Big Data: Big Opportunity or Big Risk for Actuaries?
The Institute and Faculty of Actuaries (UK) 2nd Asian Conference was well attended by actuaries from the region, including a number of Australians. Senior Vice President of the Actuaries Institute, Jenny Lyon, was there to hear expert insight into Data Analytics and the profession.
‘Big Data: Big Opportunity or Big Risk for Actuaries’ was the title of the final Plenary session at the 2016 IFoA Asia Conference. In addition to this, there were two concurrent sessions which addressed big data or data analytics, highlighting the level of interest in this field for actuaries globally.
The premise of the Plenary talk lay around the question: can actuaries manage 21st Century data?
The presenter, Peter Banthorpe, is the Global Head of R&D at RGA and a Fellow of the Institute and Faculty of Actuaries. He was very positive about the opportunities for actuaries while recognising it will take some time and ongoing development within the profession. While these were his final messages, they are worth highlighting at the beginning of my summary.
“Actuaries need to actively consider how to expand their skill sets to continue to grow and serve the public, employers and the profession. The business interface is where actuaries can play best, they add value by defining problems and need to consider how they can have their value recognized.”
Most importantly he noted :
“We have an individual responsibility for the future. If we don’t do it, then who will? What do we have to lose?”
Peter gave a lightning tour of some of the things which are happening in the field of data analytics. As he noted there can be five-day conferences on just one aspect of the field so he covered the surface rather than any depth. The following are some of the things he said which interested me.
We need to recognize that the world (and insurance) is going digital and what this means for us is that there is an ability for segmentation in new ways. In particular, there can be a move to delivering personalised solutions – ultimately at the individual level, and data can be used to significantly improve the customer journey.
The Internet of Things – by placing a sensor on/in a physical item information can be transmitted via the internet. For example, impending failure of a component can be identified early and replaced before it has a major impact. In insurance this is visible through the use of telematics in motor insurance, he noted that in Italy 19 of the top 20 motor insurers are using telematics.
While commenting on the fact that one set of data can be used to provide information right across every stage of the value chain, he also noted that in addition, by linking your data with other data the level of insight can be increased. For example, by linking telematics data to other contextual data such as the weather/places of interest you can tell more about whether/why a person is making the same trip daily or for a particular reason.
He asked whether your company is making the most of your data and what other data can you use to give you better information – unstructured (eg tweets), labelled and unlabeled data. As an example he provided a case study of an organisation which had identified that a challenge with investing was that humans can’t read quickly enough so they developed an algorithm and machine to extract information from Bloomberg articles and then use this to drive investment decisions and achieved exceptional results relative to the S&P particularly in the first couple of years.
As an example of the potential for disruption Peter had recently visited a group of academics who by applying mathematical modelling to some publicly available unlabeled data groups had been able to get insights on life insurance which would take years of experience to build in a more traditional model.
Peter confirmed that Excel is no longer enough for actuaries to use partly because of the greater speeds needed. He felt that R was a great tool for actuaries not just for data analytics but also for stochastic mortality projections.
In addition to the insights which can be found in the data, there are other issues to consider, in particular owning data brings new risks:
Regulations are needed around holding and using big data – it is a privilege to have data and you need to remember that it can have an impact on peoples’ lives.
Reasonable expectations of the public around the use of data should be a consideration.
Risks – as an example some work he had seen had suggested that it was 99% probable that underwriters would be replaced by a machine (actuaries 21% probability)
Rewards – predictions are that working in data science is the place to be and that opportunities are on an upward trend
There is an increasing need for collaboration and companies working in partnerships and in a multidisciplinary environment will continue to expand. For example, when considering wearables and their emergence in the insurance field, insurers will need to think about the role they want to play and probably work with partners who can manipulate the data and provide insights.
In summary, I came away excited about the potential for actuaries working in this field but very aware that, as in most areas, there will be a strong need for individual motivation, learning and development along with a flexible and adaptable approach.
CPD: Actuaries Institute Members can claim two CPD points for every hour of reading articles on Actuaries Digital.
