Tag Archives: Association of British Insurers

UK Floods: Don’t Say We Weren’t Warned (Part 2)

In my last post, I looked at the the recent history of climate change policy with respect to flood risk in the U.K.; in particular, a series of benchmark reports that starkly set out the flood risk choices the government and British people would have to take.

The realisation that climate change had transformed the flood risk game, together with the fallout from the 2007 floods as documented in The Pitt Review, led the Environment Agency to establish a long-term strategy (2010 to 2035), which was set out in a 2009 report entitled “Investing for the Future: Flood and Coastal Risk Management“.

The report identified 5.2 million households at risk of fluvial flooding (river and coastal), pluvial flooding (surface water and ground water) and both fluvial and pluvial.

Properties at Risk jpeg

And of the 2.4 million fluvial category, a more detailed breakdown in terms of expected frequency of flood was given.

Fluvial Flood Risk jpeg

Moreover, here are the critical regions for such fluvial flooding:

Properties at Flood Risk by Area jpeg

The media still reports the 5.2 million ‘at risk’ number of properties, despite climate science having moved on since 2009. Indeed, the above figures are all from the Environment Agency’s “Flooding in England: A National Assessment of Flood Risk” which came out in 2009 but is based on data for 2008. In my previous series of posts on flood risk, I pointed out that the current risk categories do not adequately take into account climate change or non-fluvial flooding (i.e., non-river and coastal flooding). Accordingly, all these numbers now look to be significant underestimates.

A few leaks have emerged with respect to the 2013 National Assessment (for example, in The Guardian here), but not enough to get a good impression of how the Environment Agency has moved the numbers around.

More controversially, the 2009 “Investing for the Future” report also gives us some cost-benefit estimates out to 2035. We start with 1.2 million houses at significant or moderate risk and then see how many still fall into these categories 25 years later. The base is £800 million per year divided between £570 million in tangible investment (building new flood defences and maintaining existing ones) and £230 million in intangible investment (mapping, analysis, warning systems and so on). Note that this was a forecast expenditure for 2010/11, since the report was written in 2009.

Flood and Coastal Risk Expenditure jpeg

From this future base period, five different expenditure scenarios were plotted, ranging from conservative to aggressive.

Investment Scenarios jpeg

Under the most conservative investment scenario, number 1, expenditures would flat line at £800 million per annum. Due to inflation, however, this would mean a reduction in spending in real terms, with the result that the U.K.’s flood defences would not keep up with rising climate change-related flood risk. As a result, an additional 330,000 homes would be added into the ‘significant’ and ‘moderate’ risk categories by the year 2035. By contrast, under the more free-spending scenarios, properties ‘at risk’ would fall.

The preferred scenario was Scenario 4, which would give a benefit-to-cost ratio of seven. In the words of the report:

It provides the greatest overall benefit to society, because it generates the greatest net return on investment. However, with this scenario, spending needs to increase from the £570 million asset maintenance and construction budget in 2010-2011 to around £1,040 million by 2035, plus inflation. This equates to an increase in investment in asset construction and maintenance of around £20 million plus inflation each and every year.

Cost Benefit Flood Investment jpeg

So what actually happened? Well, for a start, the Lehman shock. Then, the global economy fell into its worst recession since the Great Depression. Subsequently, David Cameron replaced Gordon Brown in May 2010, and the new Chancellor of the Exchequer, George Osborne, pledged the party to a policy of austerity. And the flood defence agenda was not immune from the cuts.

The House of Commons Library has all the up-to-date annual expenditure numbers in a research briefing called Flood Defence Spending in England. This shows that spending fell in real terms under the coalition government:

Flood Defence Spending jpeg

Accordingly, the expenditure path since the 2009 report has been lower than Scenario 1. In other words, a fall in spending in both real and nominal terms was witnessed, against a recommended incremental increase in real terms of £20 million per year.

Would a different expenditure path have made much difference to the flood-related losses currently being experienced? Probably not. The kind of precipitation seen over 2013/14 would have required vast expenditures to defend against. But the path taken has certainly increased future risk: each yearly incremental shortfall leads to a growing overall flood defence deficit.

