REACT stands for Reinsurance Event-Attributed Carbon Tax, a system designed to put a fair price on the financial loss and damage arising from carbon emissions.
REACT measures the expected losses that will occur for each tonne of carbon dioxide emitted. Further, it determines the allocation of the costs of those losses to each energy supplier based upon the proportion of carbon emitted and the carbon intensity of the fuel type used. The total revenue paid by all suppliers equals the total loss and damage caused by their emissions.
In summary, the methodology is as follows:
PAL’s PALgamma algorithm predicts the number, type and location of disasters worldwide.
From this is determined the insured and uninsured losses from disasters and catastrophes; PAL’s method is more accurate than conventional moving-average analyses and it works completely independently from previous history.
These losses are scaled to determine the global, overall losses from weather-related incidents (roughly $5 trillion per year).
Finally a method has been developed to determine the extent to which events can be attributed to anthropogenic (human) emissions of CO2. This is called REACT – reinsurance event attributed carbon tax – and it combines the various categories of losses with appropriate ‘global attribution factors’. This is complementary to the much more detailed analyses being carried out at Oxford University and elsewhere to determine the extent to which human-induced climate change has influenced individual disasters and catastrophes, such as the 2015 Storm Desmond in the UK. REACT uses a global temperature anomaly (global warming) to determine the global attribution factor.
Predict Ability Ltd (PAL) has encoded these methodologies into its real-time, online product suite PALgamma and PALcarbon. Spot and one-year carbon prices are published and they can be used in carbon auditing to determine short, medium or long-term costs of carbon emissions.
For preliminary audits, such as that presented in this report, the PALgamma and REACT methodologies can be approximated by simple exponential formulae, as explained in Chapter 8 of Predicting the Price of Carbon. In every case the price of carbon is y = L . x / C, where L = weather-related losses (determined from PALgamma), x = global attribution factor (a function of the global temperature anomaly Ta), and C = carbon emissions. The basic ‘spot price’ formula uses the instantaneous value of PALgamma and PAL’s real-time Ta to provide a dynamic insight into the trends in carbon price and losses – as Ta falls, so do the number of disasters to some extent.
Finally, PAL has determined the most appropriate methodology for charging for CO2. This is the PALcarbon 1 year carbon price (PAL-1Y). The logic is this: until recently the damage caused by CO2 was not widely known and, even where it was known, it was not paid for. Thus, the damage caused by historic CO2 emissions is a sunk cost. But now, both scientifically, politically and legally, carbon emissions must be accounted for. Yet, 1 tonne of CO2 emitted today continues to cause damage for many years ahead. So what is to be done? If all the emissions in a rolling one-year period are paid for by a tax or other actions then a significant part of the impact of that tonne of CO2 in future will have been paid for, and so on for each subsequent year. The tax to be paid depends on the amount of damage, the amount of attribution and the global emissions. In practice, over the 60-year life cycle of the Hinkley project, the PALcarbon PAL-1Y carbon price will cost about the same as the US Social Cost of Carbon (using the 3% discount rate proposed by the US Interagency Working Group on Social Cost of Carbon). This is because PAL predicts that future carbon costs will strongly, indeed exponentially, escalate. For this reason alone energy, transport and heating systems need to massively decarbonise. This is where politics and finance meet: PAL’s carbon price sends much the same signal as the Paris Agreement. In the long term, unabated gas-fired power stations will not be an option.
The full REACT methodology is laid out in Chapter 7 of Predicting the Price of Carbon.