By Paul Waters, Head of Digital Wealth and Ross Bagley, Associate Consultant at Hymans Robertson
From with-profits bonus setting processes for the legacy products of small mutuals, where reliance on licensed software and regular manual processes is a key challenge; to the massive internal capital models of the largest insurers, which generate challenges around documentation and version control.
This is a product of numerous factors, chief of which is a general lack of software engineering and systems expertise within actuarial teams, and a lack of available resources outside of BAU activity which contributes to operational friction preventing any meaningful overhauls to existing processes.
The cost of inefficient manual processes is high. It wastes valuable human capital that could be used to gain insights and make business decisions. Manual processes expose insurers to the risk of incorrect results, which can have severe consequences.
Mispricing products, underestimating capital requirements, and miscalculating policyholder pay-outs are just some examples. Updating these legacy systems and processes is necessary to streamline operations, reduce risk and create cost-effective solutions.
We have seen some uptick in momentum for digital transformation in response to these challenges, for example, the use of R and Python in actuarial work to automate data cleansing and calculations for some of the cumbersome processes that exist within insurers.
For example, consolidating a combination of disparate calculation spreadsheets which requires a familiar user (such as an actuary) to carefully execute actions laid out in a process note into a single unified model which, at a single click, produces all of the relevant outputs programmatically. A transformation like this can result in a reduction in resources and time required to complete such processes, whilst providing clear audibility and reducing the risk of error. We’d primarily attribute this type of work to the general trend of increasing junior actuarial talent with previous exposure to digital tools (like R and Python) from their time in education.
There is also growing recognition of over-reliance on licensed proprietary software, which can be expensive and difficult to troubleshoot for users. Management teams of insurers have in some cases already decided to strategically shift away from these solutions to open-source alternatives, as these continue to demonstrate their capability and flexibility to meet the needs of actuarial modelling teams.
For example, combining R (or Python) with R Shiny (or Jupyter Notebooks) enables users to apply user-friendly front ends to models and their outputs which can be operated in any modern web browser and makes both operation and communication easy.
All of these are good, and important, first steps towards digital transformation and the benefits can be felt immediately. Reductions to operational risk, improvements to efficiency, and the ability to reclaim the previously wasted specialised actuarial resource, that can now be leveraged in far more productive ways, are some of the most significant benefits.
Despite all of this, the industry remains far away from the cutting edge, where digital tools are used to centralise and streamline all actuarial models and their documentation, to scale calculations on demand, and to augment decision-making discussions with creative and flexible data visualisations.
By continuing on the path to digital transformation, insurers seek to benefit from the immediate efficiencies that result, but also the long-term effects that come from unlocking the potential of their actuarial teams to produce real insights and value.
|