The key question, then, is how the current flood defence expenditure path and flood trend will translate into future losses. The political battering the coalition governing is currently taking has already led to something of a U-turn, with a number of pledges being made to increase flood-defence spending. These include:

Response to 2014 Winter Floods jpeg

However, we are yet to see how such spending fits in to the original flood defence investment recommendations made by the Environment Agency.

Ironically, as part of Britain were drowning, the new public-private flood insurance agreement was winding its way through parliament as part of the Water Bill. A summary can be found here and a useful Q&A is available at the Association of British Insurer’s site here.

As with the previous Statement of Principles, the new mechanism relies on a cross subsidy from low risk properties to high risk properties, although it is all organised though a new legal entity: Flood Re. In other words, those properties not at risk of flood pay slightly more in insurance premiums than their actual risk profile would warrant, and these premiums build up to become the capital of Flood Re.

Nonetheless, every household subject to any kind of flood risk must be aware of the many caveats to the scheme. Principally:

  • Flood Re’s risk numbers are based on the Environment Agency’s “Flooding in England: A National Assessment of Flood Risk” which came out in 2009, but is based on data for 2008. Accordingly, all the property at risk numbers are out of date and do not adequately take account of climate change.
  • This is a transitional arrangement slated to run for 20 to 25 years. At the end of this period, market-based pricing is supposed to rule. Given the climate change trend, this means that ‘at risk’ households will be subject to a) major insurance premium increases in future or b) no availability of flood insurance at all. Moreover, the property market will not wait until the end of the scheme, but will mark valuation down well in advance; remember, no insurance means no access to a mortgage.
  • High-end houses in Council Tax Bank H are not included. So all those Thames-side mansions currently under water are on their own.
  • Properties built after 2009 do not fall under the scheme.
  • Businesses and buy-to-let properties are not covered.
  • ‘Uninsurable’ properties will not be included. This is suitably vague so as to provide a get-out clause for the industry. It will be interesting to see if all those Thames-side properties flooded in both the 2012/13 and 2013/14 winters will be classed as ‘uninsurable’.
  • The government has provided a comfort letter, but not subjected itself to a legal requirement to support the scheme should it fall into financial distress. It promises to back stop the scheme if a one-in-200-year event takes place, giving rise to a loss of £2.5 billion or more (note 2007 would fall into this category with a loss of £3 billion). With climate change progressing, however, what is a one-in-200-year event now will not be a-one-in-200 event in future.

To support my assertion that Flood Re is currently begin designed around dodgy ‘at risk’ numbers I refer you to a paper by Swenja Surminski and colleagues from the Grantham Research Institute on Climate Change and the Environment critiquing Flood Re, which can be found here. Two bullet points from the report summary stand out:

Grantham Flood Report jpeg

Note that Dr Surminski was the Association of British Insurer’s Climate Change Advisor from 2007 to 201o.

Against this background, if we get a couple of 2007 type floods in quick succession in the relatively near future (i.e., before Flood Re can build up capital), then the government would likely be expected to act on its comfort letter and recapitalize Flood Re.

At some point, however, the government interest in excluding ‘uninsurable properties’ will fall in line with that of the insurance industry—otherwise the government could get into the untenable position of bailing out property owners whose houses are getting flooded a number of times per decade. Or, alternatively, the government may have to investment massive amounts of money in flood defence to protect these properties. But for many properties, the flood defence expense would dwarf the value of the properties themselves, making abandonment the most sensible policy choice.

The bottom line for all this is that Flood Re will likely provide only a temporary respite for most holders of ‘at risk’ properties. As climate change progresses, more and more ‘at risk’ properties will be shunted out of Flood Re and into the ‘uninsurable’ category. And for many properties, flood defence may just prove financially unfeasible. Don’t say you weren’t warned.

UK Floods: Don’t Say We Weren’t Warned (Part 1)

The public’s reaction to the recent extensive flooding in the U.K. has been one of surprise and frustration. Surprise at the government’s lack of prior planning for flood and frustration at the speed and size of response.

