In the early days of the blog, I wrote about why Seattle quake is the scariest and most underestimated cat risk in the US (if not the world). It is a peak zone that isn’t treated like one. If you haven’t read that piece before, I suggest you start there.
Today, I am back to sound the alarm again. Not only is a giant Seattle quake more likely in the near term than I previously wrote, there is now research that suggests it has historically triggered San Andreas quakes as well.
Yes, major earthquakes may be correlated. Where’s that in your risk model?
Before I address that, let’s do a quick review of why Seattle scares the hell out of me.
What Makes Seattle Quake Terrible
1. Strength
Seattle sits on the Cascadia subduction zone. Cascadia has a long history of 9.0 earthquakes. The only other place in the US with that potential is Alaska, which is obviously far less populated and developed.
As a reminder, 9.0 quakes cause 32X more damage than 8.0s and 1000X more than 7.0s. The Tohoku earthquake in 2011 was also a 9.0 and was one of the five strongest quakes in history.
A 9.0 in a heavily populated, first world city like Seattle would be the most damaging EQ in world history.
2. Tsunami
Cascadia quakes cause tsunamis. That may sound bad, but it gets worse. The last event, in 1700, caused a tsunami that was recorded in Japan!
So, a 9.0 Seattle quake will cause a large tsunami on the US West coast, but also impact Japan and potentially islands like Hawaii in between.
3. Bad Geology
Seattle is further cursed by bad geology. It sits in a bowl which amplifies the shaking impact. It also has places like Mercer Island which have high end homes susceptible to landslide.
Furthermore, there is a Seattle centric fault that could go even if Cascadia remains quiet. While this would likely be “only” a 7+, it’s location would actually cause greater flooding in Seattle proper.
Return Period = Past Due!
It’s one thing when your library book is past due. It’s another when the greatest quake in modern history is overdue.
There has been a ton of research done on the Cascadia fault. 46 historical events (8.0+) have been found over the last 10,000 years. That suggests about a 1 in 220. 17 of the 46 were 8.8 or higher.
If you look over the last 6000 years, it drops to under 1 in 200 (frequency increased). Over that time period, the longest without an event was 344 years (over 5000 years ago) and the second longest was 330 years (over 4000 years ago).
Want to guess how many years we are since the last 8+? 323 years!
That means if we go 22 more years without a 8+, it will be the greatest drought in 6000 years. That is what overdue means.
If you use the 10,000 year history, only 19% of the time has the fault gone longer than 323 years without an 8+. If you narrow that to the last 6000 years, it drops to 7%.
Seattle is much more likely to experience a major quake than California, New Madrid, Tokyo, or any other major quake zone in the next 5, 10, or 20 years. And it is likely to be a much stronger quake as well.
Yet, it is largely ignored by the insurance industry and never spoken of as a major risk zone.
Correlation With San Andreas
So that sounds really bad, right? But as they say on those commercials, wait there’s more!!!
There was some shocking research that came out a few years ago. It is still a subject of debate, but as best I can tell, it has not been refuted.
This research, by a leading geologist who specializes in studying Cascadia, has found there is historic overlap between Cascadia quakes and San Andreas quakes. Without getting too technical, the geological findings make clear the Cascadia quakes come first and the San Andreas ones reasonably soon after.
Exact timing is too difficult to estimate. However, the evidence suggests more than half the Cascadia quakes in the last 3000 years precipitated San Andreas quakes and that this accounts for most of the San Andreas quakes in the historical record.
In other words, Cascadia is way overdue. If Cascadia goes, there is a good chance San Andreas goes too in the not to distant future.
I’m not suggesting days away like a sci fi movie. That forecasting precision doesn’t exist. Let’s say it’s months later or even a few years. That has giant, giant ramifications for cat modeling.
Basically, San Andreas becomes the equivalent of a “live cat”, such as writing a hurricane three days offshore and hoping it lands someplace other than Florida or Texas.
It becomes uninsurable – if the industry becomes aware of this correlation. Maybe it won’t, in which case, they will take major losses from Seattle and then say a year later take a major San Francisco loss.
California Capacity Crisis
The obvious response will be for CA insurers to non-renew clients ahead of time, except…this is California. You can’t do that.
So you ask for a 500% rate increase, right? Nope, you cross your fingers the commissioner will give you 6.9%. You will probably get zero, because the commish will say we don’t believe that hocus study about correlated quakes.
On the other hand, all those residents who decided not to buy quake insurance because it’s “too expensive” will start buying it like Walmart toilet paper in 2020. Take up rates will go through the roof.
You (and the CEA) are stuck.
But the reinsurers aren’t. They will leave the primaries to hang. Then what?
There will be a crisis worse than Hurricane Andrew. Primaries will already have crushed their reinsurers with Seattle losses, likely blown out the top of their programs, and now be facing a second event with no cover.
Actually, the best thing a primary could hope for is the San Andreas quake does come immediately after. Then, they’d at least have their reinstatement coverage. Once it drifts into the next policy year, there is nothing to do but pray.
Damageability
So how big does this get? I know I’ve called this “Seattle” for shorthand, but Vancouver will be hit hard (and Canada brings in separate reinsurance clients). Portland easily could be. Japan could have tsunami losses. Then, add on the San Andreas follow through potential.
We’re well over $100B. Could it be $200B? A repeat of 1906 SF alone is likely near $100B.
If you’re a reinsurer writing CA quake, there is a good chance you’re also writing Cascadia and don’t know it. Your PML is off by magnitudes. You probably don’t have enough capital to get by without severe ratings downgrades – that’s if you can pay all the claims.
And if you want to dismiss this as Doomsday nonsense, I should offer here that I’ve rung the doomsday bell a few times over the years calling out overlooked cat events with an awfully good batting average.
I was on New Orleans a few years before Katrina. Then, I moved to NYC before Sandy. Since then, it’s been Tampa hurricane (almost happened last summer), and Seattle quake.
New Madrid’s return period is a lot longer than Seattle’s. CA quakes have a lot lower severity. Seattle is the peak risk and the clock is ticking. Consider yourself warned.
Move While You Can?
Seattle is the last place in the country I would live. I don’t mind visiting, but I wouldn’t want to be there permanently.
If you happen to live in the Pacific Northwest, you are accepting a non-diversifiable risk. The cumulative probability of something really terrible happening during your lifetime is exceedingly high.
Maybe you’ll be one of the lucky ones who only suffers minor damage, but how do you know? At least you can evacuate before an oncoming hurricane. That option doesn’t exist with quake (well, unless you live on the San Andreas and sell your home immediately after Cascadia).
I’m not trying to scare people, but residents should be aware they’re effectively living on top of a ticking time bomb and MacGyver isn’t going to show up to defuse it.
If you live west of I-5, unless you’re on high ground, you’re potentially in the tsunami zone. Know where there is a safe elevated area nearby. Downtown Seattle is at high risk of being underwater.
The Cascadia zone is certainly wide enough that the next quake may be far enough south that Seattle doesn’t take the brunt of it. That’s what you’re hoping for, but do you want to make that bet?
I’d rather gamble on a 7+ in California than an 8+ in Seattle. They may sound similar, but they’re not. It’s like saying $1 billion sounds close to $1 trillion.
If you feel like you have to live in the Seattle area, further inland is better and preferably rent. Let the landlord absorb the financial loss and the rebuilding stress. If you’re going to own a home on Mercer, well, be prepared to have to dredge it from the bottom of the sound.
I hope to be wrong on this one for a long time and be called Chicken Little, too dramatic, etc. but the thing is, I only have to be right once unfortunately.