More fire, more rain, more risk. The severity and frequency of bushfires and riverine floods are increasing for Australia. The floods that struck southeast Queensland and New South Wales (NSW) in the first quarter of 2022 were the costliest in the nation's history, according to the Insurance Council of Australia. They were soon followed by further floods and torrential rain across southeastern and northwestern Australia in late 2022 and into 2023. For insurers, governments, and banks the threat is not new. But we believe the physical risks and second-order economic knocks from weather events could damage their creditworthiness in our downside scenarios.

Hypothetical Scenarios: Weighing The Risks

Our scenario analysis suggests that our ratings on many insurers, governments, and banks would be at risk if the direct financial impact from bushfires and floods were two to three times worse than in the past two decades, or if major disaster events were to become recurrent.

Importantly, our sensitivity analysis is hypothetical and seeks to test the level of stresses these sectors can withstand from any factor, including weather. We have not attempted to model a weather-related downside scenario that applies uniformly to all three sectors. Rather, our scenarios are based on the equivalent of, or multiples of, the financial impact of the worst historical weather events, or the global financial crisis in the case of banks.

INSURERS

A Fortified Line Of Defense

Our analysis indicates property and casualty (P/C) insurers in Australia have conservative reserving and appropriate reinsurance coverage, including protection against natural perils, such as bushfires and floods. Insurers' exposures are greatest in the personal lines of home and contents and motor vehicle insurance but also extend to commercial lines. Collectively, these insurance lines accounted for about 65% of gross premiums written for the year to Sept. 30, 2022.

The severity and frequency of declared catastrophes has trended upward in Australia over the past 12 years (chart 1). The largest bubble, which depicts insurance claims paid in the three years to November 2022, includes the recent floods that affected southeast Queensland and NSW (CAT 221)--with the claims totaling A$5.6 billion. This was the largest event on record for Australia, in nominal terms.

Chart 1

Rising Claims Are Fueling Premiums And Reinsurance Costs

To date, offshore reinsurers have absorbed much of the cost of natural perils in Australia. Higher declared catastrophes and significant events--including a series of bushfires, floods, and hailstorms--have affected gross claims, particularly since September 2019. Reinsurance cover has grown to soak up 35%-40% of the gross claims cost in recent years, which has cushioned the blow to insurers' earnings (chart 2).

Chart 2

The cost of reinsurance has increased at record pace. Reinsurance costs for Australian insurers have risen by about 35% over the past three years alone, and outweighed gross written premium (GWP) growth of about 30% over that period--squeezing P/C insurers' underwriting earnings (chart 3). The higher claims experience borne by reinsurers has meant higher pricing for cover, as well as stricter terms of cover provided, such as increasing insurers' retention of risks.

Australian insurers are responding by setting higher allowances for losses--and this is growing faster than premiums. P/C insurers are also adjusting their risk appetite and enhancing risk exposure assessments (such as exposure to flood). The Australian government has implemented a A$10 billion cyclone cover, or related damage, reinsurance pool, to support capacity and affordability.

Chart 3

Higher-Claims Scenario Will Challenge Insurers' Creditworthiness

P/C insurers have breathing space. In our base case, on average, their underwriting profits in the next two to three years should be able to absorb additional catastrophe claims equivalent to 1.7x times the highest of the top three catastrophe claims in recent years (table 1). Our ratings on insurers could face downward pressure if additional claims exceed this level, or remain consistently high, surpassing the worst seen in the past.

Over the previous five years, catastrophe claims peaked at about 12.2% of gross claims, while the top three net claims represented about 35% of total net catastrophe claims (for Australia's largest insurers). To approximate reinsurance cover, we've increased the top three net claims by 38%--which is the average reinsurance benefit realized over the past five years--to arrive at a gross claims figure. While the reinsurance cover provided to large risks is typically greater than the average, the scenario provides insight into the buffer available from underwriting earnings relative to catastrophe claims.

