Weekly Preliminary OS Climate Risk Review (4/8/21)

By April 15, 2021 No Comments

This week we had two Delaware issuers pop up on the presale market, both catching our attention for their high risk, each taking on a different lens to discuss the risks posed by climate change. Let us first give credit to Delaware State for their exemplary disclosure. The State issuer acknowledges its geography places it at risk from the physical impacts of climate change, and that its economy and public safety is dependent on its ability to adapt to current and future climate change. And it’s true. Delaware is the lowest lying state, and with 381 miles of shoreline, approximately 17% of the State’s landmass is in the 100-year floodplain (source). Delaware ranks #1 nationally for property value-at-risk from coastal flooding (beating out even Florida), and #2 nationally for coastal flooding induced GDP impairment. The State does a thorough job of discussing its economic vulnerabilities to climate change, as well as climate adaptation, resiliency planning, and greenhouse gas mitigation.

But, that’s not why we’re here…the State’s disclosure exists in stark contrast to the subject of this week’s non-disclosure:

Delaware Municipal Electric Corporation (DEMEC), Electric Revenue Refunding Bonds, Series 2021, Beasley Power Station Project

The bonds are dated 2021-2037, and secured by a pledge of revenue received via Power Sales Contracts with eight participating municipalities scattered throughout the State of Delaware. The purpose of the issuance is to finance the development of the Beasley Power Station, a natural gas combustion electric generating facility located in Smyrna, DE.

The key peril is flooding for each of the 8 DEMEC municipal participants

DEMEC takes a different, less ambitious stance than the State when addressing climate risk in its POS. The utility focuses mostly on the risks it faces from future regulations aimed at addressing emissions from fossil fuel power plants (better known as “transition risk”), noting the “while it’s impossible to predict if or when federal or state lawmakers will enact legislation to combat climate change…the impact of future climate change legislation on DEMEC’s operations…may be material”. In the subsequent paragraph, the utility goes even less out of its way to address the physical risks posed by climate change noting that its difficult to predict with any degree of certainty the magnitude of such impact” due to increased Atlantic hurricane activity and severe weather. At least, according to DEMEC, its physical risk isn’t “impossible” to predict like its transition risk is… 

Let’s give them an F for effort, especially when compared against the State’s disclosure. Although the Delaware Municipal Electric Corporation has a modest overall risQ Score of 2.3, the utility has a high Flood risQ Score of 3.9, driven primarily by inland and coastal flooding. risQ projects that by the year 2037, these flood hazards pose a cumulative property VaR of 21% and an impairment to 56% of the GDP within the utility service area. Indeed, DEMEC’s municipal participants — notably Milford, Newark, New Castle and Seaford — have experienced significant climate events in recent years, most recently Hurricane Isaias in August of 2020.

Table showing changes in annual hurricane probability by County. Probability represents likelihood of an event at least meeting that intensity threshold — for example, Cat 1 probability of 0.4 means that there’s a 0.4% probability of a Cat 1 or greater hurricane event striking the County in that year. Note how the probabilities increase between years 2021 and 2037 due to climate change’s influence on variables that drive storms, such as temperature and humidity. The probabilities increase even more so under RCP 8.5 — the greenhouse gas concentration scenario wherein emissions continue to rise throughout the 21st century.

Hurricane Isaias ripped through the Caribbean and up the Atlantic Coast last season, causing 100,000 Delawareans to lose power (even after being downgraded to a Tropical Storm, source). When all was said and done, $20 million in damage was sustained throughout the State, which for a state such as Delaware is a little more material. However, this price tag doesn’t include the economic losses due to business interruption and GDP impairment, both of which have an outsized impact on the utility sector given that the loss of business activity flows directly into utility revenue streams.

DEMEC’s business is heavily reliant on the kinds of industrial users whose operations are disrupted during a major weather event like Isaias. The POS mentions that in 2020, approximately 43% of DEMEC’s revenues were derived from industrial users and 20% from commercial users, with only 37% of revenues derived from its residential customer base (POS pg. 29).

We’re able to obtain an even richer view of DEMEC’s revenue base by digging into risQ’s workplace demographics data. There are approximately 61,000 jobs throughout the entire DEMEC service area — roughly ~6,000 in the Accommodation and Food Service sector (a proxy for tourism; 98th percentile, statewide), ~4,500 in Transportation and Warehousing (99th percentile, statewide), and almost 7,000 in the Manufacturing sector (98th percentile, statewide). In other words, the economies of DEMEC’s municipal participants are heavily reliant on tourism, transportation and manufacturing industries — the same industries that are most severely impacted by business-interrupting weather events, and the regulatory impacts of policy aimed at addressing greenhouse gas emissions. 

Each Project Participant’s share of the debt service (POS pg. 32)

Juxtaposing the physical vs. transition risk within each of the participating munis provides an additional dimension of risk analysis nuance. For example, take DEMEC’s largest servicer of debt: the City of Newark.

The City is responsible for one-third of DEMEC’s total administrative charges, as noted in the table above. While Newark has a modest Physical risQ Score of 1.0, only ranking it in the 36th percentile for GDP impairment and 32nd for property VaR statewide, the City’s physical risk only tells half of the climate risk story. Newark’s most valuable economic asset is the University of Delaware (UD), and of the ~21,000 workers in Newark, about one-third are aged 29 or younger (98th percentile nationally). Given Newark’s younger, more mobile demographic, the City has a unique exposure to the risk of population outflows. While renewable energy is becoming increasingly inexpensive to produce, transitioning from fossil fuels to renewables is nonetheless a costly endeavor. Increases in the cost of living (one of the adverse socioeconomic symptoms of renewable energy transition, given that the cost of transition investment is covered by increases in customer fees) will only exacerbate the City’s risk of migratory outflows. And, DEMEC’s #1 customer, the University of Delaware, has a relatively heavy renewable transition lift ahead of it. According to the Association for the Advancement of Sustainability in Higher Education, as of 2020 UD has lagged behind on greenhouse gas emissions mitigation, as well as clean and renewable energy and water use, ranking it in the bottom 20% of campuses nationally in terms of sustainability (source). It doesn’t help that the cost of living in Newark is already unsustainably high — 26% of Newark’s population fall below the poverty line (98th percentile, statewide), and 51% (!!) of the population are considered stressed renters (96th percentile, nationally). 