Yet, frankly, there is nothing really to be surprised about. The 2013/14 floods are part of an emerging pattern, which includes the 2003, 2007 and 2012/13 deluges. Accordingly, decisions have not been made in a vacuum; risks have been assessed by the government based on recent flood events and conscious cost-benefit choices made—and there is a paper trail of documents to prove this.

Tackling flood risk is also a classic case of decision-making under uncertainty. Floods have always followed their own probabilistic logic, but climate change has made the range of outcomes far more uncertain. Risk is probability times effect, and climate changes is pushing up flood frequency and severity. This is what both policy-makers and private individuals face.

The realisation that the U.K. has a problem with respect to flood and climate change moved from academia into government over a decade ago. This post looks at the major reports that have shaped government thinking, while the next looks at how this has translated into the speed of flood-defense roll-out and the extent of insurance provision that will be available going forward.

A decade ago, two benchmark reports were published on flood risk: one a private sector report under the auspices of the Association of British Insurers (ABI) called “The Changing Climate for Insurers”, the other a government-sponsored report called “Future Flooding” from the state’s futurology think tank The Foresight Programme. Four years later, “The Pitt Review” was released in response to the catastrophic floods of 2007. And following a recommendation within the Pitt Review, in 2009 The Environment Agency published its long-term strategy document called “Investing for the Future: Flood and Coastal Risk Management”, which looked out to 2035.

All four reports took climate change as a given. All four reports set out the hard choices that have to be made as climate change loads the dice for flood risk.

Let’s start with the ABI’s “The Changing Climate for Insurers” the Executive Summary of which can be found here. The report contains a forward by the then Head of General Insurance of the Association of British Insurers (ABI), John Parker, who was unequivocal about the role of climate change:

Climate change is no longer a marginal issue. We live with its effects every day. And we should prepare ourselves for its full impacts in the years ahead. It is time to bring planning for climate change into the mainstream of business life.

The report suggested that the finger prints of climate change were already showing up in the data:

Extreme Months jpeg

And the author of the report, Dr Andrew Dlugolecki, went on to put the government on notice that things would have to change:

Climate change will increase the frequency and severity of extreme weather events, as well as longer-term trends in weather. The possibility of weather-related catastrophic losses will be much greater, raising issues for insurers of both insurability and capacity. Insurance can only provide a suitable risk transfer mechanism if risks are kept to a manageable level. The insurance industry has a role on behalf of its customers and shareholders, in highlighting the scale of the risks, and examining steps that are needed by all parties (including the Government) to manage risks and ensure that financial protection remains available for the majority of customers.

He also put private-sector customers on alert that things would change for them too. Specifically, owners of ‘at risk’ properties would need to be prepared for ‘at risk’ insurance pricing. Moreover, ‘at risk’ would be less based on what had gone before than on what is likely to come:

The increasing use of risk pricing could enable future claims to be met, provided risks are assessed accurately. Historic claims and climate data will not provide adequate models, so reliable alternatives need to be developed.

Meanwhile, the government had put its futurology team on the case.  The Foresight Programme (which looks 20 to 80 years into the future) was asked to answer the following two questions:

  • How might the risks of flooding and coastal erosion change in the UK over the next 100 years?
  • What are the best options for Government and the private sector for responding to the future challenges?

The Foresight Flooding and Coastal Defence Report came out in April 2004 and was built around four global growth scenarios: World Markets, National Enterprise, Local Stewardship and Global Sustainability. The Executive Summary can be found here.

World Markets is the global capitalism scenario with minimal mitigation of greenhouse gas emissions; National Enterprise is World Markets with protectionism; Local Stewardship is green utopia; and Global Sustainability is statist in nature.

Based on the four scenarios, Foresight produced estimated annual damage (EAD) figures out to 2080 for both fluvial (river in this case) and coastal flooding as well as what they called intra-urban flooding (pluvial, i.e. rainwater run-off, and groundwater).

