articles Ratings /ratings/en/research/articles/210219-damage-limitation-using-enhanced-physical-climate-risk-analytics-in-the-u-s-cmbs-sector-11839707 content esgSubNav
Log in to other products

 /


Looking for more?

In This List
COMMENTS

Damage Limitation: Using Enhanced Physical Climate Risk Analytics In The U.S. CMBS Sector

NEWS

Turbo Finance 8 Class C U.K. ABS Notes Raised; Other Ratings Affirmed

Take Notes: How The Southern European Nonperforming Loan Market Is Performing

NEWS

AyT Genova Hipotecario IX Spanish RMBS Ratings Raised On Three Classes Following Criteria Revision; One Affirmed

NEWS

AyT Genova Hipotecario VI Spanish RMBS Ratings Raised On Three Classes Following Criteria Revision; One Affirmed


Damage Limitation: Using Enhanced Physical Climate Risk Analytics In The U.S. CMBS Sector

(Editor's Note: Rick Lord and Steven Bullock at Trucost (part of S&P Global) and Alka Dagar of S&P Global Market Intelligence also contributed to this article.)

Acute, extreme weather or natural disaster events--such as storms, hurricanes, wildfires, and earthquakes--are commonly perceived as unpredictable, rare phenomena with potentially severe consequences. Such "black swan" events may cause severe physical damage to the real estate backing commercial mortgage-backed securities (CMBS) loans and consequently pose serious threats to the entire industry. Depending on the circumstances, the damage may go well beyond repairs. The associated loss in property value may be significant, potentially reducing the value of a property to its land--that in extreme cases may even turn negative.

To put things in context, the five most costly natural disasters in the U.S. all occurred in recent history:   the 2005 hurricane Katrina ($170 billion of CPI-adjusted overall damages), 2017 hurricane Harvey ($131 billion), 2017 hurricane Maria ($94 billion), 2012 hurricane Sandy ($75 billion), and 2017 hurricane Irma ($53 billion), according to National Oceanic and Atmospheric Administration (NOAA). The U.S. has had to deal with 285 weather and climate disasters since 1980 (where the total damage costs reached or exceeded $1 billion) for a total cost that NOAA estimate at $1.875 trillion over the same period. What's more, 2020 set a new record, seeing the occurrence of 22 of these events (compared with 16 previously in both 2011 and 2017), with a total cost of $95 billion. Last year was also the sixth consecutive year (2015-2020) when 10 or more billion-dollar weather and climate disaster events took place in the U.S., says NOAA. According to the California Department of Forestry and Fire Protection, four of the top five largest California wildfires occurred in 2020, with more than 4 million acres burned.

For several years now, we have been commenting about the impacts of physical risks, themselves manifesting as chronic, longer-term shifts in climate patterns or acute risks from extreme weather events (see Related Research). Further, we've described how our methodology and ratings can capture the dual threat of chronic and acute physical risks when they are sufficiently visible and certain and actually or potentially material (see "The Role Of Environmental, Social And Governance Credit Factors In Our Ratings Analysis," Sept. 12, 2019).

While a building may change hands over its entire life, it represents a long-term investment (typically 10 years or more) for its current owners and therefore requires adequate forms of protection against physical risks and natural disasters.   With an expected economic life of typically 50 years, real estate is also built to last and therefore requires significant capital to design, build, maintain, renovate, and potentially redevelop when it no longer meets users' requirements--starting a new life cycle. Physical climate risks and natural disasters have long been considered within our analytical framework for rating U.S. CMBS transactions (see "Insurance Criteria For U.S. And Canadian CMBS Transactions," June 13, 2013, and "ESG Credit Factors In Structured Finance," Sept. 19, 2019). In most cases, these factors are not key rating drivers because the risks are structurally mitigated at the loan level with the use of appropriate property insurance. We believe that adequate property insurance helps focus our analysis on the credit risk of the loans and the underlying properties as factors that may affect defaults and losses. If, in our view, the property is insufficiently insured, our methodologies call for an increase to the minimum amount of credit enhancement, or we may decide not to assign ratings.

In this context, S&P Global Ratings explores the application of Climate Change Physical Risk data from Trucost (part of S&P Global) in an analysis of 11,501 properties backing U.S. CMBS transactions we rate. For our analysis, we assume no change in pool composition over the various time horizons described below. We explore how such data might enhance our dialogue with issuers and market participants regarding future physical climate risks, the possible range of exposures, and the role of insurance in lessening risk. Trucost's data is derived from publicly available information, licensed datasets, and its own models.

Climate Projections, Physical Climate Risk Dataset, And Earthquake Data

There are some inherent uncertainties associated with climate science, including the crystallization, frequency, and severity of climate risks--like water stress, wildfire, and sea level rise. Given these uncertainties and when considered immaterial, we do not apply the scenarios discussed in this report as part of our base case for rated U.S. CMBS transactions. In addition, the exposures to physical risks we describe exclude adaptation efforts that properties backing U.S. CMBS transactions have implemented or may implement in the future. However, the scenarios we explore below may lead to a better understanding of the kinds of exposures that properties backing U.S. CMBS transactions may face.

While issuers are familiar with the impacts of acute risks (such as extreme weather events), for some, other significant physical risks (such as sea level rise) may emerge over the medium to longer term, time scales that are sometimes longer than those of many issuers' financial forecasts. At the same time, the precise timing and severity of impacts remains uncertain. This uncertainty presents a challenge to understanding the potential impacts of physical climate risks and the steps required to build additional resilience. Scenario analysis has long been used as a tool to build organizational resilience and identify risks and opportunities before they emerge. However, its use for assessing climate-related risks and opportunities by many issuers is not common or is relatively new. As global economies continue their (green) recovery following the COVID-19 pandemic, it is argued by some that the real estate industry in general must take positive action to transition to a low-carbon future.

