California's insurer of last resort is the most expensive option on the board in multifamily property. The FAIR Plan sits at 62.8 cents rate on line over the trailing twelve months. Lloyd's, writing the same segment on the surplus lines side, sits at 20.8.

The story in the trade press this month is that the standard admitted market is returning to California. The terminal data says the return has not happened yet, and that when it arrives it is priced above what is already available in the state today.

Katie Dowson and Grace Schmidt go carrier by carrier through California multifamily property: who is writing the volume, why the gap between the residual market and the surplus lines market is not a like-for-like comparison, what the 2024 to 2026 trend actually shows, and why the Farmers filing reported by Insurance Journal on August 3rd looks more like a pilot than a re-entry.

Then the question the data raises but does not answer. If private capital never left California, what is actually scarce?

Every carrier level figure in this episode comes from the Advocate Market Terminal: average rate on line by carrier, California multifamily property, trailing twelve months. Four carriers sampled, not the full market.

Run the same analysis on your own book at https://advocate.app/?utm_source=spotify&utm_medium=podcast.

Chapters

00:00 The thesis: the return has not happened yet

00:41 Rate on line, explained

01:00 The national picture: property down, liability up

02:14 California multifamily, carrier by carrier

02:34 The FAIR Plan at 62.8 cents

03:27 Why this is not a like for like comparison

04:43 2024 to 2026: hardening, then a pullback

05:31 A ceiling is not the same thing as relief

06:17 Who is actually carrying the book

07:02 The admitted market is not in the room

07:51 The Farmers filing, and what it really is

09:22 Is the FAIR Plan working as designed?

11:20 What is scarce is not capital

12:47 Takeaways for California brokers

Every agency leader who has watched an AI rollout stall assumes they bought the wrong tool. Ryan Deeds thinks they had the wrong problem.

In this episode, Katie and Grace sit down with Ryan Deeds, Head of AI at ALKEME Insurance. Ryan has spent 25 years inside agency data, through retail brokerages, a conglomerate of more than 100 US and European agencies, before taking on AI across a 1,500 person organization. His argument is that technology stopped being the constraint a while ago, and what replaced it is whether the person on the other end feels safe enough to change how they work.

The Advocate Insurance Desk is a data-driven commercial insurance podcast. Most episodes we use the Advocate Market Terminal, our insurance intelligence platform, to show exactly what is happening inside specific markets: real carrier behavior, real premiums, real pricing by segment. This episode steps off the pricing side entirely and into the operating side, where the data gets clean enough to trust in the first place.

The core idea: if your employees do not feel protected when they try something new, no tool survives contact with them. Ryan builds around that. Audits that open with what people got right. Dashboards that talk back so a producer can dispute a number and a human answers. A one year old product at 40% adoption, which he calls a win, because he is not fighting for the last quarter of the org. And a hard line on where automation stops, drawn not by capability but by whether the buyer understands what they are buying.

We get into why he is tired of dashboards and wants prescriptive next steps instead, the activity taxonomy that showed him where the friction actually was, why he wants three sources before he trusts one number, what "human at the edge" means once agents do the assembling, and why he now judges vendors on whether they hand him an API or make him log into their interface.

Anybody can build anything now. The interesting question is what people will actually use.

Learn more about what we are building at https://advocate.app/?utm_source=spotify&utm_medium=podcast

Ryan Deeds is on LinkedIn and is happy to hear from people building in this space - > https://www.linkedin.com/in/ryancdeeds/.

Chapters

0:00 The two extremes of every AI conversation in insurance

1:08 Ryan Deeds, 25 years and four different seats

3:42 Resistance to change is really a safety problem

5:30 An Excel macro on Tuesday, in production by Friday

9:49 Data is worthless if it does not drive the next step

10:36 Why he is tired of dashboards

11:44 The activity taxonomy that found the friction

14:04 Three sources to trust one number

17:11 The renewal checklist that took a day to build

18:39 Getting producers to adopt the tool

22:30 Where leadership has to carry it

24:10 40% adoption on a one year old product

25:15 Risk OS and the problem it is trying to solve

28:50 Human at the edge, not human in the loop

30:28 Moving small accounts to the hive

33:27 Cyber to issuance, workers comp to a person

36:12 Why the build versus buy calculus changed

37:54 Proprietary data as the five year moat

39:46 Give me the API, not your UI

44:05 Close

Every vendor in commercial insurance now says they use AI. Almost none of them will tell a buyer which part of an answer was calculated and which part was generated. David Haddad, head of product engineering at Advocate Technologies and the builder of the World Insurance Model covered in Episode 21, joins Katie Dowson to explain why that distinction is the whole thing, and why the model itself is the least interesting part of any AI product.

