Actuary Vs Data Scientist
Data Science

Actuary Vs Data Scientist

Actuary Vs Data Science: Which is better? How do they differ?  Which should I choose? These are questions I get a lot from actuarial students these days.

First some background:

About 11 years ago I started introducing data science into one of my Actuarial Science courses.

My rationale was a belief that actuaries of the future would increasingly incorporate a lot more data science and thinking into their day to day jobs and actuaries would also possibly move beyond the traditional realms of insurance, investment and pensions to new areas where analytical skills were becoming increasingly important. I felt that actuarial students should, therefore, be learning more about how to deal with extremely large datasets and how to use machine learning techniques and be able to, when necessary, move beyond the application of more traditional statistical methodologies. That belief has turned out to be true as we have witnessed the explosion of data science more generally and it has also made further inroads into the actuarial profession.

I wanted my actuarial students to gain exposure to data scientists working in different fields so they could really start to think about the possibilities. I would therefore (and I still do) bring in data scientists from different industries and companies to talk about their experience. To tell stories about the various data problems they were solving, the tools they were using, the challenges they were facing and the opportunities that were out there. The idea was to get my actuarial students to think beyond mortality and insurance modelling and to open their eyes to the immense possibilities that were opening up for analytical and data savvy individuals to utilise their skills to help companies extract value from the ever increasing data sets that were becoming available. 

At the time of this introduction (2009), data science as a term wasn’t really in vogue.  Instead terms such as predictive modelling and predictive analytics were instead often used to describe what we now mainly refer to as data science.

Of course, since then, data science has exploded. Exponentially increasing data combined with enhanced computer processing power, better tools and sharing of knowledge has ushered in what is now often touted as “the world’s most sexy job.” Data maths and statistics becoming sexy! Alexandra Robbins wasn’t too far off the mark when she wrote that “The Geeks Shall inherit the World.”

If we compare a simple Google trends search we can see that it has left poor old Actuarial Science in the dust of late!

Actuary Vs Data Scientist
Actuary Vs Data Scientist: Source Google Trends

The process of speaking with and listening to many different data scientists over the years really made to think about and consider what the key differences were between data scientists and actuaries. My students wanted to know as well. In many ways, the difference is vague and grey. Actuaries are, after all, generally considered to be the very first data scientists when we began using data to make predictions in the insurance world over 200 years ago.

This infographic is my attempt at trying to capture some of the key differences between the two fields at present. I think the differences will become more blurred over time. When I posted this on LinkedIn back in 2017 it received a fair bit of interest and generated some very interesting debate, some of which I’d like to highlight and discuss further below (to be updated).

Actuary Vs Data Scientist

1: Old School Vs New Kid on the Block

Actuarial Science roots go back to the late 1600’s and the 1700’s when pioneers such as John Graunt (a London draper who showed there were predictable patterns of longevity and death in cohorts of lives), Edmond Halley (not just famed for “Halley’s comet,” he also constructed a life table for premium calculations) and James Dodson (pioneered work on long term insurance contracts) led the way. Actuarial Science has, therefore, evolved over many years and as such it has in place many global professional standards, certifying professional bodies and regulations.

Actuarial Science also has a well defined educational curriculum with globally agreed knowledge requirements facilitating mutual recognition agreements, such as that between the IFoA and various other actuarial professional bodies (e.g. the Society of Actuaries, the Actuarial Association of Europe, the Actuaries Institute (Australia), the Actuarial Society of India, the Actuarial Society of South Africa, the Canadian Institute of Actuaries, the Casualty Actuarial Society, the Institute of Actuaries of Japan and the Israel Association of Actuaries. 

Data Science, in contrast, is generally viewed as a new career and has not been “professionalised” to the same extent. 

2: Exams Vs Informal Learning

3: Insurance & Pensions Guru Vs Jack of all Trades

4: Business Risk Expert Vs Modeller

5: Statistical Modelling Vs Machine Learning

​Read More
self driving autonomous car
Data Science, Future of The Profession, Technology

What Will Autonomous Vehicles Mean for Actuaries?

“This is a guest article written by Niamh Loy. Niamh can be found on LinkedIn here.”

In 1886 the first true automobile powered by a combustion engine was created. Radio broadcasts could be enjoyed while driving by the creation of the first car stereo in 1930 and the year 2000 saw the first hybrid car powered by electric energy and petrol. In 2003 cars could parallel park themselves and the year 2014 saw the creation of Tesla autopilot technology. The innovation and technological improvements that have been seen through the evolution of vehicles is astonishing and it only continues to become more impressive. You therefore may be asking yourself ‘What will be next?’. It is the belief of many car manufacturers like Volvo as well as tech-giants like Google that the next step forward in the area of transportation is fully autonomous vehicles.

