Redefining Insurance Pricing for the Digital Age

InsTech recently sat down with Alex Gurr, Director of Business Development at Optalitix, to discuss advances in risk pricing, the challenges facing insurers and what the company can do to help.

‍Understanding the complexities

The way in which risk pricing is being transformed is one of the biggest challenges facing iins in the Digital Age. There is a whole new level of scrutiny of the underlying risk data and how it is used for pricing.

“One of the main drivers of pricing transformation in the market is the need for a more systematic approach which drives proper pricing techniques,” explained Alex. Insurers currently utilise several different models to manage their pricing. While many models may be found in Excel, there are also models coded in Python, R and other statistical and pricing tools. 

“It’s a complicated situation. Ultimately, bringing the pricing data together and getting that data to work cohesively with the variety of pricing models used is difficult without a system that caters for them.”

‍Recognising the limitations

For example, the ease of working with Python is perceived to become a popular coding language for new pricing transformation projects among insurers. While Python has its use for pricing, Alex has seen the industry deepen its embrace of Python – driven by actuarial professionals, rather than underwriters – with mixed results. 

“There have been some pure Python implementations in the market that have resulted in a transformation process that’s very slow and painfully expensive. It’s led to issues arising from bugs and budgets being blown to fix them.”

“With all the benefits Python brings in terms of ease of use and statistical functionality, it’s not an efficient language to code in. There are several other languages out there which are just as good if not better, that an experienced developer would choose to use.” 

Alex believes that one of the biggest limitations of Python lies in the time it takes to implement pure Python models. In addition, even after these Python models are built, there are issues with the adoption of new “black box” pricing models for underwriters. With many companies looking at multi-year transformation plans using Python, he wonders whether this is appropriate while the industry is accelerating at such a “massive pace”. 

Supercharging spreadsheets

Spreadsheet models, such as Excel, remain popular due to their ease of use. Alex explains what is driving this. 

Alex said “The thing about spreadsheets is that they are very easy to use and understand. It doesn’t matter who you are – actuary, underwriter or insurance auditor – it’s easy for you to access the formulae behind the pricing calculations. This makes these models easier to use, maintain and find pricing issues earlier. There’s no quicker way to build a pricing model and roll it out than with a spreadsheet. Underwriters still use spreadsheets like scratchpads – even when dealing with incredibly complex and large risks.” 

There are limitations, however, when insurers reach the boundaries of what’s possible, such as running simulations, according to Alex. 

“Python and R can be incredibly helpful in these situations. They speed up processes and that’s why we utilise them in our pricing engine. While we recognise the power of spreadsheets, sometimes statistical work and simulations such as stochastic frequency distributions which require large numbers of repeat calculations should be coded separately because it’s quicker from a systems and processes perspective.” 

 Four years ago, Optalitix released the Optalitix Models pricing engine. It is hailed by the developers as the “most flexible pricing engine on the market” for its use of any coding language in a single platform including Excel, Python, R, SQL, GLMs and any other API-enabled models. 

Alongside that, the company offers its front-end solution, Optalitix Quote, an operational underwriting system for pricing risk and creating more efficient underwriting workflow.

‍Combining efforts

Alex advocates that the successful use of this mixed approach “depends entirely on the type of risk and the scenario”.  While “no two insurers look the same”, he sees this approach most often utilised within the London Market because “spreadsheets don’t have the power to run some simulations or modelling”.

“I think the key consideration here is having the expertise to draw data out of the pricing processes and utilise it to improve quote insight and accuracy, as well as to optimise risk portfolios.”

Supporting expertise

As important a role as technology plays in improving abilities and processes, Alex believesreal-world underwriting expertise is critical. “You can have the best models in the world and the best AI in the world, but nothing compares to 30-plus years of underwriting experience. Over the years, underwriters havedealt – first-hand – with unusual situations that models find difficult to account for, such as hurricanes, recessions and Directors & Officers liability insurance issues.”

“Expertise likethis is incredibly powerful when you’re taking risks onto your balance sheet. That’s why our pricing engine makes it easy for users to adjust both pricingengines or underwriting workflow. It generates real value for insurers at the end of the day by pricing risk more efficiently.”

Driving efficiency

This human intervention is particularly important among insurers that do embrace AI as part of their pricing strategy and this makes pricing engines an attractive option for them according to Alex. 

“We recommend co-pilot decision-making because it enhances the accuracy and effectiveness of business decisions. By combining the expertise of human decision-makers with the power of technology (AI), co-pilot decision-making ensures more informed, data-driven outcomes and provides valuable insight to the underwriter. This approach leverages the strengths of both human intuition and technological precision, leading to improved decision quality, reduced risk, and greater efficiency in operations.”

“Ultimately,you want that submission to be priced on the fly by an underwriter withoutloads of manual data work. Within a pricing engine equipped with AI,underwriters can review and understand the price and risk offered, as opposedto starting from raw data. The AI can then handle more of the manual,data-driven tasks.”

“A linked-upsystem like this generates a massive amount of efficiency and productivity and we’re seeing our clients grow substantially on the back of that.” 

Increasing ease of use

With the human element playing such a big role in ensuring pricing success, Optalitix recognises the importance of ease of adoption.

Alex said, “Time and time and again, we’ve seen the difficulties that arise from rolling out something fundamentally different from what they’re used to. If it’s hard for people to grasp or has a steep learning curve, you can end up with a very bad experience.”

Alongside ensuring its pricing engine offers a good user interface (UI) and user experience (UX). Optalitix aims to make adoption as easy as possible for underwriters, who can be “resistant” to change. 

This combination allows Optalitix and its pricing engine to overcome the issues of a “black box” pricing approach. This enables companies to be transparent and not hide any of the details contributing to premium changes and financial issues.

“With much larger risks, we tend to see underwriters trying to understand how a premium has been generated based on their inputs. Making this information more easily available allows them to answer broker queries quickly, such as why a premium is high or how the price cannot be driven down further due to the technical premium.”

Learn more about Optalitix

To learn more about Optalitix, you can visit the website or reach out to Alex and his team directly.  “We’re very approachable, so if you see us at events or in the City, particularly in Lloyds, say hello. We would love to know more about your goals.

‍