On this basis, the absorption of additional catastrophe claims, if they occurred, would be viewed as an "earnings" event, with capital resources providing a supplemental buffer. By way of reference, the average gross catastrophe claims for the past 10 events (excluding CAT 221) was about A$470 million. Insurers' exposure subsequently decreased where reinsurance cover was effective.

Table 1

GOVERNMENTS

Strong Balance Sheets Will Help Governments Remain The Ultimate Backstop

Government plays a critical role in recovery from natural disasters. At the onset, authorities offer cash payments, concessional loans, or tax relief to households and small businesses that require immediate support. Then comes rebuilding damaged public infrastructure, such as state-owned roads, bridges, and schools. In general, constitutional power for emergency management sits with the states.

The financial safety net for states is generous. The Australian government typically reimburses to them 50%-75% of the costs of reinstating vital infrastructure--under the legislated Disaster Recovery Funding Arrangements (DRFA) and its predecessor, the Natural Disaster Relief and Recovery Arrangements (NDRRA). The 50% reimbursement rate kicks in once disaster spending exceeds a threshold of just 0.225% of state revenues, and the 75% reimbursement rate activates at 1.75x the first threshold (or about 0.4% of state revenues).

States are also required to carry "cost effective" insurance--usually via government-owned captive insurers, often supplemented by private sector insurance or backed by reinsurance--as a condition of eligibility for DRFA. However, road and bridge infrastructure are often excluded, for cost reasons. Traditional reinsurance markets may also be closed to certain perils at certain times. For instance, no reinsurers would provide quotes to the Queensland Treasury for coverage of roads in 2011, in part due to adverse claims experience and data limitations.

In addition, when a state spends more on natural disasters than is funded by DRFA, its share of national goods-and-services tax grants will improve in future years under Australia's system of horizontal fiscal equalization. State governments may provide ad hoc post-disaster assistance beyond the usual DRFA, for instance via tax, rate, and levy relief.

Worsening Weather Could Threaten Governments' Credit Standing

All things being equal, our sovereign credit rating on Australia (AAA/Stable/A-1+) could come under downward pressure if the costs of natural disasters are more than three times as damaging as those in 2021-2022 and were to recur over multiple years. That said, we consider that the Australian sovereign is well prepared to weather its increasing exposure from wildfires, floods, storms, and sea level rise (see "Weather Warning: Assessing Countries' Vulnerability To Economic Losses From Physical Climate Risks," April 27, 2022).

The reported fiscal cost of historical natural disasters to the Australian sovereign has been small. This includes DRFA and NDRRA assistance to the states and direct payments to individuals through the Disaster Recovery Payment and Disaster Recovery Allowance.

Total outlays averaged less than A$2 billion a year over the past decade (chart 4). Even in the peak catastrophe season of 2010-2011--in which most of Queensland was declared a disaster zone due to heavy flooding--relief expenditure peaked at A$6.1 billion. More recently, the 2021-2022 fiscal year saw relief expenditure of A$5.8 billion, equivalent to about 0.9% of the central government budget or 0.3% of GDP. A sustained and unaddressed increase in the annual deficit of more than one percentage point of GDP, relative to our base case, could weigh on the 'AAA' rating on Australia.

Chart 4

States Face Greater Risks From Localized Disasters

When a natural calamity is localized in a particular region, the relative fiscal shock at state government level can be more material. Consider Queensland, which has the highest per capita spending on natural disaster relief of all Australian states (chart 5). Its cash operating position briefly dipped into deficit after the 2010-2011 floods (chart 6). The lag effect was due to NDRRA funding paid in 2011-2012, in advance of recovery spending peaking in 2012-2013. The net fiscal cost to Queensland, after accounting for NDRRA reimbursements, was about A$3.5 billion, equivalent to a hefty 7% of annual nonfinancial public sector operating revenue. Nonetheless, this event did not hurt our rating on Queensland because of its one-off nature.