What Exactly Are Fully Autonomous Vehicles?

autonomous vehicle

Simply put they are vehicles that will be able to drive themselves with no human interference. Yes, you did read that correctly and no, this isn’t something taken from a sci-fi film set hundreds of years from now! A report from Statista claims that one in ten cars will be self-driving by 2030. Five levels of self-driving have been clarified. Level 1 should sound familiar where the vehicle is controlled by the driver with some driving assist features included in the design – think cruise control. There a steady progression through levels 1 to 5 with 5 being the ultimate goal. Under level 5 the car is fully automated and can carry out all driving functions under all conditions. This means there would be no necessity for a wheel or accelerator (although it is still likely there will be the option for driver control). It is difficult to imagine that a car could follow all the road rules and driving basics that we have all had to do numerous driving lessons and tests to be allowed to do. However advancements in technology has made it that sensors can pick up everything a human would with the key benefit of autonomous vehicles being that they can actually pick up more than a human would.

What Are the Benefits and Concerns of Autonomous Vehicles?

The most beneficial impact of self-driving vehicles is that the risk of human error when driving would be eliminated. In a world where autonomous vehicles are part of normality no one could do things like speed, text while driving or run red lights. All of these things contribute to the number of road accidents but no driver means no human error. In fact, US Department of Transportation researchers estimate that road traffic fatalities caused due to human error could be reduced by as much as 94% by using autonomous vehicles. According to the World Health Organisiation, approximately 1.35 million people die due to road accidents each year. Imagine this number sliced by as much as 90% and the amount of lives that could be saved.

As well as the striking benefit of a reduction in road deaths McKinsey research suggests that as much as 50 minutes per day could be deducted from each person’s driving time (that’s 50 more minutes to watch Netflix per day) and the issue of parking spaces will be hugely decreased, reducing the need for space to park in the US by more than 5.7 billion square meters. Say goodbye to the days of looping trying to find a parking space for half an hour or worrying that someone is going to bump into your parked car! The same report also claims the reaction time of self-driving vehicles is already a third of that of a human with engineers continuing to work to further improve this time. These vehicles can therefore pick up on potential hazards much quicker than a human and avoid even minor road accidents. With all these potential benefits associated with self-driving vehicles you might imagine there should be a huge global push for them to be rolled out as soon as possible. Unfortunately as we well know (through many a zoom meetings/quizzes and online classes) technology, no matter how advanced, can sometimes malfunction and this is where issues arise.

man in a self driving car

Autonomous vehicles cannot 100% guarantee that no road accidents will occur because there will always be the risk (no matter how small) of a technical failure. In March 2018 an Uber self-driving car collided with a 49 year old pedestrian killing her, after struggling to correctly identify her on the side of the road and therefore not stopping as she crossed. It is accidents such as these that have contributed to the public concern over the true safety of self-driving vehicles. However, although this accident is tragic and could have been avoided, research still suggests that the overall number of road traffic deaths will be decreased. After all this exact event has occurred time after time due to human error and the probability of it occurring because of humans is much higher than it occurring due a technical fault of self-driving vehicles. A lot of people exhibit over confidence bias and have more trust in their own abilities than that of someone else or in this case something else and this brings a lot of opposition to the forefront of the topic.

Insurance and Pricing Actuaries – Why Are They Involved?

Engineers, scientists and test drivers are the first that come to mind when considering who are the people most important to getting autonomous vehicles on the road. It would be easy to forget that insurers are also extremely important in the process- remember an uninsured vehicle is not allowed on the road. In particular, pricing actuaries working for these insurance companies will be extremely important and play a key role to get these vehicles on the road.

Taking a step back let’s answer a basic question that very few people know the answer to and I know I get asked on the regular- what is an Actuary? Simply put actuaries are concerned with managing financial risk. Everything we do has risk associated with it. We take out insurance policies to protect against the risk of uncertain future events that could have a high adverse impact on us whether this be life insurance, health insurance or car insurance. The premium insurance companies quote you for insurance isn’t plucked out of the sky! It is actuaries that we trust to access specific risks and come up with an informed and reasonable premium to charge for insurance coverage in the event these risks occur. Much like the weather risk is not stagnant but it is constantly changing and actuaries need to continue to adapt and respond to changes in this risk through updating their models, statistical methods and mathematical techniques to keep up to date with new innovations and changes in the world that affect the risk they access. Autonomous vehicles are one of these changes.

