Using financial logic to price carbon emissions

May 23, 2024|Written by and
A coloured graph displayed on a black computer screen, with three curving graph lines and a line of range bars.
© Maxim Hopman

By now, it’s no secret that climate change is altering the economic and financial landscape. An increasing number of jurisdictions around the world are pricing carbon, and the US is embarking on a massive green industrial policy experiment that could hand out upwards of a trillion dollars’ worth of subsidies over the next decade. Much of these efforts are framed around meeting the Paris Agreement global average warming target of “well below” 2°C, with further efforts being made to limit global warming to 1.5°C above pre-industrial temperatures.

Despite these overall warming targets enshrined in the Paris Agreement, there is plenty of disagreement to be had about the details of climate policy: its overall ambition; the rate at which it should change or grow; whether subsidies or taxes are more effective; and the myriad of trade-offs that exist in achieving an economy fully rehabilitated from its fossil fuel addiction, where presently each tonne of coal and each tonne of oil burned cause more in external damages than they add value to GDP.

Perhaps the most well-known economic critique to the Paris Agreement warming targets comes via the most prominently studied climate-economy model, the dynamic integrated climate-economy model (Dice), whose developer, William Nordhaus, won the Nobel prize in economics in 2018. Norhaus’s preferred calibration of Dice claims that the socially optimal level of warming – that is, the level of warming that maximises averted climate damages minus the cost of avoiding them – is just north of 2.5°C by 2100, and will continue to rise thereafter.

Climate scientists, who have been sounding the alarm on climate change since at least the 1980s, stand in stark contrast to the economic debate for one specific reason: risk. With a few decades’ worth of scientific studies behind them, the climate science community confidently claims that our continued emission of greenhouse gases is putting us on a fast-track towards disaster. For example, a recent study found that any amount of warming beyond 2°C could trigger a series of cascading climate impacts.

So what’s the hang up? Why is there such a glaring disagreement between economists, some of whom claim the Paris targets go too far, and climate scientists, who resoundingly support the targets?

How do we quantify climate damages?

The key sticking point in this debate can be embodied by the need to “put a number” on climate change damages.

Approaches to computing this number for climate damages vary widely, depending on the context and specific trade-off being considered. To keep things simple, we’ll use Nordhaus’s preferred calibration of Dice as an example: this weighs the costs of cutting carbon emissions and the benefits of avoiding climate damages, and in the end argues for a relatively low “optimal carbon price” in the near term that rises over time.

The primary issue with Dice – as outlined in a number of studies – is that it has an inadequate representation of climate risk and uncertainty. This includes, for example, key parameters such as the equilibrium climate sensitivity (defined as the global average amount of warming in equilibrium after a doubling of CO2 relative to pre-industrial levels), which have been known to be uncertain since the 1970s.

Thus, it is fair to say that Dice, despite its acronym alluding to the climate casino, misses the point that has been raised by the climate science community for decades: that what we know about climate change is bad, and what we don’t know is even worse.

A new framework for computing climate damages

Our recent paper, co-authored with climate dynamicist Cristi Proistosescu, puts forward a new economic framework for computing the cost of climate damages, utilising a framework naturally capable of accounting for risk and uncertainty: asset pricing. The logic of our approach is deceptively simple: every tonne of CO2 emitted into the atmosphere is a risky asset with negative payoff. The negative payoff of each tonne of CO2 equates to the cost of CO2 emissions.

The asset pricing framework we adopt allows policymakers to consider an important feedback: the more fossil fuels they emit, the riskier the world becomes. This feedback drives up our optimal carbon price, as not only are disastrous levels of warming more likely in a warmer world but other risks become more likely to be triggered as well, such as hitting a so-called “climate tipping point”. This dynamic assessment of risk further incentivises tightening the spigot of fossil fuel emissions in the near term.

Analogising our findings to portfolio management is natural. Finance 101 tells us that one must balance holdings of risky stocks and safe bonds to maximise payoffs; the riskier the asset, the more bonds one must buy to hedge against potential losses or the more one divests from the risky asset altogether. Our model tells us that greenhouse gas emissions are an incredibly risky asset, and that it is prudent to invest heavily in abating CO2 to insulate oneself from the potentially catastrophic damages CO2 emissions can cause.

To directly compare our results to Dice, we find that our carbon price is more than four times Dice’s (less than US$200 per tonne of CO2, as opposed to around $50 per ton). Furthermore, considering both climate damages and assumptions about mitigation costs, we find an optimal level of global average warming in the neighborhood of 1.5–1.8°C, giving explicit economic support to the Paris Agreement warming targets.

Placing risk front-and-centre in the climate policy debate

Going forward, our work has a number of implications for financial regulators and policymakers.

First, it shows that risk should be a dominant player in the climate policy discussion, and its treatment can have dire implications for economic and financial planning. Indeed, climate risk is financial risk. Regular stress tests and sensitivity analysis to worst-case climate scenarios are essential in building out a robust policy or investment strategy, especially as emissions are projected to remain stubbornly high if current policy pledges aren’t met. (Incidentally, the possibility of failing to meet current climate pledges provides yet another source of climate-related risk often not accounted for in climate-economy models.)

Based on our work, we would recommend stress-testing investment strategies with carbon prices up to 2-4 times the baseline levels, if not (much) higher, to account for worst-case scenarios.

Second, as many others have advocated for, climate-related risk disclosures are vital to avoid a so-called “carbon bubble”, whereby the value of climate-exposed companies inflates owing to a lack of stakeholder knowledge of the risks involved with their investment. How a company would respond to potential climate-related risks (such as a collapse in insurance markets or coastal housing prices) is important financial information for consumers to know, so their finances can be invested accordingly.

Finally, we would advocate that climate change is, fundamentally, an issue of risk management. And while it is true that trade-offs abound, we must ensure that our financial system is robust to the worst climate change has to offer, just as we would with any other risky asset class.

It is far past time to view carbon dioxide emissions themselves – not just the capital that generates them – as an investment, one that does societal harm and needs to be hedged against. Our work, ultimately, puts this reality front and centre, and should motivate regulators to heed the warning that climate scientists have been arguing on behalf of for decades.

This page was last updated July 4, 2024

Share this article