The climate scenarios (or Representative Concentration Pathways), RCPs describe possible pathways of future greenhouse gas (GHG) emissions and were produced by the Intergovernmental Panel on Climate Change (IPCC) and used in its Fifth Assessment Report (AR5). In our analysis, we apply three of the scenarios (RCP2.6, RCP4.5, and RCP8.5; see box) to gauge how properties backing U.S. CMBS transactions are exposed to a range of climate hazards, including acute risks such as wildfire and hurricane, and how chronic risks like heat waves, water stress, and sea level rise may evolve to 2050. Because U.S. CMBS industry standards still generally use a 10/30 amortization schedule (referring to a 10-year term loan with a 30-year amortization profile), this time frame fits relatively well with the financial horizon of the long-term debt used to finance these properties. Due to limited data availability, we do not apply the RCP6.0 scenario.

Our analysis also focuses on nearer timepoints (that is, 2030) where changes in physical risk scores for certain hazards may be material. However, a certain amount of change is locked in due to the lag in the climate system owing to historic GHG emissions. Therefore, there is sometimes little difference between the RCPs (and resulting physical risk scores) for timepoints before midcentury. Given the effect of this lock-in and uncertainty in modelling some hazards (climate scientists have greater certainty over the direction and magnitude of change in average temperature than, for example, wind), less emphasis should be placed on physical risk scores with nearer timepoints.

Note also in our analysis, exposure to hurricanes is taken as present day (that is 2020) as reliable projections for this hazard are unavailable. Nevertheless, many scientists expect an increase in extreme wind speeds in the future over Europe, parts of Central and North America, the tropical South Pacific, and the Southern Ocean; a poleward shift of storm tracks; and associated changes in wind patterns.

Owing to the exposure of some properties backing U.S. CMBS transactions to seismic hazards, our analysis considers earthquakes using U.S. Geological Survey (USGS) data. We use peak ground acceleration (PGA) with a 10% probability of exceedance in 50 years as a proxy for earthquake exposure. The USGS data represents seismic site classifications B and C. The classifications are broadly representative of the contiguous U.S. and are used in the International Building Code (IBC) and seismic provisions in Minimum Design Loads for Buildings and Other Structures (ASCE 7-05). Specifically, we identify U.S. CMBS transactions that are located within areas deemed highly exposed to earthquake risk, where PGA is greater than or equal to 0.4g, equivalent to seismic zone 3 or 4.

U.S. West Coast Exposure To Wildfires Worsens

Wildfires are not a new phenomenon to states on the U.S. West Coast, but the 2020 wildfire season was particularly devasting, shattering numerous records. About 4% of California, or about 4 million acres, were affected (more than double the size of the state's previous wildfire record) and the largest fire (known as the August Complex fire) was notable as the biggest on record, burning over 1 million acres.

The relationship between rising temperatures under climate change and fire extent, particularly in the U.S., has been known for some time. It is unsurprising then that the U.S. West Coast contains eight of the top 10 states with the highest average exposure to wildfire in 2050 under a high stress (RCP 8.5) scenario (see our commentary on the exposure of U.S. public finance to this hazard "Better Data Can Highlight Climate Exposure: Focus On U.S. Public Finance," Aug. 24, 2020).

Out of the 11,501 properties backing the U.S. CMBS transactions we rate, 605 are already highly exposed to wildfire (with a score of 70 or greater on our 1-100 scale). These represent 5% of the total number of properties. We calculate that 99% of the highly exposed assets are spread across 10 states, with California concentrating most of the risk.

Under the high stress (RCP8.5) scenario, we expect the number of properties exposed to wildfire at first to increase by 1% (up to 612 properties) in 2030. However, by 2050 we see a significant 8% increase (up to 664 properties or 6% of the U.S. CMBS universe we rate; see chart 1) in the absence of adaptation measures. In California, exposure is expected to remain broadly unchanged under the same high stress scenario (but with relatively high scores). We specifically note the increase in the number of properties with high exposures in Washington (from 0 to 61, placing it in the top five most exposed states in 2050).

By way of comparison, the low stress (RCP2.6) scenario suggests significantly lower exposure in 2050 (509; 4% of the U.S. CMBS universe). California would still be the most exposed state, but its relative weight for this hazard is less than what it is today.

Chart 1

image

California is projected to remain the most exposed state under all climate scenarios in our analysis (see table 1). The state's population is also greater than the other states and research (Radeloff et al., 2018) has shown a causal link between population density and sensitivity to wildfire risk. This is because the probability of human ignitions increases where people are more highly concentrated, and the wildfires that do occur pose a greater risk to property, while letting the fires burn becomes more difficult.

Table 1

Change In High Wildfire Exposure (Scores Of 70 Or Greater) Of The Top 10 U.S. States Under Several Climate Scenarios
Baseline, 2030 under RCP8.5, and 2050 under RCP8.5 and RCP2.6
State Population density (people/sq. mile) No. properties backing rated U.S. CMBS transactions No. properties with high exposure to wildfire in 2020 (baseline), 2030 (RCP 8.5), and 2050 (RCP 8.5- RCP 2.6) % properties with high exposure to wildfire in 2020 (baseline), 2030 (RCP 8.5), and 2050 (RCP 8.5- RCP 2.6) Ranking in 2020 (baseline), 2030 (RCP 8.5), and 2050 (RCP 8.5-RCP 2.6)
CA 254 1,210 266-275-279-187 22-23-23-15 1-1-1-1
AZ 66 284 111-110-111-113 39-39-39-40 2-2-2-2
NV 29 167 82-82-82-82 49-49-49-49 3-3-3-3
UT 40 77 62-62-62-61 81-81-81-79 4-4-4-4
ID 23 29 23-23-23-18 79-79-79-62 5-5-6-5
LA 107 126 20-20-14-18 16-16-11-14 6-6-8-6
CO 57 217 17-17-19-16 8-8-9-7 7-7-7-7
MS 63 78 9-9-0-7 12-12-0-9 8-8->10-8
OR 45 95 7-7-8-6 7-7-8-6 9-9-9-9
AR 66 56 3-2-0-0 5-4-0-0 10->10->10->10
Top 10 N/A 2,339 600-607-598-508 26-26-26-22 N/A
States outside the top 10 N/A 9,162 5-5-66-1 0-0-1-0 N/A
Total N/A 11,501 605-612-664-509 5-5-6-4 N/A
N/A—Not applicable. Data as of Nov. 11, 2020. Sources: U.S. Census Bureau Population Estimate 2021, S&P Global Ratings, Trucost.