Fewer vehicle crashes means fewer claims to an insurance company but the company cannot simply take this claims reduction as a win and continue to price in the same way. General insurance actuaries will need to adjust their pricing to reflect the expected reduction in claims so we could expect lesser premiums. Further to this, how the policies will be priced will be entirely different. Currently, insurance companies price their personal vehicle insurance using models based on a range of factors that include driver age, gender, number of additional drivers and many more factors related to individual driver characteristics but if there is no driver these factors become irrelevant. Even the no claims bonus that all insurance companies implement in their policy pricing will have little meaning when it comes to autonomous vehicles because there will be a random nature to when malfunctions occur that result in a claim rather than an indication that there is a lack of driver skill. When considering autonomous vehicles the risk will shift from the driver to technical failures. Therefore things such as connectivity issues, equipment failure and even cyber-attacks will be extremely important. How to collect data, model and incorporate all these new risk factors will be a big challenge that pricing actuaries (who build the pricing tools you would input your details into online to get a quote) will need to overcome.

Another difficult question becomes one of liability in the event of an accident, which is extremely important in the case of insurance. If a malfunction occurs while the car is driving itself then surely it must be the fault of the vehicle manufacturer and not the owner of the vehicle? Therefore it is possible that future vehicle insurance will shift from individual policy holders to car manufacturers instead. But wait- does there will be no need to buy car insurance? Unfortunately, if this became the case then manufacturers will experience higher operating costs and therefore this insurance cost will passed through in a higher vehicle price. To bring the prices of autonomous vehicles down then more likely is that of a joint ownership between vehicle manufacturer and individual holder meaning that insurance companies and in particular pricing actuaries will have to find a way to balance individual coverage with product liability coverage on the manufacturer’s part. If this doesn’t sound like a difficult enough task as it is the question of liability and the implications for insurance companies becomes even more difficult when we consider the transition period where we have both human operated vehicles on the road along with partially and fully autonomous vehicles. It could be very difficult to argue whether human error caused a crash or a technical failure and finding a way of accessing this will need to be addressed. Insurance companies will need to be able to provide a new form of insurance for autonomous vehicles while also providing traditional vehicle insurance and both will need to be separated in terms of claims and for pricing actuaries to examine the profitability of their pricing. This could lead to double the workload!

Will Other Types of Actuaries Be Affected?

It is clear general insurance pricing actuaries will face the most drastic changes and challenges due to autonomous vehicles but other types of actuaries will see some changes too. Reserving actuaries calculate what cash an insurance company should hold in order to be able to pay out expected claims. Since general insurance companies will see a reduction in claims it is likely that their reserving actuaries will see a reduction in their necessary reserves. Incorporating past claims experience to calculate their reserves could pose a challenge for the first number of years due to a lack of data in relation to autonomous vehicle claims and they will need to be careful not to overestimate the improvement in claims experience each year as autonomous vehicles are gradually rolled out.

As I have mentioned road deaths are expected to hugely decrease therefore actuaries should experience a decrease in mortality across all ages. Life policies could expect a slight decrease in premium due to a decrease in the likelihood of death while pension policies may experience a slight increase in premium due to a longer expected pay out period as a result of longer survival. Actuaries can also expect the ‘accident hump’ in the mortality curve to become less prominent. This is the part of the mortality curve that doesn’t follow the general increasing across ages mortality trend and in which mortality increases slightly before decreasing again across the age range of approximately 15-30. A big part of this is due to younger and newer drivers being more reckless when driving and having less experience on the road resulting in more accidents.

Health insurance companies should also experience a decrease in claims that health actuaries will need to take into account when pricing due to the reduction in injuries and necessary healthcare associated with road accidents.

autonomous vehicles on the highway
So although roll out of autonomous vehicles is still a while off yet it is clear that there are many factors that need to be considered before they can be offered to the public. General insurance companies and in particular the actuaries that work for them will need to begin planning for how insurance for these vehicles will be approached so that when the manufacturing and technical side of the vehicles has been perfected a lack of planning and consideration by actuaries doesn’t hold back the process of getting them on the road and making a beneficial impact within society!
​Read More
Autonomous Vehicles: Mortality & Insurance Implications
Data Science, Future of The Profession, Technology

Autonomous Vehicles: Mortality & Insurance Implications

In 1908, Henry Ford profoundly changed the automotive industry by developing and manufacturing automobiles at scale. The Ford Model T is generally considered to have been the first affordable car, subsequently ushering in the era of mass-market transportation and leading to widespread societal changes around the world.

111 years later, in 2019, the recent advances in computing power and artificial intelligence have made the previously science-fiction idea of living among unmanned vehicles, capable of navigating their landscapes without human input, a reality. A number of companies are already testing their vehicles in various locations and, since 2009, Google-owned Waymo has already driven more than five million (real road) miles, using self-driving technology (Waymo, 2018).

