Concerns around the growing impact of physical risks associated with global warming call for a more in-depth analysis of their potential economic consequences – by Irene Lauro and Samar Khaana

Global average surface temperature rose to 1.45°C above pre-industrial levels in 2023, making the year the warmest on record. As global warming intensifies, extreme weather events have become more frequent and their economic damage more significant. With climate change becoming a more pressing issue, it is extremely important for investors to understand its impacts on the economy when making asset allocation decisions.

This is not new. But what is, is that climate scientists are forecasting things will get a lot worse a lot sooner than many mainstream economic models predict. Economies and financial markets face greater downside risks. There is uncertainty in all of these projections, and this is only one possible scenario. But investors should understand how their portfolios could fare if the forecasts of climate scientists materialise, and what steps they can take today to mitigate against the risks.

What do popular economic models predict?

While there is a strong consensus that a warmer world means a poorer global economy, economic estimates tend to lie in a very broad range. This is unsurprising given the high degree of uncertainty surrounding future climate scenarios and the deep link of various sectors and ecosystems. For example, the total loss in global GDP by 2100 with a 4°C warming is estimated to be between 4% and 23% (chart 1). Here we show the estimated impact of higher temperatures in recent economic studies.

It is important to highlight that while a 23% drop in GDP might seem a big hit to the economy, when looking at the annual damage, the economic cost is estimated to be less than 0.4% per year, even in the most severe case estimated by economists.

This is in contrast with expectations of climate scientists, who are forecasting mass extinction (i.e. the end of much life on earth) if temperatures rise more than 5°C from preindustrial levels. While 5°C would be towards the upper end of the figures in the table above, the impact on GDP would obviously be far greater than any of these figures suggest.

A group of international scientists from the Potsdam Institute for Climate Impact Research and the Stockholm Resilience Centre predict that even at 1.5°C economic models are overlooking the true degree of risk. There is a significant likelihood at these temperatures of passing multiple climate tipping points, particularly around major ice sheets. When large parts of the climate system cross a warming threshold and start to change by themselves, this is called a tipping point. It can lead to serious consequences, such as a big increase in sea levels from melting ice sheets and carbon emissions from frozen ground that thaws. Another international team of scientists predicts that even if the carbon emission reductions called for in the Paris Agreement are met there could be a severe climate impact. For instance, within the Paris scenario there is a risk of Earth entering what they call “Hothouse Earth” conditions, where sea levels would increase by 10-60 metres.

Revising the damage from climate change

There are growing concerns that current economics models, including those we use in our own 30-year return forecasts, may be underestimating the risks of climate change.

For our 30-year returns forecasts, a “damage function” incorporates the economic impact of climate change via the long-term productivity forecasts. The one we use is derived from the methodology of Burke, Hsiang, and Miguel (BHM; 2015). This function shows a non-linear relationship between productivity growth and average temperature.

Hotter countries are hurt by higher temperatures but, as explained in our 30-year returns paper, “cold country” productivity is assumed to increase as annual temperatures increase (chart 2). This relationship suggests that climate change is set to create winners and losers, as economies with a low temperature are forecast to see a boost on the back of a warmer world, while warmer countries suffer.

However, there are growing arguments from climate scientists that the physical risks of climate change could be much higher and be distributed differently around the planet. This is primarily due to a combination of two factors. Firstly, confidence is increasing among climate scientists that, as temperatures rise more than 2°C above the pre-industrial level, the global economy is likely to be negatively impacted by changes in the physical environment.

As highlighted by a recent academic study, scientists are now more confident that rising temperatures will result in stronger tropical cyclones, extreme heat impacts, more frequent and intense floods and droughts. In addition, they increasingly believe global warming will result in the destabilisation of ice sheets and glaciers and consequent sea level rises.

Secondly, most of the damage functions used by policy makers to quantify the risk the economy faces as a result of climate change only look at global average surface temperatures. They exclude other important measures, such as temperature volatility. This matters as temperature volatility directly impacts the frequency and severity of extreme weather events. As will be discussed in the next section, this second factor could have a profound and negative impact on the outlook for colder countries in particular. Developed markets would be the worst affected.

Our joint research with Oxford Economics introduces a revised damage function that incorporates these temperature anomalies (i.e. deviations from long-term temperature averages), in addition to temperature averages. This allows us to consider factors like temperature volatility and extreme weather events – giving us a clearer picture of the shape of temperature distribution beyond the mean.

By accounting for temperature anomalies, volatility and skewness in temperature distributions, the damage function produces a more severe relationship with economic activity, with important implications for the growth-inflation outlook of the economy.