Annual Average Flood Damage jpeg

For the World Markets scenario—the dash for growth and ignore carbon emissions one—flood damage is estimated to rise 20-fold by 2080. On a GDP basis the position at first glance looks a lot better with damage only rising by 50%.

Annual Average Flood Damage % of GDP jpeg

Against current annual average cost of £1.4 billion from floods, the report estimated that £800 million per annum was spent on flood and coastal defences. Moreover, it also calculated than an annual increase in above-inflation spending of £10 to £30 million would be required just to maintain flood risk at its current level.

The report then looked at a suite of policy measures. Some would be at the macro level in terms of carbon emission mitigation, others at the local level such as increased flood defence and flood proofing. Rather optimistically, the report believed that if all their micro measures where enacted, there would be a dramatic reduction in flood-related damage. Note ‘high’ risk means a greater than one in 75 chance of flooding in any given year.

People at Flood Risk jpeg

Finally, Foresight also realised that the private sector would likely take a much more hard-nosed approach to flood insurance going forward, with the result that the government may have to step in to fill a potential insurance void.

The availability of insurance to cover the costs of flood damage will vary depending on changes in risk and society’s ability to pay. Cover could range from a continuation of the current situation to progressive withdrawal of cover for areas at greatest risk of flooding. Government might have to consider how to respond to pressure to act as insurer of last resort if the insurance market withdrew cover from large parts of the UK, or if there was a major flood which the insurance market could not cover.

The issues raised by the 2004 reports were brought into sharp relief when the British insurance industry took a £3 billion hit in 2007 after major fluvial (river and sea) and pluvial (rain) related flooding. The 2007 floods were unusual in a numbers of ways:

  • The floods took place during the summer
  • Surface water flooding (pluvial) caused as much damage as river and coastal (fluvial)
  • 55,000 houses were flooded and 13 people died
  • Half a million people lost access to mains water and electricity
  • Insurance industry claims reached £3 billion and total damage over £5 billion

In response to the 2007 floods, the government commissioned a report by Sir Michael Pitt, which was published in June 2008 as the Pitt Review. You can find the Executive Summary here.

In conjunction with the review, Sir Michael Pitt requested an update of the Foresight Future Flooding Report that I referred to above. The new Foresight report contained a more pessimistic assessment of climate change impacts on flooding. In particular, it estimated higher sea level rise and rising precipitation leading to more fluvial and pluvial flooding as compared with the 2004 analysis. It did not, however, provide updated figures for estimated annual damage from flood.

Based on the experience of 2007 and the new numbers from Foresight, Pitt recommended that the government commit to a long-term approach to investment in flood risk management, planning up to 25 years ahead. Following Pitt’s recommendation, “Investing for the Future: Flood and Coastal Risk Management” was published by the Environment Agency in 2009, and this has provided the foundation for all cost-benefit decisions regarding flood defence investment since then.

At the same time, Pitt supported the maintenance of the status quo with regard to insurance provision as contained in the Association of British Insurer’s “Statement of Principles”. Under this agreement, the ABI agreed to continue providing insurance to ‘at risk’ households at, in effect a discounted price. Low risk households would foot the bill, being charged slightly more than their risk profile would warrant. The government, meanwhile, would agree to beef up flood defence, but would not get directly involved in either insurance provision or in back-stopping major catastrophe loss.

The ABI had other ideas. The 2007 floods saw 180,000 claims and insurance payouts of £3 billion. The worry for insurance executives was that if 2007 could produce a £3 billion loss event, then what could happen once climate change really accelerated? The possibility existed of an open-ended loss of inestimable proportions. The insurance industry can absorb large losses by just hiking premiums in future years, but a super-large tail risk loss event is a different matter. At a certain point, insurers can be knocked out of business, and thus not survive long enough to hike future premiums.

Faced with the ABI’s implicit threat to walk away from ‘at risk’ households and leave them uninsured, the government realised Pitt’s idea of a continuation of the status quo was a non-starter—and so was born Flood Re. I will look at this agreement in my next post, along with the cost-benefit choices the government is being forced to make with flood defences.