Chart 5

Chart 6

The credit rating on Queensland (AA+/Stable/A-1+) would come under pressure if the state were to incur continual and unaddressed operating deficits. In our base case, we project that Queensland will achieve operating surpluses of about A$2 billion-A$3 billion over the next few years. A downgrade to 'AA' could loom in the unlikely scenario that events as damaging as the 2010-2011 floods were to recur for, say, three or more consecutive years.

Sound Financial Management May Protect Government Creditworthiness

The overall impact of natural disasters on credit metrics will hinge on how governments adjust their fiscal strategies. Given Australia's strong institutional settings and many states' self-imposed fiscal principles, we see persistent operating deficits as unlikely even in our downside scenario.

For example, in 2011 the central government imposed a one-off levy on middle- and high-income earners to raise A$1.8 billion, enough to recoup about one-third of its NDRRA outlays. In Queensland, the post-flood rebuilding years coincided with an independent commission of audit, which resulted in cost savings from a large public servant redundancy program.

There are also innate market constraints on spending money. East coast states currently struggle to deliver their large infrastructure budgets in the face of labor and material shortages. As such, we expect capital expenditure on disaster recovery this year to displace other lower-priority projects, minimizing the impact on states' overall cash deficit positions.

Damage To The Economy Would Have Second-Order Effects On Governments

It's not just direct fiscal costs that governments face from floods or forest fires. A blow to the economy from natural disasters would inevitably dent taxation revenues and fuel demand for higher social spending. In our downside scenario, however, we see a diminution in economic strength of Australia's central or state governments as unlikely over the next two to three years. This is because we assess the economic losses from even severe natural disasters to be digestible in the context of a wealthy and diversified economy.

According to the Australian Treasury, the Queensland floods of 2010-2011 subtracted a relatively manageable three quarters of a percentage point from real GDP growth that fiscal year, driven by production losses in the mining and agricultural sectors. Floodwaters also inundated coal pits, leading to a reduction in export earnings, though this was partly offset by higher global coal prices.

The direct effects of the 2019-2020 "black summer" bushfires reduced real GDP growth by about 0.2 percentage points across the December 2019 and March 2020 quarters, according to estimates by the Reserve Bank of Australia. The more recent floods across southeastern Australia will detract a quarter of a percentage point from real GDP growth in the December 2022 quarter, according to estimates by the Australian Treasury.

It Pays To Spend On Prevention, Not Clean-Up

Australia invests little in adaptation to acute weather events and climate change. Indeed, over 2002-2014, about 96% of central government spending on disasters was allocated to ex post response efforts and just 4% to ex ante disaster risk reduction (table 2). This imbalance may be gradually shifting. In 2022, the central government announced a Disaster Ready Fund (rebadging the former Emergency Response Fund) to provide up to A$200 million a year for projects such as flood levees, cyclone shelters, fire breaks, and evacuation centers.

State governments are acting, too. Both NSW (Northern Rivers Resilient Homes Fund, A$800 million) and Queensland (Resilient Homes Fund, A$741 million) recently announced programs to retrofit, raise, or buy back eligible properties in flood-prone areas. The states are also investing in more disaster-resilient infrastructure in their multiyear capital works budgets.

Table 2

Government policy in insurance markets can influence disaster resilience. For instance, numerous independent tax inquiries have found that state-based insurance levies and stamp duties make insurance cover less affordable. However, state governments have had little political appetite to phase out these lucrative taxes without an obvious source of replacement revenue.

A new Northern Australia Cyclone Reinsurance Pool, which commenced operation in July 2022, is designed to be notionally cost-neutral to the Australian government but saddles it with a new contingent liability in the form of a A$10 billion guarantee. We see some risk that political pressures will impel the government to expand this pool's coverage over time, expanding its role as insurer of last resort. The standing nature of the DRFA means it is also an explicit, but unquantifiable, contingent liability of the sovereign.