His own job is the evidence. David went from writing code effectively all day to writing very little of it, and what replaced it is planning, specification, customer conversation, and testing. Unit tests used to tell you a thing worked or it did not. Probabilistic systems do not offer that, so the work moves into evals, harnesses, and pipelines that decide what a model is allowed to do and catch it when it is wrong. He also walks through the week a feature his team had spent real time building was made redundant overnight by a vendor release, and why the right response was to stop defending it.

The episode covers the dependency almost nobody puts on a slide. A company building on frontier models sits on a chain of counterparties it does not control, the same shape as the managing general agent chain from the last episode. Prices move, versions change, providers go down, and the tone of a generated document can shift while the customer assumes nothing changed. The Advocate Market Terminal is built on carrier pricing, premium, and compliance data across commercial real estate lines, and the standard is the same in both directions. If a number is not testable, it does not ship.

The takeaway is three questions to ask anyone demoing AI software. How do your own engineers use it, how do you benchmark the output and show the math behind it, and how much of this rests on a single model.

Learn more about what we are building: https://advocate.app/?utm_source=spotify&utm_medium=podcast

#AdvocateInsuranceDesk #AdvocateTechnologies

0:00 The model is not the product

1:22 How the engineering job changed

3:47 Why model selection is overrated

5:10 What code is no longer worth writing

8:44 Hiring for problems, not for code

10:20 When a model absorbs what you built

13:12 What stays in human hands, and who is liable

15:03 Model supply chain, borrowed from MGAs

19:38 Tone drift in the proposal generator

22:58 Accuracy versus precision, and the bullseye

25:38 How to pressure test an AI vendor

28:17 A year out: the gap widens

The name on your declarations page is probably not the company that priced your coverage, and it may not be the one holding your risk either. In a growing share of the commercial market, a managing general agent does the underwriting, a fronting carrier lends its license and its name to the paper, and reinsurers you will never be told about hold the actual dollars. You only ever meet your broker.

Katie Dowson and Grace Schmidt walk through what that chain looks like from the buyer's seat, why it grew so quickly out of the hard market, and the two ways it can go wrong. One is a renewal that comes back drastically higher or does not come back at all, for reasons that have nothing to do with your building or your loss history. The other is what happens to a claim when the coverage chain is under stress, which the 2023 Vesttoo collapse demonstrated in public.

This episode also covers the part most buyers miss. When coverage sits on a non-admitted excess and surplus lines policy, the state guarantee fund backstop generally does not apply, which means the counterparty behind your policy matters more at exactly the moment you can least afford a problem. The Advocate Market Terminal cannot look inside a program, but it can tell you whether the program is even the right place for your risk, and it surfaces AM Best financial strength ratings next to the carrier so you are not looking up an unfamiliar name yourself.

The takeaway is three questions you can answer tonight from a document you already have. Name, rating, capacity.

Learn more about what we are building: https://advocate.app/?utm_source=spotify&utm_medium=podcast

#AdvocateInsuranceDesk #AdvocateTechnologies

Chapters

00:00 Why this is an industry data episode

01:01 The renewal that looks completely normal

02:06 What an MGA actually is

03:37 Why the pen started getting handed out

04:39 MGAs are not the villain

06:10 What a fronting carrier is

06:54 Three counterparties, one declarations page

07:19 Walkthrough: a mid-sized apartment portfolio

09:15 Failure mode one: capacity disappears at renewal

10:04 Failure mode two: claims under chain stress

12:06 Surplus lines and the guarantee fund gap

12:43 What the Advocate Market Terminal shows here

14:07 AM Best ratings surfaced next to the carrier

14:42 The catch: the rating is the front, not the reinsurance

15:48 Three tells you are in a program

16:51 Three things to check on your declarations page

18:08 Name, rating, capacity

The property market is softening for the first time in years, and for once the good news is landing on the independent agent's side of the table. But if capacity is opening up for everyone, how does a two-person shop actually win against a national brokerage?

In this episode, Katie and Grace make the case that the gap was never really about size. It was about who could see the market. We walk through five concrete moves a small shop can run in a softening property market, and for each one we open the Advocate Market Terminal and show what it looks like in practice on real placed-policy data.

