"Climate Change is no longer some far-off problem; it is happening here, it is happening now" – Barack Obama.(1)
In recent times, the importance of the phrase ‘Climate Change’ has grown significantly. It represents the long term, large scale changes in the Earth’s average temperature and weather patterns. There are beliefs that climate change is a naturally occurring phenomena occurring due to cycles arising from variations in the Earth’s Orbit around the Sun(2) . However, there is no shortage of scientific evidence proving elements of climate change to be caused by the very actions of us humans.
Burning fossil fuels, deforestation and farming livestock are just a few of the many ways that humans are increasing the concentration of greenhouse gases in the atmosphere. This in turn is causing a gradual increase in the average temperature of the Earth, also known as Global Warming. The last three decades have been warmer than the preceding decades for the first time since records began in 1850. The planets population is growing at an exponential rate meaning there are more mouths to feed through farming livestock, more fossil fuels to burn (for example, more car journeys), and an increasing rate of deforestation, which in turn will continue contribution to global warming. The increase in temperature is leading to an increase in extreme weather conditions and natural catastrophes such as hurricanes, earthquakes.
To put the issue into perspective, according to Swiss Re, 2015 had the highest number of Natural Catastrophe’s in a year (198), which displaced the previous highest number of natural catastrophes (191) in 2014(3) . Having two consecutive record-breaking years in a row for natural catastrophes cannot be a coincidence. There are devastating instant effects of these catastrophes such as destroyed homes, destroyed infrastructure, shortage of food and water, and even loss of life. Furthermore, there are also secondary effects such as economic loss due to inability to work, unemployment, direct costs of rebuilding the homes and area and destroyed land for farming. Shockingly, during the aftermath of Hurricane Katrina, criminal gangs roamed the streets stealing from vulnerable homes and businesses, which further contributes to the devastating effects of climate change(4)
Around the planet, there are certain areas that are hotspots for natural catastrophes (i.e The Americas and Caribbean for Hurricanes, and Asia for Earthquakes) SOURCE; for those living these areas, the natural catastrophes aren’t one-off’s but recurring occurrences. The sheer economic and social costs of one of these natural catastrophes is huge problem on its own, but what happens when areas are in constant risk of these disasters?
Insurance provides the residents, businesses and governments in the high-risk areas with a safety net by spreading the risk of a natural catastrophe. Without the insurers, the area would be inhabitable due to the huge risk of natural catastrophes occurring, along with the steep costs of rebuilding the area, as these costs would be too high for individuals to pay. Furthermore, hypothetically, if an individual could afford to rebuild what they owned, the rest of the area would be destroyed, meaning the individual would not be able to purchase or grow food, work or even travel due to the destruction of infrastructure. Multiply this across natural disaster-prone areas, and the economies would be incredibly unsustainable, and crash over time, whilst the social costs and infrastructure damage would be catastrophic and costly. Insurance is a key competently in the economy in order for it to run smoothly and efficiently.
In a time when the characteristics of more frequently occurring natural catastrophes are becoming increasing difficult to predict; the methods that insurers use to capture this risk and unpredictability are paramount in ensuring the insurers remain solvent. This dissertation will give readers an insight into the causes and effects of climate change, discuss the methods used by insurers to insure against natural catastrophes and explore the practices of reinsurance and retrocession, before critically analysing the effect climate change is having on the UK Home Insurance industry.
Climate change and Insurance
In this section, we will discuss evidence behind climate change and natural catastrophes, and what is causing them. Furthermore, the relationship between insurance, reinsurance and retrocession will be looked at, whilst
2.1 Climate change – History and Evidence
However, as shown in figure 1, the levels of atmospheric Carbon Dioxide had never passed the 302 parts per million level until 1950’s, which was after the Industrial Revolution.(5)
The industrial revolution, starting in the 18th century in Britain before spreading to the Americas and then the rest of the world in the 19th Century, is the time period that oversaw the industrialisation, or the transformation of an economy based mainly on handicrafts and agriculture to an economy dominated by industry and machinery(6) . This happened due to several factors such as growing population (workforce), financial innovations (banks, stocks), agricultural revolution (new techniques to harvest food) and scientific revolution (better understanding of scientific principles and mechanics).