Looking at the same dataset by Metropolitan Statistical Area (MSA),smaller but growing MSAs, such as Sacramento-Roseville-Arden-Arcade and Riverside-San Bernardino-Ontario, and Phoenix-Mesa-Chandler are the most exposed (see table 2). These MSAs each have greater exposure than Los Angeles-Long Beach-Anaheim. While the exposure of the five most exposed MSAs are broadly similar from the baseline to 2050 under a high stress (RCP8.5) scenario, we expect most of the changes to take place further down the stack in smaller MSAs, such as Seattle-Tacoma-Bellevue. Notably, in 2050 under the low stress (RCP2.6) scenario, the exposure of Sacramento-Roseville-Arden-Arcade decreases.

Table 2

Change In High Wildfire Exposure (Scores Of 70 Or Greater) Of The Top 5 U.S. MSAs Under Several Climate Scenarios
Baseline, 2030 Under RCP8.5, And 2050 Under RCP8.5 And RCP2.6
MSA No. properties backing rated U.S. CMBS transactions No. properties with high exposure to wildfire in 2020 (baseline), 2030 (RCP 8.5), and 2050 (RCP 8.5- RCP 2.6) % properties with high exposure to wildfire in 2020 (baseline), 2030 (RCP 8.5), and 2050 (RCP 8.5- RCP 2.6) Ranking in 2020 (baseline), 2030 (RCP 8.5), and 2050 (RCP 8.5-RCP 2.6)
Sacramento--Roseville-Arden-Arcade, CA 126 82-82-83-27 65-65-66-21 1-1-1-5
Riverside-San Bernardino-Ontario, CA 120 77-77-77-77 64-64-64-64 2-2-2-1
Phoenix-Mesa chandler, AZ 225 73-73-73-73 32-32-32-32 3-3-3-2
Las Vegas-Henderson-Paradise, NV 143 58-58-58-58 41-41-41-41 4-4-5-3
Los Angeles-Long Beach-Anaheim, CA 406 48-48-48-48 12-12-12-12 5-5->5-4
Top 5 1,020 338-338-339-283 33-33-33-28 N/A
MSAs outside top 5 10,481 267-274-325-226 3-3-3-2 N/A
Total 11,501 605-612-664-509 5-5-6-4 N/A
N/A—Not applicable. Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

Three MSAs on the U.S. West Coast are particularly exposed to wildfire risk in the absence of adaptation measures, comprising Las Vegas-Henderson-Paradise, Reno, and Provo-Orem. These MSAs have the greatest number of commercial properties with maximum wildfire exposure scores (scoring 100 on our 1-100 scale) in 2050 under the high stress (RCP8.5) scenario.

Table 3

Most Exposed MSAs On U.S. West Coast
Under baseline scenario compared to RCP 2.6 and RCP 8.5 in 2030 and 2050
MSA No. properties backing rated U.S. CMBS transactions No. properties with high exposure to wildfire in 2020 (baseline), 2030 (RCP 8.5), and 2050 (RCP 8.5- RCP 2.6) % properties with high exposure to wildfire in 2020 (baseline), 2030 (RCP 8.5), and 2050 (RCP 8.5- RCP 2.6) Ranking in 2020 (baseline), 2030 (RCP 8.5), and 2050 (RCP 8.5-RCP 2.6)
Las Vegas-Henderson-Paradise, NV 143 58-58-58-58 41-41-41-41 1-1-1-1
Reno, NV 20 20-20-20-9 100-100100-45 2-2-2-2
Provo-Orem, UT 11 11-11-11-0 100-100-100-0 3-3-3,>5
Riverside-San Bernardino-Ontario, CA 120 7-7-7-7 6-6-6-6 4-4-4-3
Los Angeles-Long Beach-Anaheim, CA 406 4-4-4-4 1-1-1-1 5-5-,>5,4
Top 5 700 100-100-100-78 14-14-14-11 N/A
MSAs outside top 5 10,801 13-14-23-3 0-0-0-0 N/A
Total 11,501 113-114-123-81 1-1-1-1 N/A
N/A—Not applicable. Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

The causes of wildfires may be natural, for example, lightning or ignition of dry vegetation by the sun; or human, like unattended campfires. Many other factors contribute to the number of wildfires in an area in any given year, including how high summer temperatures are, how low precipitation is, and wind conditions. Research suggests a strong relationship between temperature and fire extent, particularly in the U.S., with warmer years generally having greater fire extent (principally due to fuel aridity) than relatively cooler ones, since the early 1980s. On a global level, other factors are at play including the frequency of human-set fires for agricultural conversion, particularly in Africa and Southeast Asia. In the future, evidence suggests that climate change will have a bigger effect in areas outside the tropics than human-caused factors (see Yang et al., 2014). While the long-term change in climate that may increase the risk of wildfire events is relatively visible, it is not possible to precisely predict where and when specific wildfire events will happen and what damage they may cause. By their nature, wildfires (like heavy summer rainfall events in many parts of the world) are highly localized. Notwithstanding this, the potential increasing exposure over time highlights the importance of dialogue and learning about how properties backing U.S. CMBS transactions within these areas consider these risks and whether they have measures in place to reduce wildfire risk.

Modelling highly localized events, like wildfires, is challenging as local conditions (including topography and wind patterns) are not easily replicated at scale in global climate models. It is currently a challenge to model changing wind patterns (which can fuel wildfire intensity) in wildfire projections with the available science. Model limitations could obscure some of the likely changes in intensity that may happen over the next 30 years. However, wildfire risk in California has affected some of our (non-U.S. CMBS) credit ratings (see "California Public Power Utilities Face Disparate Physical And Credit Exposures To Wildfires," Aug. 4, 2020).