In a similar fashion to Ford’s global impact, autonomous vehicles will also change society, by significantly altering how we travel.

The areas of potential impact are wide and far-reaching and could include:

  • reduced car ownership
  • radically different car design geared more towards comfort and luxury
  • more older drivers, fewer taxi/bus/truck/delivery drivers
  • lighter burden on hospital and emergency services from fewer road accident injuries
  • significant improvements to rush-hour traffic

However, perhaps the most significant and important implications, at least to the actuarial profession, is the potential for reduced mortality and morbidity from traffic-related accidents and an overhaul of personal auto-insurance risks.

Mortality & morbidity implications

Previous research has indicated that over 90% of road accidents today result from human error. For example, the National Motor Vehicle Crash Causation Survey conducted between 2005 and 2007 attributed critical crash causation as follows:

Vehicle cr​​​​ash attribution

mortality statistics

As we try to forecast and imagine the future driverless world implications, we should first note that nearly 1.3 million people die globally in road crashes each year and an additional 20 to 50 million people worldwide are injured or disabled (Association for Safe International Road Travel, 2013). Indeed, road traffic injuries are currently estimated to be the ninth leading cause of death across all age groups globally and the leading cause of death among people aged 15-29 years (World Health Organisation, 2015). Given the potential for driverless cars to reduce accidents caused by human error the mortality and morbidity implications from autonomous vehicles are profound.

It is of particular interest to consider where these mortality effects are likely to have most impact. Unsurprisingly, traffic related deaths are not uniform across geographic location, socio-economic status, gender and age groups.

The World Health Organisation (WHO) highlights some of these disparities:

  • Income: The global average number of deaths per 100,000 population is 17.4. However, the breakdown between low income, middle-income and high-income is 24.1, 18.4 and 9.2 respectively (WHO, 2015).
  • Location: The African region has the highest fatality rates (26.6 per 100,000 population) and Europe has the lowest (9.3 per 100,000 population) (WHO, 2015).
  • Age: 60% of road traffic deaths are among 15-44 year olds (WHO, 2013).
  • Gender: 77% of all road traffic deaths are men (WHO, 2013).

Proportion of road traffic deaths by age range and country income status

WHO statistics

In terms of the potential for improvements in vehicle accident related mortality and morbidity, this may depend on the degree to which drivers in society can and wish to transition from fully operating vehicles to vehicles that are completely automated. Despite recent advances, there are still many hurdles and obstacles to overcome, and like any innovation there will be a prolonged period of transitional change before autonomous vehicles become mainstream. According to the Society of Automotive Engineers’ (SAE) J3016 standard there are six different levels of automation from level 0 (no automation) to level 6 (full automation), as shown below:

Insurance implications

Inevitably, the motor insurance world will change drastically as we move through the six levels of autonomy. As previously discussed, it is estimated that over 90% of road accidents today result from human error. Hence, personal car insurance will be redefined as risk moves from vehicle users to vehicle manufacturers and software/hardware suppliers.

Attribution of liability will become a much more grey area as discussed in AIG’s 2017 report, ‘The Future of Mobility and Shifting Risk’. In a survey they carried out asking “who is liable in a fully driverless world?” respondents identified various parties that might be liable in crash scenarios involving driverless cars. The parties identified included (AIG, 2017):

  • the car manufacturer,
  • software programmer,
  • vehicle occupant,
  • vehicle owner,
  • parts manufacturer,
  • internet service provider,
  • pedestrian and road manufacturer.

As the inevitable driverless world takes over, many traditional auto-related risks will no longer be as prevalent. Risks such as those caused by reckless or distracted driving, speeding, ignoring stop signs/red lights, unsafe lane changes, tailgating and road rage will be replaced by new, emerging risks such as malfunctioning software and cyber security.

The migration and ensuing calculation of risk will be particularly challenging during what has been called the ‘chaotic middle’ transition period where vehicle owners and the AI software share responsibility for the vehicle’s operation and any resulting liability.

Clearly, we are entering a new era of transportation. Despite the many challenges ahead, it appears that significant changes will be increasingly felt across many different aspects of society, as autonomous vehicles make their way into our everyday lives.