This effectively led to an increase in the demand and use for fossil fuels (hydrocarbons formed over time from the remains of plants and dead animals), mainly coal, natural gas and oil (Figure 2)(7) . Coal was one of the first major energy sources and was used to power steam powered engines for trains and ships, as well as power fuel furnaces in the 1700’s, whilst the development of combustion engines in the mid 1800’s and the more efficient kerosene powered lamps increased use of oil. After this, due to the abundance of natural gas, it was initially used to power street lamps, but the development of gas pipelines in the late 1800’s allowed natural gas to be accessible for new and larger markets. These fossil fuels are still being used to this day.
The combustion of fossil fuels releases carbon dioxide in the air which contributes to the greenhouse effect: the ability of greenhouse gasses (methane, carbon dioxide, nitrogen) to reflect radiation within the earth’s atmosphere. The radiation is reflected back into the atmosphere due to there being increased levels of reflecting greenhouse gases in some parts of the earth’s atmosphere.
This causes slight differences in the amounts of Solar energy received on the planet by the sun; which in turn can have significant impacts on the features on the planet. For example, glacial advance or retreat (forming or melting of the ice caps) which in turn may cause rising or falling sea levels, leading to a greater surface area of water on the planet, which can then cause increased rainfall. Furthermore, rising temperatures also cause expansion of the seawater, which could also cause flooding by either covering more land mass, or increasing rainfall. These changes in the behaviour of the weather can have detrimental effects on populations around the world.
This is illustrated by the alarming fact that between 1901 and 2012, the Earth’s temperature has increased by 0.89 degrees Celsius(8) , whilst the temperature increase over the last 5000 years before this was roughly 4-7 degrees Celsius(9) . One of the consequences of Climate Change is an increase in the quantity and scale of Natural Catastrophes, defined as a loss event caused by natural forces (Swiss Re). Examples of these include floods (arising from storms such as Edna in the UK), Hurricanes (Katrina and Andrew in the US) and Earthquakes (especially in the Far East Asia region). Science has shown that
Insurance and Reinsurance
Insurance is when one party (insurer) agrees to pay an indemnity to another party (insured) the occurrence of a predefined ‘unwanted’ random event (e.g. Natural Catastrophe) generating damage or catastrophe for the latter (10). The insurer therefore receives payments in the form of a premium and is compensation for the risk ceded from the insured to the insurer. Insurance can be split into two general areas; personal, where the individual and Commercial. Furthermore; Insurance can also be split into property, pecuniary, motor, liability, marine & aviation, life assurance and health insurance.
As mentioned earlier, with the rate of catastrophes gradually increasing, it is vital for insurers to be able to effectively predict the frequency of these catastrophes in order to calculate premiums to reflect the potential claims that may arise in the insured period. Therefore, insurers use various statistical and actuarial techniques, as well as Catastrophe Models in order to calculate appropriate premiums to offset the risk of paying out in the event of a claim. Furthermore, being a business, insurers also need to be profitable to survive in the industry.
The shifting of risk is paramount in the efficient functioning of our economy. Insurers will have pre-determined acceptance limits, which are levels of risk that they are willing to undertake. In order for an insurer to stay within their risk limits, they must ensure the pool of premiums for each class of insurance they are covering is adequate in order to cover any future possible claims. Therefore, if an insurer is offered a risk that is greater than the acceptance limits it has set, then they theoretically could just decline to provide a quotation. However, with the level of competition in the market, as well as basic business principle, it would be unacceptable to decline good business in that manner purely because of size.