The five moves:

1. Read each carrier's game plan from where they've actually landed on price, not from rumor.

2. Shop strategically. Present the incumbent renewal plus two well-chosen alternatives, not fifteen.

3. Negotiate beyond price with a defensible number and the full distribution behind you.

4. Protect your revenue and turn a softening market into a retention and trust move.

5. Use technology to spend less time on market research and more time in the conversations that matter.

One honesty note we keep on the show: the softening we describe here is on the property side only. Liability is still hardening, so bring your clients the good news about the part of the book where it's actually true.

And when we say the terminal shows you something, we mean real placed commercial policy data, kept separate from broader market context.

Create a free account and pull your first comp group at

https://advocate.app/?utm_source=spotify&utm_medium=podcast

The Advocate Insurance Desk is a data-driven commercial insurance show built on the Advocate Market Terminal, our insurance intelligence platform. New episodes go inside a real market and show you actual carrier behavior, premiums, and pricing by segment.

Subscribe for more, and find us on Apple Podcasts, Spotify, or wherever you listen.

Chapters

0:00 Welcome

0:27 Why this one is a playbook, not a market outlook

1:24 The real gap was never size, it was sight

2:19 Move 1: Read each carrier's game plan

4:13 Move 2: Shop strategically, not broadly

6:02 Move 3: Negotiate beyond price

7:39 Move 4: Protect your revenue

9:14 Move 5: Leverage technology to enhance your value

10:05 The takeaway

10:41 Where to watch and listen

What if the number everyone is quoting to say the hard market is over is quietly lying to almost the entire market?In this episode, co-host Grace Schmidt is back and Katie catches her up on three headlines that moved commercial insurance while she was gone. No long market outlooks, just the stories that actually changed what we are building at Advocate, one at a time.The Advocate Insurance Desk is a data-driven commercial insurance podcast. Most episodes we use the Advocate Market Terminal, our insurance intelligence platform, to show exactly what is happening inside specific markets: real carrier behavior, real premiums, real pricing by segment. This episode is a fast catch-up across three of them.The core idea: a single market-wide average is hiding a split. After 32 straight quarters of rate increases, the industry index posted its first broad decline since 2017, down about 1.2 percent, and everyone called the hard market over. But that average blends two things moving in opposite directions. Property is easing while liability never turned, and the flat headline number describes neither side accurately. The only question that matters for a specific policy is where it sits inside its own market, and you get that from carrier-level data, not an industry average.In this conversation we cover the property and liability split and the courtroom-driven social inflation behind it, with average commercial auto verdicts climbing from roughly 3.6 million dollars in 2010 to north of 30 million in recent years. We pull up two Advocate Market Terminal reads: the US National Liability Index on the trading tab, and California multifamily liability on the pricing comps page, where the same line of coverage runs close to eight times more expensive from the cheapest quartile to the top. That is Joe Zuk's K-shaped market, organized around risk quality cohorts, playing out in real time. We then get into the data center buildout piling up more value than carriers can comfortably insure, echoing Rachel Nixon's point that capacity, not demand, is the real constraint. And we close on an industry that spent the year buying AI before realizing the hard part was always the data underneath it, which is exactly why Advocate built the World Insurance Model as a deterministic engine instead of pointing a big model at the problem.The takeaway: averages describe nobody. Property softening and liability firming are two different markets wearing one number, and the only way to price a real account is carrier-level data, structured and connected, not one more dashboard.Industry data referenced from the Council of Insurance Agents and Brokers, broker projections, and third-party estimates on data center exposure is outside the terminal. The carrier-level reads are the terminal's own.Sign up for the Advocate Market Terminal: https://advocate.app/?utm_source=youtube&utm_medium=podcastChapters0:00 Grace is back, and here is what you missed0:49 The setup: three headlines, one at a time1:14 Headline one: the 32-quarter streak just broke2:22 The terminal: US National Liability Index2:53 Why liability is firming: social inflation3:55 Pricing comps and the eight-times spread4:54 Headline two: data centers break the math6:25 Rachel Nixon and the capacity constraint7:43 Headline three: everyone bought AI8:21 The data problem sitting underneath it9:42 Why Advocate built WIM instead10:38 Recap and what to watch next#AdvocateInsuranceDesk #AdvocateTechnologies

What if the most accurate way to run an insurance compliance check is to stop the AI from thinking?In this episode, Katie sits down with David, head of product engineering at Advocate, who built the World Insurance Model (WIM) over roughly five years of R&D. WIM is the deterministic engine underneath the Advocate app, the thing that turns thousands of manual policy checks into consistent, testable results.