Earthquake Exposure Remains A Key Concern For U.S. CMBS

The exposure of properties backing U.S. CMBS transactions to earthquakes remains a material threat for West Coast states and elsewhere, including parts of South Carolina. Although the relationship between human-induced climate change and earthquake risk may not be immediately obvious, evidence has existed for some time (see Liu et al., 2009) and has also emerged (see NASA, 2019) showing a causal link between seismic activity and increasing frequency of some climate hazards. For U.S. CMBS, earthquakes represent a material risk and are typically mitigated through insurance coverage, specifically when mortgaged properties are located in areas that are considered at high earthquake risk and the probable maximum loss on the property exceeds 20% of the replacement costs under a 475-year return period earthquake. The 475-year return essentially translates into a 10% probability that a severe earthquake would occur during a 50-year period (which is akin to the economic life of a property). Earthquake severity assessments are based on the PGA that occurs during earthquake shaking at a given location.

As noted earlier, we use PGA with a 10% probability of exceedance in 50 years as a proxy for earthquake exposure. Specifically, we identify properties in our database that are located within areas deemed highly exposed to earthquake risk (where PGA is greater than or equal to 0.4g, equivalent to seismic zone 3 or 4). With this in mind, properties in California are most highly exposed to this hazard as the state straddles the San Andreas Fault (see chart 2).

Chart 2

image

Three hundred sixty-one (361) of 608 (or 59%) properties highly exposed to earthquake (where PGA >0.4g) are located in Los Angeles-Long Beach-Anaheim and San Francisco-Oakland-Berkeley (see table 4). The remaining exposures are spread out through California, with the exception of 15 in Charleston-North Charleston and eight in Reno and Carson City.

Table 4

MSAs With The Highest Number And Proportion Of Properties Exposed To Earthquake Risk
(PGA >0.4g)
State MSA (rank) No. properties backing rated U.S. CMBS transactions No. properties highly exposed to earthquakes (PGA >=0.4g) % properties highly exposed to earthquake (PGA >=0.4g)
CA Los Angeles-Long Beach-Anaheim, CA (2) 406 236 58
San Francisco-Oakland-Berkeley, CA (12) 131 125 95
Riverside-San Bernardino-Ontario, CA (13) 120 70 58
Santa Rosa, CA (112) 23 23 100
Others MSAs 445 46 10
CA total N/A 1,210 585 48
SC Charleston-North Charleston, SC (74) 33 15 45
NV Reno & Carson City, NV (115, 384) 23 8 35
Grand total 11,501 608 5
Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

Table 5 describes the exposure for all properties included in our analysis to earthquake risk using PGA. The dark blue CMBS in chart 2 are the values listed in row 3 of the table, where PGA is between 0.4-0.8g. Note, there are no properties located in areas where the PGA is greater than or equal to 0.8g.

Table 5

Exposure Of Properties Backing U.S. CMBS Transactions To Earthquake Risk Using Peak Ground Acceleration
Peak ground acceleration (g) % and no. properties* highly exposed to earthquake (PGA >=0.4g)
>0.8 0% (0)
0.40-0.80 5% (608)
0.20-0.40 6% (696)
0.15-0.20 1% (167)
0.10-0.15 3% (302)
0.07-0.10 3% (375)
0.05-0.07 6% (692)
0.03-0.05 27% (3,096)
0.02-0.03 20% (2,320)
0.01-0.02 16% (1,876)
<0.01 10% (1,154)
Not available* 2% (215)
Total 100% (11,501)
*Some properties fall outside of the USGS dataset (that is, Hawaii and Alaska). Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

Hurricanes Remain A Challenge For U.S. CMBS

Properties backing U.S. CMBS transactions in Florida are the most exposed to hurricanes under the baseline (2020) scenario (see table 6 and chart 3). As noted earlier, due to the uncertainty associated with projections of this hazard, our analysis considers only the baseline scenario.

Chart 3

image

A total of 826 properties are located in Florida (7% of the portfolio). Of these, 22% (183) achieve the maximum score of 100, while 39% (322) are also highly exposed (that is, they receive a score of 70 or greater on our 1-100 scale). Florida's high population (it is the third most populous state in the U.S.) also amplifies the state's sensitivity to hurricanes since research by Ashley and Strader (2016) has demonstrated a link between expanding "targets" (that is, people and their built environments), which occurs as populations grow and spread, otherwise known as the bull's eye effect. Further, a recent study by Fisher et al., (2018) described the long-term effects of hurricanes on five property types (apartments, industrial, office, retail, and hotels), finding that valuations decreased by almost 6% one year after the storm and further declined to 10.5% two years out. At the MSA level, Miami-Fort Lauderdale-West Palm Beach scores on average 82 out of 100, with 82% (164 out of 201) of scores 70 or greater--and all of these scores are incidentally the maximum score of 100. Key West (score 95 out of 100), Port St. Lucie (score 92 out of 100), and Pensacola-Ferry Pass-Brent (85 out of 100) in Florida all receive scores of 70 or greater, and therefore are highly exposed in the absence of adaptation measures.

Texas is another populous state (the second most in the U.S) with noticeable exposure to hurricane risk. In our analysis, 28% (358) properties in Texas are highly exposed to hurricane risk (achieving scores of 70 or greater on our 1-100 scale). In terms of MSAs, Houston-The Woodlands-Sugar Land is particularly exposed with 338 (84%) of the properties in our analysis highly exposed. In contrast, no properties in the Dallas-Fort Worth-Arlington MSA are highly exposed to hurricane risk.

Outside Florida and when averaged at the state level, the hurricane exposure of properties backing U.S. CMBS transactions is muted given the relatively large number of properties that receive low scores (30 or less out of 100). The only exception, with an average score of 82, is Puerto Rico. There, however, the exposure is limited to only two properties.

