APG raises questions on quality of climate risk data
Leading Dutch pension fund APG encounters significant variation in third-party data on climate risk in the property sector, calling into question the reliability of providers’ methodologies.
Derk Welling, senior responsible investment and governance manager for global real estate investments at APG, who is based in Amsterdam in the Netherlands, told AsianInvestor that methodologies designed to evaluate buildings’ climate risks were frequently unclear, especially when hidden within “black boxes”.
“We never want to rely on one data source, as we have observed conflicting risk assessments for the same asset. We cannot make informed investment decisions based on black box models. We want to understand the climate models that are used by these data providers. For example, one data provider [may] see no sea level rise risk, whilst the other comes with an extremely high sea level rise risk,” he said.
The fund encountered the variation when establishing specific geolocation data on all private and public investments it owns, which form part of a global asset-level database it has assembled that covers more than 100,000 buildings.
The database employs inputs from scores of data providers to determine specific discount rates on an individual asset level.
APG recently hired a climate data expert, in part to make sense of providers’ models, Welling said.
“The asset-level data is enriched with data from multiple data providers, including exposure to climate hazards,” he said.
ASIAN HAZARDS
Disagreement in data providers’ evaluations of climate risk is of particular concern in Asia, where the risks associated with different buildings within the same city often vary considerably.
“Most of the major cities in the Asia Pacific region have a large spread in individual assets’ value at risk, due to physical climate risks,” said Niel Harmse, vice president of real estate solutions research at MSCI, based in Cape Town, South Africa.
He noted that risks varied significantly, even for properties that were geographically near to each other. “This is especially true for coastal and fluvial flooding, where a property’s micro-location can have a major influence on its value at risk,” he said.
Harmse pointed to the example of Hong Kong. MSCI classifies the city’s average physical climate risk, which arises principally from flooding, as significant — an average of 18% value at risk. But 83% of properties have only a moderate risk (meaning between 0.5% and 5% value at risk), while 16% of assets are at severe risk, due to their micro-location.
“Since flood risk is mostly dependent on elevation, assets located in close proximity to each other [in Hong Kong] could face varying degrees of climate risk,” he said.
“In cities where coastal or fluvial flooding poses the major risk, a property’s micro-location can have a major influence on its value at risk. This contrasts with a hazard like a tropical cyclone, which has much more uniformity on the likelihood and its expected impact across assets in a city, since local topography is less of a determining factor,” he added.
Harmse therefore cautioned investors against relying solely on market-level data in a sector where location and topography were decisive factors behind the impact of physical climate change. “This underlines the importance of asset-level analysis and benchmarking,” he said.
GOVERNMENT ROLE
Welling said that governments could help investors by increasing the stock of publicly available information related to physical climate risks, pointing to the Climate Impact Atlas (Klimaateffectatlas) in the Netherlands.
“[Such models should] ideally be maintained by climate research experts,” he said. He cautioned on the effectiveness of global models, adding they always come with a level of adjustments that might overestimate or underestimate the risks in a particular country or location.
In Asia, where many of the region’s largest and most economically significant cities face substantial risks from natural disasters, there is significant variation in risk.
According to MSCI research published in May, Mumbai is the region’s most at-risk city, with 64% of properties facing severe risk, followed by three of China’s largest cities, Guangzhou (59%), Shenzhen (51%), and Shanghai (40%). Hong Kong was 10th, and Tokyo 16th. Singapore, Melbourne, and Sydney were among the five least at-risk cities.
Seventeen percent of all Asian cities were exposed to coastal flooding, with assets in Japan and Eastern China the most exposed. Meanwhile, 74% were exposed to tropical cyclones, where assets in the Philippines, Japan, and the Eastern coast of China had the highest exposure to the phenomenon.
One hundred percent of the assets were exposed to extreme heat, with South Asian and Southeast Asian assets having the highest exposure.
This story has been updated in para 6.