The ABI had performed a sensitivity test rather than a projection, and this difference was confused by none other than the Stern Review Report. At the time I wrote of Stern's reliance on ABI:

The Stern Review’s methodological error is based on treatment of a sensitivity analysis focused only on the effects of climate change in the context of total GDP as a projection. Thus, rather than telling the reader what losses might be expected in the future, the Stern Review’s results instead indicate the effect that future climate change would have on today’s world GDP. This is misleading because both GDP and societal vulnerability are increasing at a rate that will increase catastrophe losses much faster than the independent effects of climate change.ABI has released a new report and while it takes a small step in the right direction, it continues to confuse the issue. The report notes that it is conducting a sensitivity analysis, but then does something odd with GDP (p. 39):

Pielke, Jr., R. A., 2007. Mistreatment of the economic impacts of extreme events in the Stern Review Report on the Economics of Climate Change, Global Environmental Change, Vol. 17, pp. 302-310.

The loss estimates presented in Section 3 isolate the effects of climate change by holding all other parameters constant. In examining the potential impact on insurers’ operations in this section, projected 10 year GDP growth is considered alongside an assumption of no growth. The GDP projections are meant to serve as a proxy for combined increases in the number of insured properties resulting from population growth and increases in the total sums insured resulting from increased wealth.The report's examination of climate change effects are for the 2040s and 2070s. So it is comparing the effects of ~35 and ~65 years of climate change with 10 years worth of change in GDP. If one wants to look at 35 or 65 years worth of climate change, it would make obvious sense to look at 35 or 65 years worth of GDP change. The ABI is thus looking at apples and oranges.

The ABI report states that GDP growth over 10 years leads to a cumulative increase in expected losses of 28% for 2.5% and growth over 10 years and 79% for 6.0% annual growth over ten years. It then uses these 10-year growth figures to compare with 35 and 65 years of climate change:

As an example, a conservative long-run estimate of annual GDP growth for the UK and China (based on annual real growth of 2.5% and 6% respectively) could be adopted as a proxy for increases in the number of insured properties resulting from population growth and increases in the total sums insured resulting from increased wealth.What would happen if they did an actual apples-to-apples comparison of losses based on 35 and 65 years of GDP growth?

Assuming a 4°C increase in temperature and ten years of GDP growth at 2.5%, insured 100-year Great Britain flood losses could rise by 38% compared with a possible rise of 30% if GDP growth were not taken into account. Assuming a 4°C increase in temperature and ten years of GDP growth at 6%, insured 100-year China typhoon losses could rise by 16% compared with a possible rise of 9% if GDP growth were not taken into account.

The independent effect of societal change on increasing losses from their 2010 values would be as follows:

2045 -- 240% (@ 2.5% annual growth) -- 770% (@ 6.0% annual growth)

2075 -- 500% (@ 2.5% annual growth) -- 4,415% (@6.0% annual growth)

In other words, for 2075 the ratio of the independent effect of societal change on losses to the independent effect of climate change, under the assumptions of the ABI exercise, would be about 17 times larger in the case of the UK and almost 500 times larger in the case of China. This means that the societal influences on future losses overwhelmingly dominates the projected effect on losses due to climate change.

The ABI report is extremely misleading because it compares 10 years worth of societal change to 35 and 65 years worth of climate change.