It places the utility in a challenging predicament when a) the City of Newark (largest DEMEC participant by 2x) and b) the University of Delaware (largest energy consumer in Newark by >5x) both appear to have an outsized yet undisclosed impact on DEMEC’s long-term credit outlook. We totally get that these are abstract risks that are difficult to quantify, let alone properly convey in a financial disclosure…but, maybe DEMEC could at least try? Sometimes data science isn’t rocket science.

The University of Delaware is DEMEC’s largest customer, responsibly for 11.7% of revenue (POS pg. 31)

To their credit, DEMEC isn’t standing flat-footed and in the last 10 years has invested in more than 26MW of solar generation and 69MW of wind generation. DEMEC is funding its Municipal Green Energy Program through customer charges, and as of 2020, 18% of the Corporation’s power supply originated from renewable sources (POS pg. 35), putting it on target for the State-mandated goal of 28% by 2030 (source). Given the already high rates of poverty and stressed renters in DEMEC’s service region, one should be wary of how economically feasible it is for DEMEC to achieve State-mandated renewable energy goals. DEMEC is obviously aware of this incoming risk — “effects of compliance with rapidly changing environmental [requirements]….could have an adverse effect on the financial condition of DEMEC…and DEMEC’s ability to make payments on the Series 2021 bonds. (POS pg. 33) However, this sentence tells investors little, and leaves it to the people buying their debt to figure out what the actual transition risk might be. Thanks…?

Long on physical risk, very long on regulatory uncertainty, and regrettably short on disclosure. DEMEC’s Series 2019 2044 paper is currently priced above par value, at ~114.

Honorable Mentions

Hernando County Water & Sewer, FL (risQ Score 3.2; Flood risQ Score 3.7)

While only around the 50th percentile in terms of over risQ for counties in Florida, that’s still robust. As we highlighted for Miami Dade’s POS last week, there are a ton of considerations for these water utility issuers in Florida. Hernando is not as extreme, but is still obligated to address more severe storm water events, salt water incursion that will increase the cost of water and the energy footprint required to produce it, and the requisite damage to infrastructure from hurricanes. The Hernando County POS addresses the latter using cursory, copy-paste language down on page 52, but falls well short of Miami Dade’s POS last week (which had shortcomings of its own). No discussion of the other issues or active programs to prepare for such events. We’ve got maturities stretching out to 2034 in Series 2021A and even further to 2037 in 2021B and that allows for even further growth in climate risQ and potential salt water incursion to the system. Just for hurricane risQ alone, in 2036 the Property VaR will be just over 60% higher than is it currently, and that’s assuming no increase in underlying property values over that time. Most of that property is not insured for flooding, which makes it all the more important that the local water and sewer utility stays in front of the stormwater that’s coming. The lives, property and financial wellbeing of their customers depends on it.

Palm Beach County, FL (risQ Score 4.4; Hurricane risQ Score 3.3; Flood risQ Score 4.1)

In contrast to Hernando County, much more specific and action-oriented discussion of climate risQ on pages 37-38 with call-outs of specific programs described elsewhere. This is more akin to a Miami Dade County quality of disclosure, which is not surprising as the two counties (along with Broward and Monroe) have joint forces to a degree. Of particular note – and a lesson that issuers in Florida, Louisiana and elsewhere could learn from – is there is specific language and thought on climate risk for unincorporated and western parts of the county, and not just on the coastal area. After all, inland flood and hurricane precipitation risQ combine to represent ¾ of the overall climate risQ to the county. The best coastal defenses in the world won’t help with that. This POS suggests Palm Beach are actually aware of that compared to far too many other coastal issuers’ tunnel vision focus on sea level rise.

Salina, KS (risQ Score 1.9; Flood risQ Score 4.0)

We see you, Kansas. The potential for funding flood control projects is mentioned in a section of “Special Assessments” but there are no actual specifics of projects completed or being considered. No other mention of flood risQ or any other reference to climate risQ for that matter in the POS. That’s a little bit of a problem given there is a history of flooding in the city and wider area. No evidence of flood risQ mitigation from other sources either. Neither Saline County or the City of Salina show up in the NFIP’s Community Rating Survey (CRS), so FEMA isn’t getting any adaptation or mitigation work done there, and the residents aren’t getting discounted flood insurance as a result of the local jurisdictions being in that program. Not surprisingly, the vast majority of property in the area is not insured for flood. Note that the county overall has been losing population as has the city, and property value growth is in the lowest 15% of counties nationally. These are both signs consistent with areas with higher inland flood risQ.

Birch Run Area School District, MI (risQ Score 1.8; Flood risQ Score 3.9)

Recent history is littered with flood events in and around Birch Run, bit no mention of “flood”, not one, in the POS. Notably, property buyouts of serially flooded properties have begun, which retires those properties from the ad valorem tax base. Like Salina above, there is no evidence of participation in the NFIP CRS by towns in the district, nor by Saginaw and Genesee counties which the school district spans. Also like Salina, population loss is evident – 5% over the last 10 years – and the flood insurance uptake rate is dismally low leaving residents who stay financially vulnerable to events that are only going to get worse and more frequent.

Leave a Reply