Table 6

Exposure Of States At High Risk Of Hurricane
In the baseline scenario
State No. (%) of properties backing rated CMBS transactions with high exposure (scores of 100) No. (%) of properties backing rated CMBS transactions with high exposure (scores of 70 or greater) % of properties with high exposure (scores of 70 or greater)/total properties backing rated U.S. CMBS transactions
Florida 183 (22%) 322 (39%) 3
Miami-Fort Lauderdale-West Palm Beach, FL MSA 164 (82%) 164 (82%) 1
Texas 22 (2%) 358 (28%) 3
Of which Dallas-Fort Worth-Arlington, TX MSA 0 (0%) 0 (0%) 0
Of which Houston-The Woodlands-Sugar Land, TX MSA 17 (4%) 338 (84%) 3
Other states 12 (0%) 56 (1%) 0
Total 217 (2%) 736 (6%) 6
Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

Exposure To Sea Level Rise, Flooding, And Heat Waves Are Muted

Our analysis found that no states containing properties backing U.S. CMBS transactions are likely to be highly exposed to sea level rise in any scenario or timepoint (that is, have scores of 70 or greater out of 100). This is due to the averaging effect, with far greater numbers of properties with low exposure scores. Despite this, individual properties representing less than 1% of the total (67 out of 11,501) are currently highly exposed to sea level rise with scores of 70 or greater out of 100. We expect the number of highly exposed properties to increase from 67 to 95 (representing a 42% rise) by 2050 under a high stress (RCP8.5) scenario. The exposure is broadly similar under a low stress (RCP2.6) scenario by midcentury (94 out of 11,501). A large number (41 or 43%) of these 95 are located in California, Louisiana, and Florida (see chart 4). None of the 95 represent single assets within single-asset, single-borrower transactions. We identified only seven properties in the New York-Newark-Jersey City MSA and 10 in the San Francisco-Oakland-Hayward MSA but once again, none form the sole and unique collateral within a given transaction.

Chart 4

image

Given their location on an archipelago, all properties in Key West (Florida) scored a maximum of 100 on our scale under all scenarios and timepoints. Changing exposure to unmitigated sea level rise remained generally muted. An exception is New Orleans-Metarie, which may expect to see an increase in exposure to sea level rise (in the absence of adaptation measures) of 15 points under the low (RCP2.6) scenario (from 17 to 32) and 16 points under the high (RCP8.5) scenario (from 17 to 33) by 2050. Although the average state scores remain low, this corresponds to an increase in the proportion of properties highly exposed to sea level rise (with scores of 70 or greater) from 6% (3 of 54) to 31% (17 of 54) under both scenarios by 2050.

For river flooding, less than 1% (41 out of 11,501) of properties backing rated U.S. CMBS transactions are highly exposed (a score of greater than 70) to river flooding in 2050 under the high stress (RCP8.5) scenario (see chart 5). Our analysis shows that changing exposure to this hazard is marginal by midcentury. Further, exposure scores to coastal flooding are lower than river flooding, and changes are minimal. Of note, our analysis does not capture impacts from storm surge, which are attributed to hurricane events and that can cause severe flooding in coastal areas.

Chart 5

image

For heat waves, we expect no properties backing U.S. CMBS transactions in our analysis to be highly exposed (receiving a score of 70 of more) by 2030 or 2050 under a high stress (RCP8.5) scenario (see table 7). Southeastern U.S. states (but principally Florida), are projected to experience the greatest exposure to heat waves in 2050 under a high stress (RCP8.5) scenario (the state may experience about an extra three weeks of heat wave conditions in 2050 under RCP8.5). However, no properties in our analysis are located within highly exposed areas. That said, 8% (933 out of 11,501) of CMBS may experience moderate exposure (scoring between 30 and 69 on our 1-100 scale) in 2050 under RCP 8.5 (for 2030, the proportion is close to 0% or three out of 11,501 under the same scenario). Under the low stress scenario (RCP2.6), no properties are highly exposed to heat wave risk by 2050, however, less than 1% (10 out of 11,501) are moderately exposed over the same time frame.

Chart 6

image

Florida is the state that we expect to have the greatest proportion of properties moderately exposed to heat wave under all scenarios, increasing from 0 in the baseline scenario to 82% (679) under the high stress scenario (RCP 8.5) by 2050. Hawaii is the state we expect to see the greatest increase in number of heat wave days (38 days), corresponding with no properties in the baseline scenario to 100% of properties (204) in the high stress scenario (RCP 8.5) by 2050.

Table 7

Change In Moderate Heat Wave Exposure (Scores Between 30 To 69) In Florida And Hawaii
Under the baseline scenario and RCP8.5 in 2050
State No. of properties backing rated U.S. CMBS transactions No. of properties with high exposure to wildfire in 2020 (baseline), 2030 (RCP 8.5), And 2050 (RCP 8.5-RCP 2.6) % of properties with high exposure to wildfire in 2020 (baseline), 2030 (RCP 8.5), And 2050 (RCP 8.5-RCP 2.6) Ranking in 2020 (baseline), 2030 (RCP 8.5), and 2050 (RCP 8.5-RCP 2.6)
FL 826 0-0-679-7 0-0-82-1 10->10-1-1
HI 204 0-0-204-0 0-0-100-0 >10->10-2->10
Rest of U.S. 10,471 0-3-50-3 0-0-0-0 N/A
Total 11,501 0-3-933-10 0-0-8-0 N/A
Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

Exposure To Water Stress Is Widespread, Though Less Material

Of the modelled hazards, exposure to water stress is greatest. A greater score on this scale means that water demand is likely to outstrip the renewable supply, and so water resources could deplete over time. As we describe in a U.S. public finance commentary ("Better Data Can Highlight Climate Exposure: Focus On U.S. Public Finance," Aug. 24, 2020), this is problematic for those states that draw significant resources from those with greater supply and may necessitate a shift toward groundwater supplies when surface water supplies decline, increasing costs.