References

AIG (2017). The future of mobility and shifting risk. (Innovation + Risk series; 8). December 2017.
https://www.aig.com/knowledge-and-insights/k-and-i-articlethe-future-of-mobility-and-shifting-risk

Association for Safe International Road Travel ([2013]). ‘Annual global road crash statistics’; ‘Annual United States road
crash statistics’. http://asirt.org/initiatives/informing-road-users/road-safetyfacts/road-crash-statistics

Society of Automotive Engineers (2014; 2016). Taxonomy and definitions for terms relating to driving automation systems for on-road motor vehicles. J3016_201609. https://www.sae.org/standards/content/j3016_201609/

Singh, S. (2018). Critical reasons for crashes investigated in the National Motor Vehicle Crash Causation Survey. (Traffic Safety Facts: Crash; Stats. Report No. DOT HS 812 506). March 2018. Washington, DC: National Highway Traffic Safety Administration. https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812506

Waymo (2018). Waymo safety report: On the road to fully self-driving. https://waymo.com/safety

World Health Organization (2013). Global status report on road safety 2013: supporting a decade of action. http://www.who.int/violence_injury_prevention/road_safety_status/2013/en/

World Health Organization (2015). Global status report on road safety 2015. http://www.who.int/violence_injury_prevention/road_safety_status/2015/en/

Note

A version of this article was originally published in the Institute and Faculty of Actuaries Longevity Bulletin (Issue 11, September 2018) and reproduced here with kind permission. The original article can be downloaded here.

​Read More
data science - future of actuarial profession
Data Science, Insurtech, Learning, Misc

6 Thoughts on Future Strategies for the Actuarial Profession

I was recently contacted by another actuary – an IFoA council member – asking for my thoughts on strategies the IFoA should think about for the future.

I thought I’d share my response on here, and would love to hear what other actuaries think:

“…. Like many other professions, I truly believe that our profession is facing a critical time given the pace of change.

So what can the actuarial profession do to remain relevant well into the future? Here are a few of my personal thoughts:

1. Perpetual Learning

Going forward, I think, there needs to be a greater focus on “on demand learning/upskilling” and accessibility to resources as and when needed for learning on the job. Eg understanding and applying artificial intelligence, machine learning, blockchain, IoT, wearables, dealing with new emerging risks. The list goes on and is constantly evolving. I think the profession can and should continue to play a role in continuous ongoing learning after qualification. It increasingly seems that qualification is only the beginning and we need to be constantly learning and evolving as actuaries, to avoid getting left behind. Learning to learn and being flexible and having an ability to reinvent oneself are critical skills for the future. I believe the profession can help us stay nimble as well as fostering a culture of continuous ongoing lifelong learning. Otherwise, I believe, others (eg “data scientists”) that often have the more flexible mindset may continue to make inroads into traditional actuarial areas (eg pricing).

2. Innovation & Creativity

I would like to see a greater focus on innovation and creativity as well as more training in these areas. These skills are becoming increasingly important. The insurance world is now beginning to change and “catch up” with other sectors, in terms of technology adoption. We should be aiming to claim a slice of the InsurTech pie and not just seen as the regulator checkbox guys or technicians. Let’s learn to bring more of an agile mindset to our work. I think it will be increasingly important to be able to think like an entrepreneur. Building, testing, iterating. Let’s continually learn to take initiative and find unique solutions to the new emerging business problems.

3. Data Science

Data science, especially machine learning, is becoming more and more important as a tool for actuarial work. It’s great to see the profession doing a lot in this space. Can the actuarial qualification act as a more general route into a career as a data scientist? I think it possibly can. We have a unique position in this space due to our code of ethics and highly regarded professional qualification/recognised credentials. Hence, we have a unique advantage over many others claiming to be data scientists (given anyone can now say they are a data scientist!). The problem though is the opportunity cost of learning irrelevant material (for much of data science) to get the actuarial credential.

4. Professional Actuarial Judgement & Business Acumen

Related to the above – there’s a lot of talk about automation and people losing their jobs to robots. Whilst there is a certain amount of hype around this, I do believe automation will change the nature of many jobs in the future. Digital and automation is the future and we need to be part of that. Let’s, therefore, ensure that we continue to add value beyond the number crunching by having the skills to exercise professional judgment and business acumen whilst staying on top of developing trends.

5. Supply & Demand

I think the profession has a role to play here to help ensure this is managed carefully in different parts of the world. India, in particular, seems to be currently feeling the effects of an over-supply of new “aspiring actuaries” having difficulty getting on to the actuarial career ladder.

I believe we can collaborate more with other actuarial bodies to effectively build brand “actuary” (trusted advisors of risk). There will always be risk and especially today risk is everywhere and changing all the time. I would like to continue to see brand “actuary” as the trusted advisors in financial risk, but with more creativity and innovation. I believe one of the key positives of our profession is that it attracts really bright people. We should focus on continuing to make the profession attractive to the brightest and best.

Our new IFoA President, John Taylor, has talked about “growing the membership,” so I think it will be interesting to see what direction this takes.