The Advocate Insurance Desk is a data-driven commercial insurance podcast. Most episodes we use the Advocate Market Terminal, our insurance intelligence platform, to show exactly what is happening inside specific markets: real carrier behavior, real premiums, real pricing by segment. This episode goes one layer deeper, into the engine that powers the platform itself. The core idea: frontier models on their own are not good enough for compliance work. They perform decently and then leave you to clean up the rest. Pair a model with WIM as a tool, and it offloads the reasoning to a deterministic engine that returns the same output for the same input every time. Accuracy roughly doubles while token cost stays flat, because the model stops guessing at requirements and starts asking WIM which fields actually matter.In this conversation we cover the scale of the problem (around 30 million commercial policies and 500 billion dollars in annual premium running through compliance every year), what deterministic actually means and why it matters when one missed check can cost millions, how documents flow through the platform from upload to compliance report, and the Advocate App Labs benchmarks: Sonnet moving from 26 percent of coverage gaps found on its own to 63 percent with WIM, the rule engine alone reaching about 74 percent, and a licensed human reviewer still leading at 96 percent. David also walks through the model harness that mixes engine and frontier models step by step, how hallucinations are handled with citations and a human in the loop, why review time compresses from about 90 minutes to a couple of minutes, and what WIM means for brokers looking to expand into new asset classes.The takeaway: the reading and data-pulling part of a review is already being automated. The judgment calls stay with the human, and the accuracy gap keeps closing.See WIM at work. Create a free account at ⁠https://advocate.app/?utm_source=spotify&utm_medium=podcast⁠The full benchmark study is on Advocate App Labs.Connect with David: https://www.linkedin.com/in/david-a-haddad/Chapters0:00 Why this episode goes one layer deeper1:16 Meet David, who built WIM2:24 The scale of the problem: 30 million policies3:30 What the World Insurance Model actually is4:52 Deterministic vs probabilistic, explained6:05 Dropping documents into the platform7:56 Benchmarking WIM against the frontier models9:12 How a model uses WIM as a tool12:18 The numbers: accuracy gains and cost15:05 The harness and mixing models16:20 Hallucinations and the human in the loop18:18 Will AI take the reviewer's job?19:19 Trying WIM yourself20:11 What WIM means for brokers22:16 What to watch for next

#AdvocateInsuranceDesk #AdvocateTechnologies

Florida just mailed $1 billion back to 830,000 policyholders. New York thinks it can force the same result by law. The data says the bill is aiming at the wrong line.

In this episode, Katie and Grace put New York's new commercial insurance bill (A11298) up against the Florida tort reform playbook, then test the whole thing against live transaction data from the Advocate Market Terminal.

Florida fixed its courts and the rate relief showed up on its own, all of it in personal lines. New York is betting that prior approval, forced rate disclosure, and a filing delay can do for commercial property and commercial liability what tort reform did for Florida homeowners. So we pulled a single New York multifamily archetype, pre-war construction, five stories, three miles off the coast, and looked at what is actually moving.

What the data shows:

Commercial property runs about a 2.9x spread across the middle of the market and is already correcting downward on its own

Commercial liability runs a 6.9x spread, with the average rate sitting roughly 3.4x above the median

On comparable risk, the factor analysis pulls property pricing down while pushing liability up, the same K-shaped split Joe walked through a few episodes back

The biggest driver on liability is not catastrophe or distance to coast. It is location, density, and the local litigation environment

The bill puts prior approval on the line that is already healing and adds lag to relief that is already on its way. It discloses the line that actually hurts, but disclosure and a 60-day delay cannot reach a courtroom. The lever that fits the problem, tort reform, is the one New York did not pull. And this is not law yet, it is one member's bill that most likely stalls as the session wraps.

Sign up at advocate.app and run your own asset class and market: see which lines are moving, by how much, and what is actually driving your price before your next renewal.

Subscribe for more on YouTube, Apple, Spotify, or wherever you listen.

Chapters:

