a decrease in the supply, resulting in higher market pricing. After Katrina, catastrophe rates increased 50% to 100% for peak-zone exposures in the United States. While some similar factors are present in 2011, other significant factors are not. Absent another event, we do not foresee the same level of price increases that we saw at the beginning of The losses are widely believed to be more of a significant earnings event rather than an impairment of capital. Additionally, though rating agencies may adjust their views of risk because of the international catastrophe experience, we do not expect the same type of changes witnessed in 2005 and Although it is early in the evaluation process, there is some evidence of price firming. The cumulative effect of events in Chile, Australia, New Zealand and now Japan are likely to cause some insurers to reevaluate both their diversification strategies and pricing levels. Their decisions will also likely be affected by the new RMS release, which has towerswatson.com 5

6 increased expected losses for U.S. hurricane risk. Finally, the required spreads in industry loss warranty and insurance-linked securities markets have also increased, and there is some anecdotal evidence of increased pricing for other catastropherelated insurance products. Although we do not believe that price increases will be nearly as dramatic as post-katrina, we do believe they will be noticeable. Market estimates include projected 20% to 50% rate increases for Japanese programs, likely increases in India (April 1 renewals) and Australia and New Zealand (July 1 renewals). Impacts on U.S. programs may be different, depending on the type of exposures reinsured, but many industry observers are expecting 5% to 15% increases for the upcoming U.S. hurricane renewal season (June 1 and July 1). Figure 3. Residential charts Overview of insured losses (US$ billions) $ Non-life/Government market losses (US$ billions) $ Total insured losses Non-life market (pre-jer cession Non-life market (post-jer cession Non-life market (post-jer cession) Japan government Co-op insurance losses (US$ billions) $ Co-op portion of total loss JER A Need for Better Distribution of Risk There are some important risk management lessons we can learn from the Japan earthquake, which has highlighted the need for tsunami risk coverage, the implications for business interruption coverage and prices, and the related supply-chain problems. For comparison s sake, Figure 4 shows insured and economic losses for the Japan disaster compared with other recent catastrophes around the globe. Tsunami Risk Coverage Given that the Japan earthquake was the second Pacific Ocean tsunami to cause significant damage in recent years, policyholders and insurers are more than likely thinking about whether they need tsunami coverage. Coverage definitions surrounding flooding and earthquake coverage will likely also be revisited, with special consideration given to yet-tobe-developed commercial tsunami models. Pricing of this exposure is clearly not very well developed and, consequently, so-called Ring of Fire coastal exposure properties enveloping the Pacific Ocean could eventually have insurance availability problems similar to Florida coastal properties. Business Interruption Coverage Additionally, business interruption insurance is not used nearly as much in Japan as in Western markets. Most covered business interruption losses are for Western interests insured in Japan or overseas from the loss of parts affecting production. Insurers would be smart to consider revisions in limits or coverage to help minimize or quantify these exposures. And while many large manufacturers have already invoked business continuity plans to overcome short-term supply issues, concern is rising over the need for medium- to long-term plans to overcome a global shortage of certain products. In particular, some semiconductor manufacturers have confirmed they are unsure how long existing inventories of materials will last or how logistics, power or staffing disruptions will impact supplies. Fuel shortages in Japan are significantly disrupting logistics, which is hampering alternative supply routes Co-op portion of total losses Zenkyoren s portion Remaining co-ops 6 towerswatson.com

7 The United States Geological Survey (USGS) warns there is a 99% chance of having one or more magnitude 6.7 earthquakes in the California area. The likelihood of an even more powerful quake of 7.5 is 46%. The time frame for both events is within the next 30 years. A Global Approach Towers Watson also believes that taking a more global approach to insurance, reinsurance and capital markets can increase risk-sharing efficiency among the world s peak catastrophe zones and other markets with concentrated risk exposures. As we noted, a key difference of this event is the Japanese market s retention of most risks within its market, a stark contrast to how risk is spread in New Zealand. Taking a more global view in providing coverage to peak catastrophe zones would send appropriate price signals to local markets and increase global capacity for the private market to take a larger share of the economic loss associated with major catastrophes. While catastrophe models do exist across Asia, most do not have the rigor of the U.S. and Japanese models. Understanding and improving upon these models is paramount as companies look to transfer or assume risk in the broader Asian market. With the growth of Asian economies and increasing insurance penetration, Asia may someday make us rethink our traditional views of peak catastrophe exposure zones. Are We Prepared? The United States Geological Survey (USGS) warns there is a 99% chance of having one or more magnitude 6.7 earthquakes in the California area. The likelihood of an even more powerful quake of 7.5 is 46%. The time frame for both events is within the next 30 years. We have already withstood two significant earthquakes in recent memory: the 1989 Loma Prieta 6.9 magnitude earthquake and the 1994 Northridge 6.7 magnitude earthquake. The USGS s predictions and California s earthquake history make it worth examining how we would handle a significantly larger event, especially with the difference in earthquake preparedness and mitigation when compared to Japan. Figure 4. Comparison of economic and insured losses for recent global catastrophes (US$ billions) Economic Insured Event loss loss Northridge earthquake $49 $15 (1994) Kobe earthquake (1995) World Trade Center attacks (2001) Indonesia earthquake 10 1 and tsunami (2004) Hurricane Katrina (2005) New Zealand earthquake (2011) Japanese earthquake $300+ $20 $45 (2011) towerswatson.com 7

8 For more information, contact your local Towers Watson consultant or: Americas Matthew Ball Christopher Bozman Michael Christian Ed Hochberg Asia Pacific Ron Kozlowski Chiaki Tanaka chiaki.tanaka@towerswatson.com Europe Cormac Bradley cormac.bradley@towerswatson.com Robert Brown robert.brown@towerswatson.com James Hole james.hole@towerswatson.com François Morin francois.morin@towerswatson.com About Towers Watson Towers Watson is a leading global professional services company that helps organizations improve performance through effective people, risk and financial management. With 14,000 associates around the world, we offer solutions in the areas of employee benefits, talent management, rewards, and risk and capital management. Copyright 2011 Towers Watson. All rights reserved. TW-NA towerswatson.com