6. Innovative Research

A greater focus on truly innovative research. We live in exponentially changing times. Let’s be at the forefront of change and bring a creative, innovative mindset to our work underpinned by relevant cutting-edge research.”

​Read More
The Last Actuary - Proactuary
Data Science, Future of The Profession, Technology

The Last Actuary

This essay was awarded an "Honourable Mention" in the Society of Actuaries' Actuarial Innovation and Technology Essay Competition in 2020. It has been republished here with kind permission from the Society of Actuaries. The original essay and other entries can be read in full here.

The Last Actuary

The church was packed.

Hundreds had gathered to pay their respects.

As the eulogy was being read out, my thoughts began to drift. My father was loved. Eccentric and slightly odd in his ways, but very intelligent and, most importantly, he always acted with complete integrity. Whether it was work, family or community related, he put integrity at the core of his approach. He partially thanked his chosen profession for instilling in him the importance of integrity in how he conducted his life.

These attributes had served him well, judging by the vast crowd of friends, family and former work colleagues that had gathered to say their last goodbye.

The church minister was humorously explaining to the silent crowd how my father had accurately predicted the exact week of his death—no doubt using the tools and expert judgement he had become so proficient in when he was working as an actuary.

Born in 1968, he had certainly had a good life, dying just 3 months shy of his 119th birthday. As the eulogy progressed, I became more and more intrigued about the profession that my father had made such an impact upon and that he had clearly loved.

When he finally retired, in 2038, the actuarial profession had also literally died a death.

I vividly recalled father speaking with passion at his retirement dinner. A poignant moment of his retirement speech, that remained with me, was when he was shaking his head, showing tangible regret and disappointment, as he talked nostalgically about his actuarial career.

He had discussed how times, and in particular the business world, had changed drastically. Eventually making his skills and profession obsolete.

Technology and data had been the driving forces of change and, as father used to say, these same drivers should have been what propelled his profession to greater heights and not towards oblivion.

But it wasn’t just the actuarial profession that had vanished. Paralegals had disappeared, doctors were pretty much gone. All victims of technology advancements and the merciless artificial intelligence (AI) algorithm. Even household maids were a thing of the past as robots took care of cooking, laundry, and other household chores.

The fuel behind AI was data, algorithms, and smart individuals. Father had, therefore, felt that actuaries were well placed to play a key role in the emerging paradigm. Particularly given the pressing need to consider the ethical and regulatory implications.

The New World

On my way home from the funeral I again began to think about the world we now lived in and the prophetic words father had uttered towards the end of his career.

My car was traveling at speed as I lay back in my seat and closed my eyes, safe in the knowledge that my autonomous vehicle would have me safely home at exactly 17.05. Technologists had finally overcome the moral and legal issues with self-driving cars and they were now of course ubiquitous bar the occasional hobbyist manual driver, that was confined to the 'leisure only' roads. It seemed strange to think of a time when the majority of people actually drove themselves from place to place and thus took on the potential liability of damaging someone else or their property.

Father, of course, had tried to warn his fellow actuaries that this new reality would have drastic repercussions for the many P&C actuaries. Risks and liabilities still remained, but they had shifted from the driver towards the manufacturers of the cars and their parts (e.g., sensors). As human input had all but disappeared from the driving experience, cutting out the 90% of risk previously caused by human error, risk and liability had also transferred to the producers of the software used to make the AI driven decisions as well as those responsible for building the transport routes and networks on which the driverless cars operated.

My father had been one of the more outspoken actuaries of his time, calling for widespread changes and a need to grasp the opportunities that were arising in the new digital and technology-focused world. If only they had listened to his advice, instead of laughing at his claims of the potential end of the profession, perhaps the actuarial profession would still be alive.

personal investment chatbot - Proactuary

As I arrived home and walked through my apartment door, my retina mail immediately downloaded via my permanent e-contact lenses. The e-letter was from my employer pension fund provider showing that my crypto and equity investments had passed the $5,000,000 mark and according to my personal investment chat-bot it was now time to start moving towards bonds to lower the fund volatility, given my planned retirement age of 86.

Scrolling further through the letter, using 2 successive blinks from my right eye, the sophisticated AI powered algorithms informed me that my expected death was still 112 with a 90% level of confidence.

I recalled father telling me how his first actuarial job involved calculating the liability for an employer's pension scheme where the employer made a promise to pay out a pension which was a fixed percentage of each individual's final salary. I wasn't sure if he was joking or not when he told me about this strange type of pension scheme, but a quick retina search revealed that these so-called "defined benefit" pension schemes had indeed existed and had kept many actuaries in work for a number of years. How times had changed!