A fully efficient insurance market is an extreme hypothetical concept and means an acceptable amount of cover is provided to all within the market at an acceptable premium due to the law of large numbers. However, the presence of informational problems such as inaccurate modelling of the risk, solvency issues arising preventing full loss payments and fraud can cause competitive insurance markets to become inefficient. Additionally, there is also a trade-off in allowing efficiency as having a large number of competitors in the market weakens the law of large numbers by diminishing average insurance reserves.(11)
Reinsurance, a key tool to increase the efficiency of the market, is an insurance contract in which a party (known as the reinsurer) agrees to indemnify another party (the reinsured – a first line insurer or another reinsurer) for specified parts of the underwritten insurance risk it has undertaken. In return, the reinsured party (cedent) pays the reinsurer a reinsurance premium as monetary compensation. This practice can therefore be looked at as the insurance principle but lifted up one level, so the first line insurer is the party in need of an insurance contract.(12)
Whilst Reinsurance is often seen as a form of Insurance, and they both share several similar characteristics, it is important to note that Reinsurance does have numerous distinct traits. Reinsurance is a form of risk sharing between two ‘professional’ insurance entities, so will have tend to have tighter regulation and guidelines to adhere to. The types of data available for risk analysis can also differ significantly, i.e. historical data for first line insurers is scarce, but there is a much smaller limit for data for reinsurers. The magnitude and classification of risks being undertaken can also greatly differ from first line insurance.
Insurance companies have a primary purpose to adequately take on risk within the economy, which is similar to financial organisations. In both scenarios, capital provided by shareholders is leveraged in order to raise debt. However, financial debt tends to have a pre-determined expiry date, as well as a known face value (severity), whereas insurers raise this debt by selling policies to those requiring cover in return for a premium, making it a riskier debt due to the uncertainty around the severity and timing of the claim. For example, property insurance claims due to natural catastrophes such as flooding tend to come in large numbers due to the number of policy holders being affected.
There are two main methods that can be used to cede the risk between insurers and reinsurers; one is called ‘Quota Share’ reinsurance whilst the other is called ‘Excess of Loss (XL)’ reinsurance. Both methods can be applied to claim distributions.
Quota Share Reinsurance:
Under a Quota Share reinsurance scheme, all claims and premiums are shared in agreed proportion by the insurer and the reinsurer:
X Ã claim amount
Î± =Ãproportion retained by insurer (0â‰¤Î±â‰¤1)
Y=Î±X Ã the amount the insurer pays
Z=(1-Î±)X Ã the amount the reinsurer pays
Excess of Loss (XL) Reinsurance:
Under an XL Reinsurance scheme with retention level M, when a claim occurs:
Y=minâ¡(X,M) Ã the amount the insurer pays
Z=maxâ¡( 0,X-M) Ã the amount the reinsurer pays
Insurance linked securities
Other investment related vehicles – relate all to retrocession, reinsurer and insurer financial markets
Catastrophe Models (or Cat Models) are computer programs that mathematically represent the physical characteristics of catastrophes such as terrorism, extreme casualty events and natural catastrophes(13) . The concept, developed in the late 1980’s, provides various organisations and institutions (such as insurers) the ability to assess the likelihoods and therefore, severity, of future catastrophes in order for them to be able to adequately prepare for the financial impact.
Prior to the development of Catastrophe models, simple elementary methods such were used to estimate catastrophe losses, albeit from a deterministic point of view. To forecast acceptable risk levels, underwriters would use spatial risk mapping (14) (the action of allocating risk indices to unit areas depending on their vulnerability to hazards) and the measurement of hazards to forecast the frequency of these catastrophes (but used each method separately). This was then used to generate simple Probabilistic Maximum Loss (PML) estimates which forecasted a worst-case scenario for catastrophe losses. Regression techniques and Actuarial Analysis was also used to statistically model the catastrophes based on historical data. However, the processes did not forecast frequency due to the lack of scientific input into the process (such as characteristics of disasters) meaning there was no way to forecast the timings of potential future catastrophes. Furthermore, historical data for ‘low-frequency, high-severity’ events such as natural catastrophes is scarce due to the variables in episodes of events, and in addition to this, the majority of limited available data available may be outdated due to huge advances in technology and catastrophe prevention measures such as flood barriers, and advances in communication in-between catastrophic events. Subsequently, AIR (1987), RMS (1988) and EQECAT (1994) recognised the need to create a dynamic forecasting solution which incorporated the exposure, hazard risk, vulnerability and loss, and therefore, the Catastrophe Model was born as we know it. Catastrophe modelling has now established itself as standard practice in the Insurance and Reinsurance sector and is also being adopted by various other areas such as the Government and the Financial sector.