Despite significant exposure to water stress, properties backing U.S. CMBS transactions are not typically high water users (with the exception of multifamily and lodging properties) and water resources tend to be well managed. However, increasing water scarcity and drought severity and frequency, particularly in already water-scarce states, may exacerbate exposure to this hazard and increase costs in the absence of adaptation measures. Large developments located close to properties could lead to significant impacts, while increasing water scarcity may increase property owners' operational costs.

Although water stress is a less material risk given that properties backing U.S. CMBS transactions are not high water users, some states do receive the maximum 100 score on our 1-100 scale, including Utah, Wyoming, and Colorado. These states receive the maximum 100 score under a high stress scenario (RCP 8.5) in both 2030 and 2050, as well as under the low stress scenario (RCP2.6) by 2050. MSAs receiving the maximum 100 score under all scenarios include Phoenix-Mesa-Scottsdale, Denver-Aurora-Lakewood, El Paso, and Salt Lake City. We do not expect the number of CMBS highly exposed to water stress to increase significantly under either the high (RCP8.5) or low (RCP2.6) stress scenarios in 2030 or 2050. However, states that draw significant resources from those with greater supply may necessitate a shift toward groundwater supplies when surface water supplies decline, which may increase production costs. Chart 7 presents the exposure to water stress under the high stress (RCP8.5) scenario in 2050.

Chart 7

image

Climate Change May Also Benefit Some CMBS

Exposure to cold waves drops by midcentury under all scenarios owing to warmer temperatures in many regions. Reduced exposure to cold waves may lower properties' operational costs and disruption related to snow, ice, and severe frost events in affected territories. Less use of heating, ventilation, and air conditioning (HVAC) in winter may also correspond to reductions in GHG emissions. However, it is unclear if increased use of HVAC systems in summer months may offset any benefits (including increased GHG emissions) gained through warmer winter temperatures.

We do not factor these opportunities into our ratings given the uncertainties, including the precise timing and geographic areas that may benefit from the manifestation of opportunities that could emerge due to climate change.

Insurance Is A Key Credit Risk Mitigant

The quality of insurance coverage is an important point of consideration, although we recognize that modern and refurbished or retrofitted buildings (that is, designed to resist or mitigate extreme weather events) as well as certain actions taken by local authorities to minimize exposure to extreme weather events may enable landlords, in some cases, to offset some of the potential physical impacts of climate change and natural disasters. Indeed, its main purpose is to protect property against losses and ultimately preserve the lender's security in the underlying asset in case of damage. While traditional U.S. CMBS loans require property insurance coverage equal a property's full replacement cost to protect against fire and casualty events, extreme weather events and natural disasters typically require additional policies to backfill potential exclusions from traditional insurance policies, especially when assets are located in zones considered at risk. For example, borrowers are required to obtain additional earthquake insurance if the mortgaged properties are located in an area considered to be a high earthquake risk and the probable maximum loss (PML) on the property exceeds 20% of the replacement costs. Similarly, properties located in federally designed special flood hazard areas require additional coverage to mitigate flood risk.

Although actual CMBS loans are already structured to mitigate physical climate risks, there is agreement among commentators that climate change is acting as an accelerating factor for natural hazards. They expect the impacts from (unmitigated) physical climate risks to worsen over time. Depending on the region, climate experts anticipate more frequent wildfires, longer periods of drought, an increase in the number and more powerful storms, and an increased number of heavy precipitation events, for example.

The higher frequency and potential higher severity of these events will likely have knock-on effects on commercial real estate:

  • First, with immediate impacts upon occurrence, will be the potential for an increase in the amount and value of damage that may result from more frequent and harmful natural disaster events (also partially due to more expensive technologies embedded in the assets).
  • Second, potentially less visible, will be the incremental increase of insurance premiums and property capital expenditures to prevent or offset these risks.

These two catalysts, in our view, may weaken financial returns for the commercial real estate sector depending on insurance providers' ability to adapt their offers and tenants' capacity to absorb these additional costs.

Physical climate risks have long been considered within our analytical framework for U.S. CMBS transactions. In most cases, they are not key rating drivers because the risks are structurally mitigated at the loan level with the use of appropriate property insurance. We believe that adequate property insurance helps focus our analysis on the credit risk of the loans and the underlying properties as factors that may affect defaults and losses. If, in our view, the property is insufficiently insured, our criteria call for an increase to the minimum amount of credit enhancement, or we may decide not to assign or withdraw our ratings.

How Data Can Enhance Transparency Of Physical Climate Risks For U.S. CMBS Participants

We welcome additional information about the anticipated impacts of physical climate risks and actions available to mitigate them and more generally about natural hazard exposure. Additional insights about physical climate risks and trends over a longer period will likely at first complement, then probably supersede, some or all the traditional tools market participants currently use to assess such risks. This type of data will help to improve risk awareness, enhancing the quality and depth of the dialogue among all stakeholders, and refine the risk analysis to ultimately support more informed decisions and holding strategies.

Although environmental credit factors have not been key rating drivers to date, alternative climate physical risk data can enhance our ability to identify future emerging physical risks and deepen our dialogue with rated entities that may face material exposure in the future. As a result, we believe that making physical exposure information available to all our stakeholders should be a priority, with the ultimate goal of supporting transparency and awareness. With that objective in mind, we are exploring the possibility of sharing more transaction-specific information about climate hazards in our publications.

Ultimately, if climate scenarios highlight material unmitigated exposures or point to major uncertainties and therefore risk, they can weigh on our views of current and future credit quality, if not mitigated by other credit factors such as capital and financial planning and coordination with regulators and other entities. While some risks may not be credit factors in the short term, their consequences could be gradual or lead to increased uncertainty of revenues, higher operating costs, and volatility of cash flows. As such, they could weigh on our assessment of long-term sustainable recovery values of those more exposed U.S. CMBS.