My trail of thought was interrupted by my fiancé announcing that dinner was now ready. As I sat down to eat, the digital counter on the kitchen table automatically showed the exact amount of calories and breakdown of protein, fats and carbohydrates I was about to consume. Of course, this information would be automatically relayed to my insurer, Baidu, and a quick glance at the figures told me that my health premium score wouldn’t be adversely affected.

It had been a bad few days, since father’s death, as far as my real time health insurance premium was concerned. My usually well-organized diet had suffered and my implanted glucose sensor's historic data showed that I had been eating a lot of sugar heavy meals. As a result, my automatically calculated fasting glucose had crept over 100 mg/dL this morning, which had implications for my health insurance premium. But worst of all, my jewellery sensors had picked up the emotional stress response from having to drop everything and help with funeral arrangements whilst also dealing with the emotional shock that father was no longer with us. Despite his accurate predictions about his death and recent humorous quip that the trustees of his employer’s pension scheme would be “jumping with joy"  with the fact that another one of their scheme's liabilities was now extinguished, the news of his death still came as a blow, as if out of nowhere, and I could see the toll it was taking in cold hard stress figures.

A New Insurance Paradigm

Blockchain

All this health-related data was feeding directly into the autonomous blockchain enabled insurance company where the machine learning algorithms predicted my health risk with near 100% accuracy.

Ubiquitous blockchain proliferation had occurred. Not by 2025 as the hype had suggested, but by 2035 the exchange of value across the world was facilitated by the irrefutable and distributed technology.

actuaries and blockchain

Smart contracts were set up and utilized with ease which had led to a gain in trust in both the banking and insurance industries. Middle-men were a thing of the past. Unnecessary friction across nearly all services and transactions had now disappeared.

Adverse selection was also a thing of the past, as individual insurance risk predictions were now completely specific and accurate to each individual, ensuring that the insurers were not impacted by potential asymmetry of information.

Hence, bad risks were no longer being subsidized by the “good risks”. The traditional insurance pooling of risk no longer existed as every insured individual was accurately assessed on a completely personalized granular level. However, lack of pooling had created wider societal problems as the poor risks in society were now finding it very difficult to get the necessary cover.

Insurance companies had also changed in many other ways:

Baidu now had 90% of the global insurance provision market share, as they had leveraged off their advanced analytics and strong customer loyalty and engagement. Everything had moved from static to dynamic and real-time. From purchase and annual renewal to continuous offerings and interactions.

Data availability was ubiquitous with sensors everywhere feeding in voluminous data cheaply and easily for the sophisticated and automated algorithms to perform complex and extremely accurate calculations, which had moved beyond the realm of human understanding.

Wearables, social media, geolocation, weather and news were just some of the real-time data continuously feeding into insurer’s dynamic blockchain enabled data systems, where AI was doing the work previously done by actuaries.

A Shift to Prevention

My own insurance company acts not only as a financial security blanket but also a valued trusted advisor. The insights and health information it provides me on a daily basis has truly helped to overhaul my health behaviors and constantly nudge me in the right direction with alerts and monetary motivations.

The fact that all my coverage comes from a single provider, where I also get nearly all my other digital services, has meant that my customer experience is drastically efficient and simple. They even provide me with new unique bespoke personalized and dynamically priced products based on my changing needs.

Manual underwriting no longer exists. Claims processing is fast, accurate, cheap and efficient. Fraud is almost non-existent as the data and algorithms, delving into online social data can spot them with ease.

Everything is frictionless. The industry has shifted largely to one of prevention, risk monitoring and mitigation. I enjoy interacting with my insurer as they provide such valuable information that is provided in a clear simplified format. I truly trust that they will provide the necessary claims and/or advice should I need to call upon them at any point.

A World of Actuarial Opportunities

My father had foreseen these changes and viewed them as being catalysts and opportunities for actuaries to “move up the value chain” adding value in innovative creative ways, completely outside traditional actuarial work. His view had been that creativity, flexibility and the ability to innovate and reinvent oneself were going to be key skills for the future. He spoke about the need for actuaries to embrace technological changes

His message to actuaries of the future also included:

  • The need for actuaries to become more creative as new data sources and complex risks continued to appear and evolve.
  • The need for actuaries to evolve their mindsets so that perpetual ongoing learning was viewed as being very important for actuaries to not only thrive, but survive.
  • Barriers to entry will reduce in the future. Actuaries would no longer be protected, to the same extent, by credentials from exams and by regulatory work. Actuaries must ensure they are always adding value and remain focused on the needs of employers and their customers.
  • As part of being flexible and continuous innovators, actuaries must learn to work in diverse teams that often take an agile experimental approach to new problems which would involve failing fast and iterating as necessary.
  • As key members of these diverse agile problem-solving teams, actuaries would, at times, be expected to apply their specialist knowledge but they would also need to be able to act as generalists with an ability to see the bigger picture and to connect people and ideas.

