Consultants should ensure pension fund trustees are communicated the “huge uncertainty” in estimates of different climate outcomes’ impact on portfolio returns, according to an EDHEC academic.
Earlier this year the climate scenario analysis conducted for pension funds came in for sharp criticism in a spate of reports. The latest of those came from one of the UK’s largest funded pension schemes, which worked with University of Exeter specialists to produce ”new scenario narratives for action on climate change”.
In a new position paper, Riccardo Rebonato, who is scientific director of EDHEC-Risk Climate Impact Institute, said EDHEC agreed that pension fund trustees have been poorly served by their consultants and that the models that have been used underestimate climate risk.
However, he said that the models (DICE-like Integrated Assessment Models) should not be jettisoned as they can be modified to handle scenario analysis (for which they were not designed) and that it is incorrect to posit that there is an economist consensus on the severity of climate damages.
At the same time, Rebonato, who is also professor of finance at EDHEC Business School, said that the degree of uncertainty about climate outcomes was “the single most important piece of information that pension trustees should be given” and yet it was exactly what was “more conspicuous for its absence in the consultants’ reports”.
“Pension fund trustee would not be necessarily better served by mechanically substituting a set of aggressive damage estimates for a set of tame ones”
“Given the huge degree of uncertainty in climate damages, pension fund trustee would not be necessarily better served by mechanically substituting a set of aggressive damage estimates for a set of tame ones, especially because excessive prudence can have investment consequences as severe as an underestimation of climate risk,” he said.
Instead, approximate probabilities need to be attached to climate scenarios, according to Rebonato.
He said that if a probabilistic description was deemed too difficult, consultants could at least provide estimates of expected returns (and other statistics) obtained from a variety of models, damage functions, economic growth assumptions, and other sources.
He acknowledged that consultants may be responding to pension fund trustees’ desire for simplified messages and said that although this was understandable, trustees needed to address this.
“The temptation is understandable, but the unprecedented nature of climate change makes single-number answers not just useless, but dangerous,” said Rebonato.
No comments yet