BANKS

Headroom But Points Of Vulnerability Too

An increase in severity, geographic spread, and frequency of weather events could hurt bank creditworthiness through several channels, including:

• Increased credit losses (or charges for bad and doubtful debts) from physical risks to assets of borrowers
• Loss of interest income as nonperforming loans rise
• Loss of fee income due to reduced economic activity
• Reduced opportunities for profitable lending and other business
• Increased overall economic and industry risks

Most of these rising risks would emerge as the second-order impact on the operating conditions for banks.

In our base case, we forecast Australian banks' pretax earnings before credit losses in 2023 at 1.5x the highest level of credit losses in the past two decades. Bad and doubtful debts in the year to June 30, 2009, rose to 93 basis points (bps) of customer loans, largely reflecting the effects of the global financial crisis.

Relatively localized damage from floods and bushfires, sound lending standards, and insurance coverage of physical risks have meant that natural disasters have had no noticeable impact on banks' credit losses. For example, Australia suffered severe flooding and bushfires in the months before the breakout of COVID-19 in the first quarter of 2020. Nevertheless, there was virtually no jump in bank credit losses or impaired loans until the breakout of the pandemic.

Furthermore, even following the outbreak of COVID-19, banks' credit losses rose to a (relatively modest) short-term high of only 40 bps of customer loans in the year to Sept. 30, 2020. Consequently, we have used the year to June 30, 2009 as the reference point in setting our downside scenarios for this analysis.

Chart 7

Bank Credit Quality Can Weather Moderate Downside

Our sensitivity analysis suggests the banking system will remain profitable in a hypothetical downside scenario where credit losses and non-accruing loans rise to as high as those in 2009 during the global financial crisis. In this downside scenario, we also assume the noninterest income to be 25% below our base case. Such a scenario--provided it's a one-off--would probably have limited effect on most banks. However, sustained credit losses at this level would likely erode systemwide creditworthiness and that of individual banks.

Table 3

Downgrades And Hit To Capital Base: Contemplating A Severe Downside

Most Australian banks would incur large statutory losses under a severe downside scenario in which charges for bad and doubtful debts are double those seen in 2009, the nonaccruing loans rise to 5% of total loans and advances, and the noninterest income drops to half of our base case. Ratings on many banks would be at risk even under a one-off event of this type. Furthermore, most banks would almost certainly face downgrades if such events were to become recurrent.

Under this hypothetical scenario, the pressure to our ratings on banks would come from several quarters within our framework for rating banks. Most of these rising risks would emerge as the second-order impact on the operating conditions for banks:

• Combined pre-tax losses of the banking system in a single year would be about 13% of its total tier-1 capital base, in the absence of additional capital injections. If this scenario were to occur two years in row, capital levels of most rated banks would fall below the thresholds consistent with our current ratings on them.
• Higher credit losses in the system and likely weakened economic resilience would very likely point toward increased economic risks within our Banking Industry Country Risk Assessment (BICRA) analysis for Australia. That, in turn, could mean a lower anchor, or a weaker standalone credit profile for an average bank in Australia. In addition, we would apply higher risk weights in our capital analysis, which would depress capital ratios of banks under our framework.
• If weak earnings became entrenched within the banking system, they would suggest weaker ability of banks to price for risks, and cope with any other unexpected losses. We would likely assess such conditions commensurate with higher industry risks within our BICRA analysis for Australia.
• A rise in credit losses and nonperforming assets, as well as other signs of deterioration in Australian banks' creditworthiness, could also make them more vulnerable to a disruption in their access to funding, especially from outside Australia.

Insurers, Governments, And Banks Face Sterner Tests

Australia's weather bill is set to climb. The country's institutions--both public and private--are robust but they face more tests. Their ability to prepare for and respond to natural disasters will in turn have a bearing on their creditworthiness and the ratings assigned to them.

Related Research:

• Australia's Insurers Won't Need To Tap Reinsurance After Floods, Oct. 18, 2022
• Weather Warning: Assessing Countries' Vulnerability To Economic Losses From Physical Climate Risks, April 27, 2022
• Australia Sovereign And State Ratings Can Accommodate Bushfire Impact, Jan. 13, 2020

Editor: Lex Hall

Designer: Halie Mustow