The Climate Change Threat to Productivity

Posted by Cam Hui -
This is the third in a series of the opportunities and threats to productivity. This week, I address the issue of climate change (also see AI Productivity and the Promised Land and Will America get old before it becomes Great Again?).

 

The World Health Organization recently issued a joint report with the World Meteorological Organization that quantified the effects of extreme heat on productivity. It warned that “Extreme heat is fast becoming one of the biggest threats to workers’ health and livelihoods…worker productivity drops by 2 to 3 per cent for every degree above 20°C (68°F). The health consequences are wide-ranging, including heatstroke, dehydration, kidney dysfunction and neurological disorders. Overall, nearly half of the world’s population is now experiencing negative effects from high temperatures.”

 

The Network for Greening the Financial System (NGFS) issued a report that modeled different scenarios of climate change and mitigation policies on global GDP growth. I would make two points about these projections. Investors should approach GDP and productivity forecasts with some humility. Models of temperature change and GDP growth vary wildly, but the general direction of change is down. The higher the temperature, the more growth slows, which equates to a hit to productivity. As well, even though the NGFS report is used by many central banks for scenario planning, the Net Zero in 2050 scenario is totally unrealistic from a practical perspective in light of the voracious power consumption appetite of AI data centres.

 

 

 

The Market Doesn’t Care What You Think

Climate change is a controversial topic, but here’s a secret. The market doesn’t care what you think about climate change. Its function is to price risk, and perceived risks are rising.

 

Consider how insurers price catastrophic risk. Twenty years after the devastation of Hurricane Katrina, Swiss Re Institute warned in a recent report:

Hurricane Katrina does not represent a worst-case tropical cyclone loss scenario. Some of the North Atlantic hurricanes that occurred during early 20th century, if they were to strike today, would cause insured losses well above USD 100 billion in 2024 prices. The Great Miami Hurricane of 1926 could cause the most damage, with simulated insured losses potentially twice those of Katrina. Our North Atlantic tropical cyclone model also shows scenarios with even greater losses if they occurred today, particularly from major hurricanes hitting New York City, Miami, Tampa, or Galveston, Texas.

 

 

While modern catastrophe models are much improved compared to the day of Katrina, secondary perils, such as tornadoes, hailstorms and floods, can be difficult to model. Jencap, a wholesale insurance broker, documented in a report of how Hurricane Helene exposed blind spots in catastrophe models. “The disconnect between models and the reality of Helene’s destruction raises significant concerns, as many models fail when the environment shifts in ways that historical data cannot account for. In the case of Helene, traditional models focused on coastal impacts, underestimating the potential for severe inland flooding.” When Hurricane Helene tore through elevated areas such as Asheville, North Carolina in 2024, the area saw unexpected levels of flooding and landslides.

 

Moreover, hurricanes are becoming more intense, which raises the level of losses. As well, one widely cited study indicates that they’re getting wetter.
 

 

The market is speaking. Already, a number of insurers have withdrawn underwriting coverage in some markets, like Florida. A Bloomberg article, “S&P Warns of Reinsurer Protections as Catastrophe Risks Escalate”, tells the story of how reinsurance companies are becoming more selective about the perils they cover.

Moody’s also notes that in the longer term, severe weather will pose a “key challenge for the reinsurance sector” as natural catastrophes grow more frequent. In the 1970s, there were almost 50 extreme weather events annually; in the past decade, there were closer to 200, Moody’s says.
Reinsurers also face bigger losses in the form of so-called secondary perils, such as severe convective storms and wildfires, Moody’s said. And figuring out how to model such events is proving challenging for buyers and sellers of insurance-linked securities such as catastrophe bonds.

Europe saw numerous heatwaves and wildfires this year. A Bloomberg article highlighted that Europe lost an area the size of Cyprus to fires this year. If the heat waves were to continue, it would severely impact productivity in southern Europe, not to mention the devastation of its tourism industry.
 

 

There is a silver lining for Americans. The NGFS model shows that, under all scenarios, North America is expected to Experience the least direct impact of severe weather events. This puts the catastrophe insurance modeling challenges previously discussed into context. It will be far worse in other regions of the world.

 

The Federal Reserve was recently the subject of some criticism of exhibiting “woke” concerns over climate change. But if a central bank is charged with ensuring financial stability, the magnitude of catastrophic weather-related hazards could strain the financial system. It is therefore prudent for the central bank to ensure the banks under their supervision have adequate financial buffers against these losses.
 

 

 

Gradually, Then Suddenly

How should investors respond to the risks posed by climate change? At the end of the day, the investment approach to climate change should be an exercise in risk mitigation and control.

 

In a 2015 video made for the Boston Security Analysts Society’s Sustainable Investing Seminar, Bob Litterman discussed the issues surrounding pricing climate change risk. Here are the main takeaways from the interview:

  • The worst-case scenario in climate change is highly uncertain, Litterman argued that you have to build in a risk premium to compensate for that uncertainty.
  • Create a sufficient incentive structure, or tax policy, to price the cost of carbon emissions to disincentivize carbon-based energy and incentivize renewable energy.
  • Uncertainty itself is risk, and therefore you should assign a risk premium to the uncertainty. The price of climate risk should Be the sum of the risk itself, and the degree of social risk aversion.

Litterman went on to reveal that the World Wildlife Fund put on a “stranded asset return swap” by shorting a basket of stranded asset companies, which are mainly coal and high-cost tar sand extraction stocks, against their portfolio. A Financial Analyst Journal article went further and detailed a quantitative technique for a constructing a “Factor-Mimicking Portfolios for Climate Risk”. The authors of the paper proposed an approach to minimize climate exposure while controlling the standard portfolio risks within a Fama-French risk framework.

However, there is a chance that the effects of climate change could suddenly become non-linear. An article in the Economist outlined how the Earth’s climate could be nearing an irreversible tipping point. It described an “Amazon dieback” scenario where the Amazon ecosystem could enter a negative feedback loop that turns the basin into a dry savannah.

[Other examples] include the breakdown of the vast Greenland ice sheet, which would raise global sea levels by more than seven metres, and the collapse of the Atlantic Meridional Overturning Circulation (AMOC), the powerful system of heat-distributing ocean currents that keeps northern Europe reasonably temperate. Should AMOC collapse, temperatures and rainfall levels could fall dramatically across Europe, greatly damaging the continent’s ability to grow crops.

The only problem is that no one can accurately model how far we are from any of these tipping points. The error terms of the forecasts are large and unknown.

 

In conclusion, the market doesn’t care what you think about climate change. Its role is to price risk, and market risks are rising. A hotter climate degrades labour productivity. While the accuracy of loss models is unclear, the direction of the trend calls for headwinds to growth and productivity. For investors seeking to mitigate climate risk, a Financial Analyst Journal article proposed an approach to minimize climate exposure while controlling for standard portfolio risks within a Fama-French risk framework.
 

The key question for investors is the likely trajectory of U.S. total factor productivity in the next decade. Our review shows that while AI adoption and deployment is a tailwind, stagnant U.S. population growth and the effects of climate change are headwinds. I conclude that, at best, factor productivity will be no better than trend. At worst, it will flatten and show little or no growth.