Many of his fellow actuaries, unfortunately, thought he was being dramatic with such views. Actuaries don’t need creativity they claimed. Technology and AI is over-hyped they retorted. We are protected by impermeable barriers to entry they countered. However, risk and corporates had evolved and changed so quickly that father’s warning about creativity and flexibility being some of the most important attributes an actuary of the future could possess, did indeed seem prophetic. As he had forewarned, AI and technology had indeed penetrated every nook and cranny of business. Anything that could be easily automated had fallen prey to the majestic combination of AI and technology. Those that failed to embrace the fourth industrial revolution had indeed been left behind.

Hindsight, of course, as all actuaries know is a wonderful thing.

The Merciless AI Onslaught

As I brought up the the news on my e-retina I was greeted by some uplifting positive news. The new president of China and free leader of the world was speaking at an address and the support from the crowd was palpable.

Politics had moved on in leaps and bounds over the last decade and of course had changed immeasurably since the spectacular collapse of America, in 2032. President Wò sēn was everything a nation could hope for in a leader - completely altruistic, non-biased, 100% committed to the good of the people and more intelligent and rational than any president that had gone before him.

This was, of course, no surprise. President Wò sēn was after all an artificial intelligent robot. What began as a joke (“Watson For President”) in 2016 had set in motion the wheels towards the current reality. Surprisingly, the vast majority were embracing the new efficient world order where national decisions were based on terabytes of historic data, aided by quantum computers that had all but replaced traditional binary computers.

The results were difficult to argue against. World crime was now down 15%, healthcare costs were down significantly as chronic conditions were recognized at an early stage and patients had the means to understand their conditions and treat themselves much of the time.

Concluding Thoughts

I should finish by saying that I actually don't think that the actuarial profession will die a death. I believe there are too many talented and intelligent people in this profession for us to lose our relevance and for actuarial obsoletion to become a reality.

However, despite our many strengths, I do have the view that we are entering an era where the world of work will continue to change at a rapid rate. Increasing connections, technology, digital information and ideas are leading to exponential change throughout the world in many ways, whether we like it or not. To use a metaphorical quote from Malcolm X:

The future belongs to those who prepare for it today.

As a profession, whose very existence is based on the premise of being skilled at predicting the future and dealing with risk, let’s not fail in these regards, at this very important juncture.

​Read More
insurance actuary
Data Science

The Extinct Actuary?

I recently took part in a panel debate interview organised by Rick Huckstep from the Digitial Insurer. The write-up was titled: “Is the insurance Actuary an endangered species, or simply facing an evolutionary makeover?” and involved Geoff Keast (CEO of Montoux in North America) and Steven Mendel (Co-founder and CEO of Bought By Many).

Rick is one of my favourite InsurTech authors, so I knew he would do a good job of bringing up some interesting debate.

We covered many areas, and I’ve briefly summarised the top 13 key points below:

  1. 1
    As we enter the fourth industrial revolution and age of technology and automation, change is upon us. Many professions, including the 300 year old actuarial profession may be under threat.
  2. 2
    The digitial economy has led to an explosion of data. This data is emerging from many different sources and is also impacting the insurance industry (think IoT, Telematics, Wearables, etc).
  3. 3
    The non-actuarial “data scientist” is encroaching into actuarial work.
  4. 4
    Many actuaries are now adding machine learning to their toolkit and the actuarial profession is embracing this change.
  5. 5
    In the era of “data is the new oil”, actuaries may be able to use their skills to break into new non-traditional areas.
  6. 6
    Actuaries understand financial performance and policy-holder behaviour and hence they are the lifeblood of many insurance companies.
  7. 7
    We are likely to see a blending of actuarial and data science in the next five years.
  8. 8
    We are moving towards more dynamic pricing models as data increases in size and availability.
  9. 9
    New risks will emerge that will likely require actuarial expertise to understand (e.g. drone risk or liability from autonomous vehicles).
  10. 10
    We may see a shift towards growth outcomes (new business, customer retention, etc.) rather than the risk/compliance/ regulatory side of modelling.
  11. 11
    Automation may enhance the actuary’s work rather than leading to extinction. Assuming you are a forward thinking perpetual learning actuary!
  12. 12
    An actuary’s strong professional code of ethics provides an oversight which I think is particularly important for helping to ensure that we are using data ethically and responsibly.
  13. 13
    Being flexible and being able to adapt and grow is key for future actuaries.
​Read More