3.1. A breakdown of a Catastrophe Model
A catastrophe model for a specific catastrophe needs to incorporate three key components; the hazard, the vulnerability/engineering component and a financial component, with an ideal output being a fully probabilistic loss distribution which represents an approximation of expected outcomes(15) .
The hazard component of a Catastrophe Model answers Catastrophe specific questions about the location, severity and frequency of the events in question. This stage requires scientific experts in fields such as meteorology, climate, seismologists, hydrologists and geophysicist, who are tasked with keeping on top of any research advances and scientific literature. Additionally, they are also required to contribute via research of their own to ensure the hazard inputs of the model remain as close as possible to the real event being modelled in order to ensure the parameters that will be inputted into the model are as accurate as possible. In the hazard component of the model, there are two key stages; Event Generation and Local Intensity Calculation(16) . Within the Event Generation sub-stage, using the various scientific parameters and characteristics relevant to the event being modelled, a large number (for example, 10,000) of simulations are generated in order to capture severity, frequency, location and other traits of potential future catastrophes. Following this, in the Local Intensity Calculation sub-stage, the intensity of the hazard is calculated at each affected site. In order to obtain a greater level of accuracy, digital elevation data (DEM) is used to understand the terrain of the area in question. This allows the simulation to further enhance scenarios. For example, the digital elevation data for land for a flood model would be greatly beneficial to understand which areas of land are more prone to flooding for reasons such as river banks and less absorbent land. The multiple simulations then consider a large catalogue of variables; for example, recent rainfall may indicate that the soil cannot absorb any more water in certain areas, causing rainfall in those areas to be a high contributor to flood risk. Also, certain types of land may be more impervious as opposed to others, so surface runoff may occur. Therefore, although Catastrophe Models are considered to be stochastic, the severity of certain natural catastrophes are also dependent on previous extreme weather patterns. The various event simulations are overlaid on hazard metrics in order to produce a metric used as an input in the next module(17) .
At this stage, the exposure data – highly detailed data representing the physical characteristics of the area or property being insured – is inputted into the model.
Next, the vulnerability component of a Catastrophe Model assesses the degree to which structures, their contents and other insured properties are likely to be damaged by the hazard previously modelled. There is a huge uncertainty in how structures and buildings may react to catastrophes, and it would be inefficient, costly and time consuming for insurance modellers to visit the premises of every site they insure in order to record parameters to predict potential damage, so they damage is described as an average using the exposure data. The metrics obtained from multiple hazard simulations of different intensities and characteristics are applied to the detailed exposure data along with damage functions in order to compute the expected level of damage to the different properties being insured. Damage functions are used to represent the relation between the maximal damage possible to a building, compared to the damage arising from the event the model is simulating. For example, for a multi-storey apartment block, a flood is only likely to cause damage to the basement and lower floor(s), so it wouldn’t be correct to assume the whole building will suffer damage due to flooding, so a function/factor is applied to reduce the asset value of the damage in the building. Damage functions are developed by highly trained structural engineers, by using laboratory testing, findings from on-site damage surveys, published research and detailed claims data provided by insurance companies. Individual damage curves can then be displayed on a single graph for categories such as areas, types of property and architectural styles on axis displaying the severity of the catastrophe on the x axis, and the damage ratio on the y axis(18) . DAMAGE RATIO DEFINITION – monetary loss.
Finally, in the Financial section of the module, the insurance and reinsurance practices are applied in order to calculate
Simulation – historical data – how many years? 50/100? – apply distributions – limitations? -EXTREMES?
