The Actuaries Climate Index in Scholarly Publications

The Actuaries Climate Index (ACI) was launched in 2016 by its joint sponsors—the Academy, Casualty Actuarial Society, the Canadian Institute of Actuaries, and the Society of Actuaries—to offer an index of weather extremes as an objective indicator of climate change experienced in the United States and Canada. The ACI relies on observed information on high and low temperatures, high and low precipitation, wind, and sea level to document changes based on a reference period, 1961–1990. In January 2020 the Academy published preliminary findings of the Actuaries Climate Risk Index (ACRI), linking changes in climate based on the ACI with property losses in the United States.

In the past two years, four academic publications have demonstrated the usefulness of the ACI as an indicator of climate change. One reports the impact of climate change on macroeconomic activity, and a second reports the impact of climate change on agricultural production, directly and indirectly. Another foreshadows the value of the revision of the ACI, currently underway. Finally, the most recent of the publications, in an analysis of the impact of air pollution on excess mortality, uses the ACI to control for the effects of climate change.

In an August 2021 working paper from the Federal Reserve Bank of Richmond, the authors examined the impact of the ACI on industrial production, unemployment, and inflation. Relying primarily on analysis of a smooth transition and a vector autoregressive model, they found clear statistical evidence of the impact of climate change on macroeconomic variables in the United States. They also found clear evidence of greater impact in more recent years than in earlier years, and they reported on efforts related to individual components of the index to macroeconomic effects. While they do not find any single component significantly associated with all of the macroeconomic outcomes, they did find certain components associated with certain outcomes.

In a 2020 master’s thesis submitted to the University of Waterloo (Canada) in fulfillment of the thesis requirement for the degree of Master of Mathematics in Actuarial Science, author Ruihong Jiang found that the ACI was strongly predictive of the performance of corporate profits and stock returns for companies in the agricultural sector. Assuming that climate change is impacting the production of agricultural crops, Ruihong wrote that increases in the ACI were negatively associated with agricultural corporate profits throughout the entire time period, although the relationship between the ACI and stock returns diminished significantly in 2017. The author concluded that increased market attention to climate change might be leading to this result.

In the first article relying on the ACI to appear in a peer-reviewed journal, published this year, the authors made an assessment of the effectiveness of the ACI in estimating extreme weather on certain statewide crop yields. Using both linear regression and probit analysis of data for eight Midwestern states, they found the ACI predictive both of crop yields and of the severity of shortfalls in crop yields. The impact was even more significant when the index is decomposed into five elements (sea level not being a factor in Midwestern states); when the weights of the elements are estimated directly rather than assigned a priori as they are in the index, the relationships are stronger.

In an attempt to see whether increased granularity improves results, the authors used climate data from another source to create an analysis on a 4-km-by-4-km grid. They used the ACI methodology to develop a pseudo-ACI and found that climate change and crop yields were still strongly related, but more weakly than observed at the state level, and hypothesized that the weakness of the more granular crop yield data may be the cause of the weakening. This more granular analysis foreshadows work underway by the Academy and the ACI’s actuarial association partners to revise both the ACI and the ACRI. By switching climate data sources, we aim to develop the ACI on a 0.25-degree-by-0.25 degree grid. Associating this more granular ACI (and its components) with more granular losses (also with new, improved data sources), we look forward to more robust and more granular results for the risk produced by climate change.

Finally, in a second peer-reviewed article, also published this year, the authors examine the impact of air pollution on excess mortality in California. Because air pollution can result from climate change and the authors had reason to believe that climate change might affect excess mortality, they controlled for climate change by using two of the six ACI components—low temperature and high temperature—for the Southwest Pacific region. They found those ACI components statistically significant in their impact on excess mortality.

With interest in climate change greatly expanded and given the success of these early attempts to use the ACI as an indicator of climate change, it seems likely that more such uses will be forthcoming. And when the revisions of the ACI and ACRI are completed, offering more robust and more granular data, the revised indexes may prove even more useful to actuaries, non-actuary climate science stakeholders, public policymakers, and the general public.