0:00 The $1 billion Florida refund

1:03 Commercial's quiet crisis

1:48 New York's bet: Bill A11298

2:39 Two theories: tort reform vs regulation

4:47 Theory two: the regulatory fix

5:48 What the bill actually does

7:39 The asymmetry: only homeowners get a forced cut

8:47 Setting the control: a New York multifamily archetype

9:33 The spreads: property 2.9x vs liability 6.9x

12:41 Factor analysis: property down, liability up

14:19 Not catastrophe, location

15:18 Umbrella, excess, and Joe's K

17:04 Synthesis: regulating the line that is healing

19:33 Politics, and why the bill likely stalls

20:24 Where we land, and pulling your own market

Property and liability are moving in opposite directions. So how do you insure the asset class that everyone is suddenly chasing?This week Advocate co-founder and CEO Ashwin Agarwal pulls up a chair as co-host alongside Katie Dowson for a conversation with Rachel Nixon of IMA, who has been placing data center coverage for more than 20 years, long before it became the story everyone is chasing, and who recently helped structure a $4 billion placement.The Advocate Insurance Desk is powered by the Advocate Market Terminal, the insurance intelligence platform that shows you exactly what's driving pricing in your market. See it for yourself at advocate.app.Rachel's point is that insuring a data center is not just insuring an expensive building. The real story is uptime. Downtime that used to be measured in days is now measured in minutes, revenue is tied directly to grid reliability, and there is a gray area between property and cyber that the market still has not solved. Get those pieces wrong and the most well-funded sponsors in the world can still find themselves underinsured on the risk that actually matters.Ashwin and Rachel get into the full coverage stack from property and business interruption to cyber, liability, and construction wrap-ups, the structure behind a $4 billion placement, why risk engineering now decides who gets the best rate, and the new frontier lines opening up around SLA and parametric coverage for uptime and power. They also dig into where insurance capacity is heading, the political and environmental backlash starting to shape the risk picture, and what a data center benchmark would actually need to track. We layer in Advocate's own terminal data on the property and liability divergence playing out across the asset classes we cover today.The takeaway Rachel lands on: in five years this is its own asset class, with its own insurance to match.Want to see how property and liability are actually moving in the markets you cover? Pull your market on the Advocate Market Terminal at advocate.app. That is what it was built for.Connect with Rachel at [email protected] and on LinkedIn: https://www.linkedin.com/in/rachelstempernixon/.Chapters:00:00 A different kind of episode01:57 Meet Rachel Nixon03:00 How scale and speed rewrote the market04:31 Underwriting the hardware and depreciation06:26 The coverage stack08:35 Coverage gaps and the property-cyber gray area10:10 Business interruption and SLAs11:31 Power, redundancy, and the backup for the backup12:43 Loss history and the funding boom15:27 Systemic risk and where capacity is heading17:44 Inside a $4 billion placement22:17 The Market Terminal: property down, liability up24:50 Politics, regulation, and public backlash28:01 Designing a data center benchmark30:53 Construction volatility, wrap-ups, and temp to perm33:25 SLA insurance explained35:19 Coverage that doesn't exist yet37:14 Rachel's prediction: a new asset class

Property rates are down. So why aren't owners feeling it?

This week we hand over the desk. Advocate co-founder and CEO Ashwin Agarwal takes the host chair for a conversation with Joe Zuk, operating partner at Altamont Capital and board member at Accelerant, and one of the few people who has operated across the entire insurance stack: brokerage, MGA, reinsurance, and the capital side.

The Advocate Insurance Desk is powered by the Advocate Market Terminal, the insurance intelligence platform that shows you exactly what's driving pricing in your market. See it for yourself at advocate.app.

Joe's argument is that "AI is going to fix insurance" is the lazy version of what's actually happening. The real story is a K-shaped market. The top of the K, well-capitalized sponsors with modern assets and clean loss history, has carriers competing hard. The bottom is commoditized and increasingly automated. And the middle, which is most of the market, is getting left behind: passed over in submission queues and priced without anyone really looking at the risk.

Ashwin and Joe get into the capital cycle pushing property pricing toward the floor while liability keeps climbing, the quiet arms race in policy language as carriers use AI to carve coverage out and brokers use it to add coverage back, and the new frontier lines opening up around AI infrastructure, from residual value cover on servers to parametric structures for data center power. We also layer in Advocate's own placement data on where property and liability pricing are actually heading.

The takeaway Joe lands on: know your lane, know your data, and remember the K.

Want to see where your asset class and geography sit on the K? Pull your market on the Advocate Market Terminal at advocate.app. That is what it was built for.

Chapters:

00:00 A different kind of episode

01:34 Meet Joe Zuk

02:17 What is the K-shaped insurance market?

03:46 What's driving the K, and what it really means

06:13 Why the middle of the market gets left behind

08:09 Fixing the middle: data, benchmarking, and telling the risk story

10:41 The AI arms race in policy language

12:58 Two faces of AI: sharper underwriting vs. new tail risk

17:04 New frontier lines: data centers, RVI, and parametric cover

20:31 The capital cycle: property down, liability up

24:25 Why there's so much capital in property right now

26:09 Does this cycle rhyme with past ones?

27:59 Positioning for the K-shape: owners, brokers, carriers

31:05 Recap: know your lane, know your data, remember the K