Climate downside scenario

Our Climate Downside scenario is developed on the revised damage function. This approach aims to create a scenario that, while severe, remains within the realm of possibility. Our ultimate goal is to explore what a climate downside scenario might look like. Investigating alternative damage functions is one of the ways to examine potential future trajectories associated with climate downside risks.

Chart 3 shows how physical and transition risks play out in our scenario analysis. Physical risks are as the name implies. Transition risks are the economic consequences of actions taken to move to a lower temperature trajectory.

In our Climate Downside scenario, transition risks are low as it is assumed that governments will fail to take the actions necessary to meet their climate objectives i.e. there will not be much of a transition. This failure results in an energy landscape that continues to be heavily dependent on fossil fuels with oil, gas and coal production increasing until 2050. In reality, the severe consequences of such a scenario mean that the likelihood of it generating no response is low. However, it can be useful to understand the scale of the impact if it did come to pass.

In contrast, in the Delayed Transition scenario which we used as our baseline for our 30-year forecasts, we manage to somewhat meet our net-zero objectives. But because the government delays its action, the transition is chaotic, harmful, and expensive, resulting in a much higher impact due to increased carbon pricing, even though the physical damages are lower.

The low level of climate ambition in the Climate Downside scenario is reflected in higher global warming. In this scenario, temperatures are expected to rise to around 2.2°C by 2050, higher than 1.9°C in Current Policies and 1.6°C in Delayed Transition. As a result, physical costs are more severe.

Impact on GDP

The higher level of warming and the adoption of a damage function with temperature anomalies lead to a more significant impact on global GDP. There is a 20% decline in global GDP under the Climate Downside scenario relative to Current Policies scenario by 2050 (chart 5). The decline in GDP expected in the Delayed Transition relative to Current Policies is around 4% by 2050, as the hit to the economy from a sudden and steep rise in carbon prices is mitigated by less severe physical risks.

It is also important to note that physical risks rise sharply over our forecast horizon in our new scenario as higher temperatures lead to higher temperature volatility. Consequently, there are more frequent extreme events, with the decline in GDP becoming more significant after 2040.

In the Climate Downside scenario, the hit to the global economy is now larger as all economies are expected to see a drop in GDP, as shown in chart 6. With the application of the revised damage function, a number of cold countries – which coincidently happen to be most of the developed economies – face more severe economic damages. This is different to standard approaches, such as the BMH model, which imply that colder countries will benefit from initial warming.

Canada, the US and the eurozone are forecast to see the largest hit to GDP in Climate Downside compared with the Delayed Transition. This is due to the fact that they are likely to warm up faster than others, being closer to the Artic, the fastest warming region of the planet. A quicker rise in temperature for regions in the northern hemisphere means higher temperature anomalies and volatility and therefore more severe extreme weather events hitting these economies.

The eurozone, Japan and Brazil are forecast to have outright negative growth over the next 30 years in this scenario. The US is barely positive. The downside risks should not be underestimated.

What does it mean for inflation?

Transition and physical risks have a significant effect on economic output, and they will also influence inflation. The impact on inflation is initially much higher in Delayed Transition than in Climate Downside. This is because in the transition scenario, we see a sharp rise in carbon prices starting from 2030, that leads to escalating energy prices. As economies tend to move away from fossil fuels and start to rely more on cleaner energy sources, the impact on inflation declines in the transition scenario.

In comparison, inflation tends to increase in our Climate Downside scenario over the next decades due to the escalation of temperature rises and the greater volatility in temperature changes. More frequent extreme events lead to resource scarcity, notably around the availability of water, arable land, and certain commodities, which results in cost-push inflation. As temperatures rise, input prices follow suit, leading to a supply-side shocks that intensify over our forecast horizon.

What it means for investors – the impact on 30-year returns

Changes in economic activity impacts our long-term return forecasts. Higher inflationary pressures in Climate Downside means higher nominal returns. If we strip out the inflationary impact, real returns provide a clearer picture to the risks to asset class returns, as shown in chart 8.

Most notably, given the hit to economic activity expected for developed markets, annualised real equity returns for the US and other developed markets are nearly half of what they were under the Delayed Transition scenario. Emerging market equity returns are also lower, but it is developed markets that suffer most. Real cash returns in the G4 economies are negative over our 30-year forecast horizon.

Real risk-adjusted returns are lower across the board in this scenario. In it, our analysis suggests investors will find it increasingly challenging to generate returns in excess of inflation at the levels they may be hoping for. This highlights the importance of incorporating physical risks in strategic asset allocation decisions.