Appendix: Aggregated Physical Risk Scores By Climate Hazard And Decile Under Different Climate Scenarios And Timepoints

Please note the below tables may be subject to rounding errors.

Table A1

Aggregated Physical Risk Scores For Properties Backing U.S. CMBS Transactions By Climate Hazard And Decile Under The Baseline Scenario
Wildfire Cold wave Heatwave Water stress River flood Hurricane Sea level rise
Low score 6,885 (60%) 0 (0%) 11,491 (100%) 2,526 (22%) 11,231 (98%) 10,058 (87%) 11,360 (99%)
Medium score 3,789 (33%) 219 (2%) 10 (0%) 4,103 (36%) 129 (1%) 631 (5%) 74 (1%)
High score 827 (7%) 11,282 (98%) 0 (0%) 4,872 (42%) 141 (1%) 812 (7%) 67 (1%)
Maximum score (100) 305 (3%) 42 (0%) 0 (0%) 3,592 (31%) 70 (1%) 236 (2%) 64 (1%)
Total 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%)
Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

Table A2

Aggregated Physical Risk Scores For Properties Backing U.S. CMBS Transactions By Climate Hazard And Decile Under The High Stress (RCP 8.5) Scenario By 2030
Wildfire Cold wave Heatwave Water stress River flood Hurricane Sea level rise
Low score 6,661 (58%) 3 (0%) 10,588 (92%) 2,622 (23%) 11,235 (98%) 10,058 (87%) 11,356 (99%)
Medium score 4,037 (35%) 2,088 (18%) 913 (8%) 4,128 (36%) 209 (2%) 631 (5%) 66 (1%)
High score 803 (7%) 9,410 (82%) 0 (0%) 4,751 (41%) 57 (0%) 812 (7%) 79 (1%)
Maximum score (100) 318 (3%) 0 (0%) 0 (0%) 3,534 (31%) 19 (0%) 236 (2%) 66 (1%)
Total 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%)
Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

Table A3

Aggregated Physical Risk Scores For Properties Backing U.S. CMBS Transactions By Climate Hazard And Decile Under The High Stress (RCP 8.5) Scenario By 2050
Wildfire Cold wave Heatwave Water stress River flood Hurricane Sea level rise
Low score 6,307 (55%) 207 (2%) 8,692 (76%) 2,668 (23%) 11,236 (98%) 10,058 (87%) 11,339 (99%)
Medium score 4,326 (38%) 11,102 (97%) 2,809 (24%) 4,368 (38%) 216 (2%) 631 (5%) 67 (1%)
High score 868 (8%) 192 (2%) 0 (0%) 4,465 (39%) 49 (0%) 812 (7%) 95 (1%)
Maximum score (100) 336 (3%) 0 (0%) 0 (0%) 3,520 (31%) 20 (0%) 236 (2%) 80 (1%)
Total 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%)
Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

Table A4

Aggregated Physical Risk Scores For Properties Backing U.S. CMBS Transactions By Climate Hazard And Decile Under The Low Stress (RCP 2.6) Scenario By 2050
Wildfire Cold wave Heatwave Water stress River flood Hurricane Sea level rise
Low score 7,091 (62%) 3 (0%) 10,737 (93%) 2,436 (21%) 11,251 (98%) 10,058 (87%) 11,340 (99%)
Medium score 3,691 (32%) 4,039 (35%) 764 (7%) 3,970 (35%) 196 (2%) 631 (5%) 67 (1%)
High score 719 (6%) 7,459 (65%) 0 (0%) 5,095 (44%) 54 (0%) 812 (7%) 94 (1%)
Maximum score (100) 249 (2%) 0 (0%) 0 (0%) 3,707 (32%) 28 (0%) 236 (2%) 75 (1%)
Total 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%) 11,501 (100%)
Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

Table B

Average Physical Risk Scores For Top 5 MSA Under Baseline Scenario And Low (RCP 2.6) And High (RCP 8.5) Stress Scenarios In 2050
Natural hazard MSA 2020 (average score baseline) 2050 (average score low stress/RCP 2.6) 2050 (average score high stress/RCP 8.5)
Wildfire New York-Newark-Jersey City, NY-NJ-PA 7 7 9
Los Angeles-Long Beach-Anaheim, CA 54 54 54
Chicago-Naperville-Elgin, IL-IN-WI 15 15 18
Dallas-Fort Worth-Arlington, TX 7 7 7
Houston-The Woodlands-Sugar Land, TX 14 12 13
Cold wave New York-Newark-Jersey City, NY-NJ-PA 72 58 32
Los Angeles-Long Beach-Anaheim, CA 65 46 35
Chicago-Naperville-Elgin, IL-IN-WI 81 69 44
Dallas-Fort Worth-Arlington, TX 79 62 48
Houston-The Woodlands-Sugar Land, TX 72 53 42
Heatwave New York-Newark-Jersey City, NY-NJ-PA 8 11 19
Los Angeles-Long Beach-Anaheim, CA 6 9 15
Chicago-Naperville-Elgin, IL-IN-WI 5 7 14
Dallas-Fort Worth-Arlington, TX 8 11 17
Houston-The Woodlands-Sugar Land, TX 13 18 25
Water stress New York-Newark-Jersey City, NY-NJ-PA 83 87 77
Los Angeles-Long Beach-Anaheim, CA 95 96 94
Chicago-Naperville-Elgin, IL-IN-WI 92 95 94
Dallas-Fort Worth-Arlington, TX 48 56 43
Houston-The Woodlands-Sugar Land, TX 67 71 46
River flood New York-Newark-Jersey City, NY-NJ-PA 2 3 3
Los Angeles-Long Beach-Anaheim, CA 2 1 1
Chicago-Naperville-Elgin, IL-IN-WI 1 1 1
Dallas-Fort Worth-Arlington, TX 4 2 1
Houston-The Woodlands-Sugar Land, TX 9 6 6
Coastal flood New York-Newark-Jersey City, NY-NJ-PA 1 1 1
Los Angeles-Long Beach-Anaheim, CA 1 1 1
Chicago-Naperville-Elgin, IL-IN-WI 1 1 1
Dallas-Fort Worth-Arlington, TX 1 1 1
Houston-The Woodlands-Sugar Land, TX 1 1 1
Hurricane New York-Newark-Jersey City, NY-NJ-PA 14
Los Angeles-Long Beach-Anaheim, CA 1
Chicago-Naperville-Elgin, IL-IN-WI 1
Dallas-Fort Worth-Arlington, TX 1
Houston-The Woodlands-Sugar Land, TX 65
Sea level rise New York-Newark-Jersey City, NY-NJ-PA 2 2 2
Los Angeles-Long Beach-Anaheim, CA 1 1 1
Chicago-Naperville-Elgin, IL-IN-WI 1 1 1
Dallas-Fort Worth-Arlington, TX 1 1 1
Houston-The Woodlands-Sugar Land, TX 1 1 1
Sources: S&P Global Ratings, Trucost. Data as of Nov. 11, 2020.

