Cyber exposure & the affordability crisis

Cyber exposure & the affordability crisis

By December 17, 2019Climate Risk, News

Welcome to the sixth edition of the Finity Climate Risk Blog, which focuses on cyber exposure as a component of transition risk and how recent assertions of an affordability crisis compare to premiums being charged in the market.

Will decarbonisation lead to new sources of risk

On 22 November a startup green energy company backed by Bill Gates called Heliogen announced that it had found a way to use Artificial Intelligence to coordinate thousands of mirrors to reflect enough sunlight in a concentrated area to create sustained temperatures of over 1,000C. This means that solar energy can be directly used to create the very high temperatures needed for production of things like steel and cement, which currently require large amounts of fossil fuel powered electricity. In Australia metal and cement manufacturing totalled 8.4% of electricity consumption in 2017[1] .

Heliogen illustrates how the transition to a low carbon economy will depend on advanced software, sensors, and other electronics, often connected to the internet. Other examples include:

  • Smart Grids, defined as a “self-sufficient electricity network system based on digital automation technology for monitoring, control, and analysis within the supply chain”.
  • Electricity Load (Demand) Management, where customers allow electric utilities to dial back use of things like air conditioning compressors during peak periods.
  • Fleet Fuel Management, where cloud based systems monitor vehicle fleet activity and recommend fuel saving routing.
  • Smart home devices which allow remote monitoring of heating/cooling systems in buildings to improve energy use.

These and other similar technologies promise to increase the viability of renewable energy and improve efficiency. But as energy, transport, and manufacturing come to increasingly rely on advanced technology and the “internet of things”, risk managers must assess whether new sources of cyber risk are being created which are beyond those commonly thought of. The problem is exacerbated by the use of many components such as sensors and circuit boards which may not all be hardened against malicious actors.

While much of today’s focus in cyber risk management involves preventing hackers from stealing information, in the future we will need to pay more attention to exposure to cyber attacks on infrastructure. While this issue has been present for some time, aggressive decarbonisation will accelerate the emergence of risks that may not have been considered previously. Cyber risk should be a consideration in assessments of transition risk.

"Uninsurability" - a study based on insurer premiums

In our last climate blog we looked into the issue of climate change and insurance affordability, presenting research on how current premiums in Northern Australia compare to disposable income. In this blog we examine how recent media reports of “uninsurable” properties compare to actual premiums being charged in the market.

In October, an ABC News report[2] triggered considerable controversy by saying exposure to losses from natural perils means “…some 370,000 properties [are] already effectively ’uninsurable’”. The story cited work by XDI (an affiliate of Climate Risk), which had concurrently released a report entitled “Climate Change Risk to Australia’s Built Environment”[3].

The XDI report characterized 383,300 properties as being “high risk” currently (about 4% of all standalone homes nationwide). This was based on XDI’s calculation of a “Total Technical Insurance Premium” (TTIP)[4] for selected perils (riverine flooding, coastal inundation, forest fire, subsidence, and wind; similar to average annual loss, or AAL) and a definition of “high risk” as where the TTIP exceeds 1% of the replacement cost based on a building sum insured of $300,000.

The ABC News story went on to quote Dr. Karl Mallon of Climate Risk as saying “If banks start to screen mortgages … we could see a very rapid, very sudden and relatively significant change in [property] values…And we don’t think this market adjustment will occur in 10 or 20 or 30 years when these hazards become a real problem … People are making these decisions today.”

We wanted to test whether the view of current risk implied by the XDI study was consistent with current insurance prices in the insurance market. To do so we collected over 7,000 online quotes from five leading insurers across the twenty Local Government Areas (LGAs) identified by XDI as having the largest number of high risk properties. We made sure our quotes covered an even geographic spread of the location of risks within each of those LGAs.

The quotes obtained were for total building premium charged to customers, which we then converted to AAL[5] and used this to calculate the proportion of “high risk” properties as defined by XDI. This provided a like-with-like comparison to XDI’s results, which is shown in Figure 1.

Our analysis suggests there is a significant discrepancy between how the insurance market is actually rating household risk and the information put forward by XDI. For example, the XDI analysis indicated that 11% of Gold Coast Properties were high risk whereas we found that less than 1% of market quotes were being rated as high risk. Averaged across all of these 20 specific LGAs, while XDI identified 6% of properties as “high risk”, the market quote data showed only 1% as “high risk”.

Without us being privy to the underlying details of the XDI modelling it is difficult for us to explain why the results of the XDI study differ so significantly from our analysis of market quotes. Of course there is always room for some sampling error in our approach; however, we think it quite unlikely to explain the magnitude of the difference observed. Our conclusion is that the insurance market, which certainly has access to an array of sophisticated perils risk pricing tools of its own, is currently pricing for significantly less risk than XDI’s figures imply. In our view more research is needed before clear conclusions can be made about the scale and specific location of the home insurability problem in Australia.

                Figure 1 – XDI Analysis Comparison to Insurer Quotes

In our most recent blog we highlighted the insurance affordability issue in parts of Australia when we examined the relationship between premiums and disposable income. So we certainly agree it is an issue; however, we think based on the above analysis of market premiums that the severity of the problem is most likely overstated by the XDI study.

Insurers have a lot of incentive to understand the impact of natural catastrophes and to reflect those risks in the premium rates they offer, and we continue to see this dynamic playing out across the more exposed parts of Australia. Looking forward we can expect the understanding of climate risk and its significance to further increase, and to continue to drive insurance losses and premiums up in various parts of Australia, inevitably leading to growing affordability issues and hence remaining an important community issue.

Did you know?

According to the US NOAA Arctic Report Card released on 10 December, “Thawing permafrost throughout the Arctic could be releasing an estimated 300-600 million tons of net carbon per year to the atmosphere.” By comparison, Australia’s emissions were 538.9 million tons over the year ending March 2019[6].

Climate Risk Team

Rade Musulin, Shirley Chau and Stephen Lau from our Climate Risk Team

[1] From Australian Energy Update, Department of the Environment and Energy, 2018
[2] See “The rise of red zones of risk”, ABC News, 23 October 2019
[3] Climate Change Risk to Australia’s Built Environment, XDI, 2019
[4] Defined by XDI as “the total annual cost of damage”
[5] By removing taxes, levies, expenses, other losses (e.g. theft, liability, etc.), and profit.
[6] See “Australia's carbon emissions continue to rise despite Government assurances about climate change policy”, ABC News, 30 August 2019