The first Actuaries Climate Index (ACI) was developed in North America, sponsored jointly by the American Academy of Actuaries, the Casualty Actuarial Society, the Canadian Institute of Actuaries, and the Society of Actuaries. This went live in November 2016. In 2018, the Actuaries Institute Australia released the first results of its climate index (AACI).
The Board of the AAE established a Working Group (WG) in February 2019 to investigate the feasibility of producing a European Actuaries Climate Index (EurACI). The WG, which includes members from 15 associations, has undertaken a detailed analysis and has had the benefit of discussions with the ACI working group in North America, which has helped to identify the key issues to be addressed in developing a EurACI.
The initial report of the WG to the Board in October 2019 concluded that the development of a EurACI, similar to the ACI and AACI, was feasible and indeed desirable, but noted that further work was needed to firm up on data availability and the likely costs of development and maintenance of the index.
The AAE Board is considering how best to meet these costs, including the possibility of seeking funding from European bodies. This article will focus on the issues which need to be considered in deciding on the datasets to use, and how to ensure the quality, reliability and availability (including historic information) of the data which is input to the model.
STRUCTURE OF THE ACI
The ACI is based on analysis of seasonal data for six index components collected since 1961. The index measures changes in extremes of high and low temperatures, high winds, heavy precipitation, and drought, as well as changes in sea level, expressed in units of standard deviations from the mean for the 30-year reference period of 1961 to 1990, for the United States and Canada combined and by region. Combining six components over a five-year measurement period, the index’s moving average smooths out monthly and seasonal fluctuations for a meaningful measurement of long-term climate trends.
Further detail on the construction of the ACI and the data sources used are available on the website.
The most recent update was issued in November 2019. As can be seen from the chart below, the ACI’s five-year moving average continues to rise and for the fifth consecutive quarter has set a new high.
The ACI uses gridded datasets for all of the components except mean sea level. Each of these grids covers an area of 2.5 degrees latitude by 2.5 degrees longitude, and the data values are in effect averages over that area. This provides more manageable amounts of data, which have been collated and checked by the data provider. However, the use of gridded datasets does lose some valuable information where the area covered by a grid is not homogenous with regard to climate, and it relies on the dataset provider to check the reasonableness of the data and to deal with any gaps due to missing data. In our discussions with the ACI working group, they have indicated that for the next version of ACI they are minded to use individual station data to address these concerns. The AACI is based on individual station data.
The European Climate Assessment & Dataset (ECA&D), which is based in Utrecht, collates data provided by European meteorological services, and produces gridded data sets for Europe. Members of the WG had a very useful meeting with ECA&D in September 2019, and based on these discussions and further information subsequently provided, the WG is of the view that these datasets (which are freely available) could be used to develop a EurACI.
A final decision on whether to use gridded datasets or individual weather station data has not yet been reached. The key question is whether the benefit of having more granular data is greater than the additional resources (time and cost) which would be required to validate and process them.
In either case, the question of missing data will need to be addressed, as the availability of readings in some parts of Europe is quite limited. The gaps may be filled by the use of reanalysis but this does introduce a level of approximation which might not be considered acceptable. It may be necessary, at least initially, to limit the scope of the Index to those parts of Europe where there is sufficient data to be credible.
Another possibility is to use satellite data, and the WG has had some interesting discussions with a provider of satellite data, Space4Climate, in this regard.
USAGE OF THE INDEX
The Index will provide information on trends in the frequency of extreme events that could be attributed to climate change. Although no final decisions have been taken on how this will be made available, it is likely that quarterly updates will be published on the AAE website, similar to those published by ACI and AACI. No decision has yet been taken on what, if any, sub-indices covering different regions or countries might be produced; this will depend crucially on the availability of sufficient data to make the sub-indices credible, and on the additional value which sub-indices might provide in return for the additional work involved.
Climate indices provide useful information for actuaries, insurers, regulators and policy makers in relation to the frequency of the occurrence of extreme climate events. They do not provide information on the losses which arise due to these events. The ACI working group are working on the development of an Actuaries Climate Risk Index, which would incorporate information on the losses arising from past events which could be of assistance in setting reserves and capital requirements and indeed pricing for such risks, and the American Academy of Actuaries has recently published the preliminary findings of this research.
The WG will finalise its recommendation to the AAE Board shortly, and if the Board decides to progress the development of the EurACI, it will commence the project with a view to having an Index available as soon as possible. This may require some pragmatic decisions to be taken about the level of data quality and scope of the index, rather than delaying significantly the production of an Index in the pursuit of perfection. An ambitious target would be to have developed the Index by the end of 2020, with the first results being published in early 2021.
The AAE will need to establish governance procedures in relation to the publication of the Index and to keep it under constant review; in time, this might lead to more detailed outputs and to the development of a risk index.
Ultimately, the target should be a global actuaries climate index, which is likely to rely significantly on satellite data to cover the oceans as well as land areas where reliable data is sparse