From this we can deduce the Annual Average Loss (AAL) (also known as ‘pure premium’ or ‘burning cost’) as the mean of the distribution, and also the Exceedance Probability Curve (EP Curve) (19) EXPAND
FloodRe (Flood reinsurance program) Case Study/Factfile
Floods are the single largest natural catastrophes in terms of losses. They cause significant damage by sweeping away homes and buildings, destroying electricity supplies and causing death. The burden on the economy after flooding is huge, and with the increase in severity and frequency of floods, a large proportion of those effected are either uninsured or underinsured. It is estimated that one in six residential properties within the UK is at risk of flooding. The industry is trying to fill this gap by offering flood protection reinsurance programs which are either government led (non-profit organisation) or insurance companies(20). In this section, I will provide an in-depth look at FloodRe – a flood reinsurance program that was specifically created to address the flood risk in the UK.(21)
WHAT IS FLOODRE and HISTORY
Launched on 4th April 2016 after years of planning, FloodRe differs from other reinsurers as it is a not-for-profit company, and it is owned and managed by the insurance industry(22) . FloodRe was created by the government working closely with insurers to find a new way to deal with flooding and offer affordable flood cover on home insurance policies. Prior to its introduction, homeowners living in areas prone to flooding often found themselves paying high premiums and excesses when making claims in order to cover their flood risk; FloodRe was established to be a solution to this. Aiden Kerr (Operations Director at Flood Re – Access, not excess, May/June 2014 ) mentions that according to modelling carried during the scheme construction phase, up to 350,000 UK Households at risk of flood could be eventually ceded to the scheme over the initial four year period.
WHAT DOES FLOODRE DO AND HOW
The FloodRe scheme allows insurance companies to insure themselves via reinsurers against potential losses due to flooding by transferring the flood risk component of home insurance policies to FloodRe. Therefore, the scheme’s primary aim is to promote the availability of flood cover to individuals who live in or own properties that are prone to flooding, whilst ensuring that costs remain low. FloodRe receives two different revenue streams; firstly, Insurers pay FloodRe a set amount per policy taken on dependent on the council tax band of the property being insured, and secondly, UK home insurers that use the FloodRe scheme contribute towards a Â£180 million levy per annum. Using these two incomes, FloodRe purchases Reinsurance through financial investors within the Global Capital Markets to cover potential flood claims linked to those properties that are ceded to the scheme. The premiums set by FloodRe for insurers are low and affordable, and there is a Â£250 excess on every policy claim that goes via FloodRe, so the insurers themselves are able to reduce their charges towards their customers. It is key to note that FloodRe is designed to only cover residential but not commercial properties, as it was found that there was no such systematic issue within commercial insurance sector that left these properties at high risk of flooding.
FloodRe is beneficial to both insurers, as well as the insured; the insurers are covered against large potential claims due to flooding, whilst homeowners are able to cover their properties against flooding at a reasonable cost. Whilst helping homeowners, the scheme also helps tenants in finding affordable contents insurance if living in an eligible property in a flood prone area. In an economic context, the scheme has created a ‘level playing field’ for healthy competition between insurers offering policies to those with homes that are prone to flood risks.
The scheme due to run until 2039, and in order to gain the most from it, it is key that in that time, those benefitting from the low home insurance prices make themselves more aware with the flood risk they are exposed to, whilst also taking appropriate steps to reduce this risk if possible.
WHY DOES FLOODRE DO THAT
Only covers houses built before Jan 1st 2009 as it is important to not incentivise houses built in flood prone areas – principle – to incentivise rigorous and responsible planning decision
Show chart with habitable land – previously and now – and use this to explain that is the impact of climate change as significant as we think?
Also, climate change shows rising temperatures, so more precipitation, but also more evaporation so
Home insurance policies are done on a yearly basis so not much change per year as assumptions at start of policy remain the same through the year so change doesn’t have to be taken into account
Government programs like floodre have more than one benefit.