Digital Design: Tom Lowenstein

Editor: Rose Marie Burke

Related Research And Criteria

S&P Global Ratings research
Other research

This report does not constitute a rating action.

Primary Credit Analysts:Paul Munday, London + 44 (20) 71760511;
paul.munday@spglobal.com
Alexandre Hanoun, New York + 1 (212) 438 8615;
alexandre.hanoun@spglobal.com
Secondary Contacts:Matthew S Mitchell, CFA, Paris +33 (0)6 17 23 72 88;
matthew.mitchell@spglobal.com
Dennis P Sugrue, London + 44 20 7176 7056;
dennis.sugrue@spglobal.com
Peter Kernan, London + 44 20 7176 3618;
peter.kernan@spglobal.com
Michael Wilkins, London + 44 20 7176 3528;
mike.wilkins@spglobal.com
Catherine Baddeley, London;
catherine.baddeley@spglobal.com

No content (including ratings, credit-related analyses and data, valuations, model, software or other application or output therefrom) or any part thereof (Content) may be modified, reverse engineered, reproduced or distributed in any form by any means, or stored in a database or retrieval system, without the prior written permission of Standard & Poor’s Financial Services LLC or its affiliates (collectively, S&P). The Content shall not be used for any unlawful or unauthorized purposes. S&P and any third-party providers, as well as their directors, officers, shareholders, employees or agents (collectively S&P Parties) do not guarantee the accuracy, completeness, timeliness or availability of the Content. S&P Parties are not responsible for any errors or omissions (negligent or otherwise), regardless of the cause, for the results obtained from the use of the Content, or for the security or maintenance of any data input by the user. The Content is provided on an “as is” basis. S&P PARTIES DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE OR USE, FREEDOM FROM BUGS, SOFTWARE ERRORS OR DEFECTS, THAT THE CONTENT’S FUNCTIONING WILL BE UNINTERRUPTED OR THAT THE CONTENT WILL OPERATE WITH ANY SOFTWARE OR HARDWARE CONFIGURATION. In no event shall S&P Parties be liable to any party for any direct, indirect, incidental, exemplary, compensatory, punitive, special or consequential damages, costs, expenses, legal fees, or losses (including, without limitation, lost income or lost profits and opportunity costs or losses caused by negligence) in connection with any use of the Content even if advised of the possibility of such damages.

Credit-related and other analyses, including ratings, and statements in the Content are statements of opinion as of the date they are expressed and not statements of fact. S&P’s opinions, analyses and rating acknowledgment decisions (described below) are not recommendations to purchase, hold, or sell any securities or to make any investment decisions, and do not address the suitability of any security. S&P assumes no obligation to update the Content following publication in any form or format. The Content should not be relied on and is not a substitute for the skill, judgment and experience of the user, its management, employees, advisors and/or clients when making investment and other business decisions. S&P does not act as a fiduciary or an investment advisor except where registered as such. While S&P has obtained information from sources it believes to be reliable, S&P does not perform an audit and undertakes no duty of due diligence or independent verification of any information it receives. Rating-related publications may be published for a variety of reasons that are not necessarily dependent on action by rating committees, including, but not limited to, the publication of a periodic update on a credit rating and related analyses.

To the extent that regulatory authorities allow a rating agency to acknowledge in one jurisdiction a rating issued in another jurisdiction for certain regulatory purposes, S&P reserves the right to assign, withdraw or suspend such acknowledgment at any time and in its sole discretion. S&P Parties disclaim any duty whatsoever arising out of the assignment, withdrawal or suspension of an acknowledgment as well as any liability for any damage alleged to have been suffered on account thereof.

S&P keeps certain activities of its business units separate from each other in order to preserve the independence and objectivity of their respective activities. As a result, certain business units of S&P may have information that is not available to other S&P business units. S&P has established policies and procedures to maintain the confidentiality of certain non-public information received in connection with each analytical process.

S&P may receive compensation for its ratings and certain analyses, normally from issuers or underwriters of securities or from obligors. S&P reserves the right to disseminate its opinions and analyses. S&P's public ratings and analyses are made available on its Web sites, www.standardandpoors.com (free of charge), and www.ratingsdirect.com and www.globalcreditportal.com (subscription), and may be distributed through other means, including via S&P publications and third-party redistributors. Additional information about our ratings fees is available at www.standardandpoors.com/usratingsfees.

Any Passwords/user IDs issued by S&P to users are single user-dedicated and may ONLY be used by the individual to whom they have been assigned. No sharing of passwords/user IDs and no simultaneous access via the same password/user ID is permitted. To reprint, translate, or use the data or information other than as provided herein, contact S&P Global Ratings, Client Services, 55 Water Street, New York, NY 10041; (1) 212-438-7280 or by e-mail to: research_request@spglobal.com.


Register with S&P Global Ratings

Register now to access exclusive content, events, tools, and more.

Go Back