While there were issuers with higher overall climate risk metrics this week, a cohort of California school districts with very different wildfire risks also published their Preliminary Official Statements, and those exemplified very different disclosure of those risks. This also provided a good opportunity to look at wildfire metrics more closely, given the inferences that can be drawn from probabilistic risk versus tail risk metrics.
With all that said, we’re going to take the deepest dive into Mariposa County Unified School District, CA, the Preliminary OS for which can be found here.
While “only” at the 85th percentile for property value at risk for school districts nationally, the wildfire risk numbers are staggering: 99.5th percentile nationally and 96th percentile for California school districts only. A 100 year wildfire event has a property VaR of 31%, while a 500 year wildfire bumps that up to 62%. Whatever way you slice this, these are large, Paradise-exceeding metrics.
First things first, lets take a look at the language in the OS. In this case, we’ll give you verbatim from page 19 of the document (where emphasis is ours):
“In addition, wildfires have occurred in recent years in different regions of the State, and related flooding and mudslides have also occurred. The most destructive of the recent wildfires, which have burned thousands of acres and destroyed thousands of homes and structures, have originated in wildlands adjacent to urban areas. Although the recent natural disasters do not include territory within the District’s boundaries, the District cannot predict or make any representations regarding the effects that wildfires, flooding, mudslides or any other natural disasters and related conditions have or may have on the value of taxable property within the District, or to what extent the effects said disasters might have had on economic activity in the District or throughout the State.” Rather than being a non-disclosure of risk, this is just outright misleading by any standard. To it, and for the record, the 2017 Detwiler Fire burned more than 81,000 acres and 131 structures and 2018’s Ferguson Fire tallied a 97,000 acre burn, killed two firefighters and cost nearly $300 million in damages and suppression efforts. Seems to run against representations in the OS, no?
As is typical of CA issuers, there is standard boilerplate language (in this case on page A-22) of a handful of statewide programs focused on wildfire, but no indication as to the specificity to Mariposa when there is reason to do so. Contrast this with two other issuers with POS documents on this week’s slate:
Sonoma Valley USD: Negligible 100 year wildfire risk, 12.7% property VaR for a 500 year wildfire, < 10% of Mariposa County USD’s overall wildfire risk, and on page 12 of the OS: “In October 2017, wildfires burned approximately 110,720 acres of land in the County and the County of Napa according to the California Department of Forestry and Fire Protection. A few of the District’s outdoor facilities such as a garden and stage were damaged by the fire; within the territory of the District, there was damage to structures, but it did not have a material effect on assessed valuation of property within the District. In October 2019, the Kincade Fire burned approximately 77,758 acres of land in the County and destroyed more than 374 structures (residences, commercial and other) and damaged more than 60 structures (residences, commercial and other) according to the California Department of Forestry and Fire Protection. Within the District, no property was damaged or destroyed by the Kincade Fire.” At least you know the reality in this case.
Oxnard ESD: Negligible 100 year wildfire risk, o.o4% property VaR for a 500 year wildfire, < 1% of Mariposa County USD’s overall wildfire risk, and on page 20 of the OS: “The most destructive of the recent wildfires, which have burned thousands of acres and destroyed thousands of homes and structures, have originated in wildlands adjacent to urban areas…” “Although fires have occurred in areas adjacent to the District, recent major wildfires have not occurred within District boundaries.” and in the next paragraph “The District cannot predict or make any representations regarding the effects that wildfires, other type of natural or manmade disasters and related conditions or economic conditions have or may have on the value of taxable property within the District, or to what extent the effects said disasters might have on economic activity in the District or throughout the State.” See anything familiar in here, setting aside the fact that then language around state-level wildfire programs follows on page A-25.
Mariposa County USD plainly misreported or failed to recognize recent events and has also managed to downplay forward-facing wildfire risk to the same level of other issuers with orders of magnitude less risk. To tie this to assets relevant to the muni universe, limited to just counties that have had any insured wildfire losses over the last 36 years, Mariposa County is at the 98th percentile on that metric. Not only does this continue to fly in the face of the POS, but with (wildfire) insurance gaps growing, wildfire risk increasing with climate change and a series of bonds extending to 2045, there is a lot to digest in terms of future risk to the tax base of Mariposa County. One could argue shortcomings across E, S and G metrics for Mariposa County USD just in the context of this one POS.
Feel free to drop us a note if and when you’re looking at issuers where wildfire risk may be tangible, or if you’re just unsure. We’d prefer to help clients in the flow of their decisions than offer perspective in hindsight when documents are made public.