Pricing risk may be helped by technology but essentially comes down to long years of intuition and experience, as addactis chief executive Pascal Mignery explains.
How do you explain the growing interest in data science and machine learning, and how does your solution support these new methods?
Our pricing solution is very comprehensive and provides all functionality in one consistent user interface to ensure data quality and improve overall usability. Data science and machine learning add another perspective on pricing and the data we use to calculate tariffs. It lets you find new criteria and factors to take into account when discovering better price points.
We want clients to benefit from both generalised linear models (GLM) and new data science formulas. However, data science is not meant to be a substitute for existing GLM methods, but rather to complement and challenge traditional GLM pricing and optimise it.
When data is drawn in from many different sources, how do insurers cope with data management?
The first problem to deal with in pricing is data. You have to get clean data to get a proper tariff. An expert in pricing is more an expert in data than GLM.
We can handle any type of data thanks to integrated modules. This allows us to connect to any kind of database, flat files or Excel files. Our solution lets users transform, test, analyse and segment data before using it for calculation. The export function is similarly powerful and versatile to ensure that the data can be used for additional scenarios.
Where do you see regulation going?
Regulation is improving in nearly every country and we really want to support actuaries in their daily work and therefore insurance companies to improve their business processes and ultimately their bottom line.
In Europe, regulation is at a somewhat mature point, albeit there will always be more specific things that a regulator wants to double-check. Governments all over the world have understood that healthy insurance companies are essential for a well-run economy and for the build up of customer trust. Therefore, we will see more "action" in other parts of the world and oftentimes they will take the European legislation as a blueprint for their own.
One way we help our customers with regulation is by implementing a full audit trail that allows for access to documentary evidence of the sequence of activities.
Asian insurers' most important consideration is with IFRS 17 right now. There remain a number of outstanding issues on software to deal with this. To help clients move forward we have set up a new hub in Singapore and will soon do the same in Africa.
Which particular issues on IFRS 17 are most challenging right now?
One of the biggest issues on IFRS 17 is reserving. Even if it is a few years away, it is important to get to grips with it now. IFRS 17 is a big driver for our software sales. We did our reserving methods according to the requirements of the Big 4 accounting firms. That's why we can give clients the solution to calculate reserves according to IFRS 17.
How easy is it for actuaries to grasp the user interface of your pricing software? How long does it take to train someone to use the technology?
Our software is built by actuaries, for actuaries and contains both the latest technical advances due to years of massive investments in research and development, and the knowledge and expertise of more than 20 years of experience in property and casualty and health insurance pricing. It is used by professionals who already understand the underlying methods. That said, user-friendliness is very important in our development. Pricing is not easy to do — it is a question of long experience, expertise and intuition.
We offer a free trial version which is used by almost all our prospects and customers. For interested parties there is nothing as useful as seeing how a solution works with their own data in their trusted environment.
Nonetheless, the software is really an all-in-one expert tool covering all pricing issues and often requires help from our team to start with efficiently. That is why we are happy to provide specialised, customised training with clients' own data, so they experience how useful the software is in comparison to existing solutions first hand. Training is a huge thing for us. Using our guided and structured process it is possible to use the classical pricing methods in a couple of days.