In 2021, risk team implemented a quantitative risk-based approach to insurance renewals which resulted in reducing the cost of insurance by 60% and improving the quality of coverage across all insurance lines. This approach allowed us to save, or rather pay a fair price for insurance across the globe. Key to success was not to rely on market “best practices”, but on testing various hypotheses in the market and relying on data.
Over the last few years the team worked with dozens of brokers across major corporate insurance lines in CIS, UK, EU, US, South America and Asia, big and small, local and global. It would be safe to say we heard all the sales speeches, promises and excuses there are. The biggest takeaway? Most brokers have no idea what their competitive advantage is. Brokers use the same arguments: quazillion years of placement and claims experience, better policy wording, know the underwriters, up-to-date IT solutions, understand nature of risk and can prepare a comprehensive insurance submission (the last two are actually a lie, but more on that later). When we were buying insurance we quickly discovered none of this mattered. All top plays have history, claim to know the risks, know underwriters, wording and even IT solutions are similar. None of these are a competitive advantage, they are barely a ticket to play. In fact, a single factor made the difference and helped us shortlist brokers – data.
Ironically, most brokers said they didn’t have it (clearly a lie) or if they did, they didn’t know how to clean and use it. And even the ones that claimed to have it and to know how to use it, did the math wrong, so the conclusions were wrong. I remember when a global broker built a model for us, estimated the fair value of the policy and came back saying our existing policy was fairly priced. We kicked them out, built own model and couple of months later reduced the cost of that exact insurance policy by $3M. We used the same math, Archer is currently working on automating, sign up here to get notified when the model is available in your country https://go.archerirm.co/Insight-RiskAcademy.
I was beginning to worry how disconnected most brokers were from quantitative risk analysis and how little effort they put into upskilling their team. And let me be very clear, unless the risk profile has been calculated as per the previous article, it is impossible to determine whether insurance policy provides good or bad coverage, whether it is fairly priced or what the deductible should be. Doing the same as last year and cheapest the market offered is just not good enough.
We were looking for brokers who could help us take the guesswork out of the equation by bring us the data, even if they can’t help with the actual modelling. As of today we managed to find few brokers that were able to speak our language in EU, only in EU weirdly. There are also brokers who refused to participate in our procurement procedures, because they could not answer the basic questions that we ask when we shortlist brokers:
- What markets generally exist for our type of insurance? We need to understand the population of the markets because we always divide the markets between brokers (in rare exceptions, we launch one broker into the market).
- Who are your friendly underwriters? The broker should have established contacts with underwriters, the broker’s unique wording of the insurance contract for various types of insurance and the availability of loss statistics with deep detail are also important. We only refer brokers to their specific “insurance friends”. The answer that the broker communicates equally well with all markets is not true and does not help much.
- What “hooks” each of your underwriting friends? We are trying to figure out what emotional and practical trigger points hooks specific underwriters, and how they ultimately “price” our risk. What formulas are used or what latest news from the market they are currently interested in. For example after Beirut explosion every underwriter asked about ammonium nitrate, which made it very clear to us, underwriters didn’t understand anything about ammonium nitrate or the associated risks and yet we still had to address this point, so we made it our point to address all these trigger points, real or not, in the submission document.
- Show us the industry loss data. At this stage, we usually lost 90% of brokers invited to tender.
- How is the brokerage (percentage or fee) calculated and whether they are ready to work on our terms?
We need these questions to shortlist 2 or 3 brokers and divide the markets between them. We would never go to market with a single broker. We would also prefer local niche brokers to global players. As it turns out good brokers all had three things in common: knows specific underwriters and is friends with them; has statistics in a format that is convenient for us to model and has policy wording that is comfortable for us.
Once we shortlisted a couple of brokers we would start working on the insurance submission. We soon discovered that we had to write all submissions ourselves, not because we wanted to, brokers couldn’t. One of our fundamental rules is that brokers are strictly forbidden to talk to underwriters until we have done our homework and built a submission document. Our standard submission is 40-60 pages of analytics with quantitative risk assessment and answers to the key underwriters’ questions. I know what you are thinking, underwriters would never read that. Well, first of all shame on them, and second we saved $13M last year by doing it our way.
I will write a separate article on submissions, road shows and working with underwriters.