TESTING MEAN REVERSION CHARACTERISTICS OF DOCTOR EFFICIENCY MEASUREMENT

Matan Abraham

Presentation with Paper

October 24, 2018, 2:15 PM – 3:15 PM

Relevant practice area(s): Healthcare

Suggested audience knowledge level: Intermediate

 

  • Medical schemes can measure relative doctor cost efficiency and quality of care in treating patients through benchmarking analysis.
  • Doctors patient risk profiles are adjusted for to improve comparability.
  • Benchmark analyses influence scheme provider networks, performance-based reimbursement and peer review engagements.
  • This research considers the effect of mean reversion on benchmarking models to test the consistency of doctor behaviour over time.

INSURTECH HYPE – TECHNOLOGY IS NOT THREATENING YOUR WORKFORCE. POOR LEADERSHIP IS.

Dimitri Anagnostopoulos

Presentation

October 25, 2018, 11:45 AM – 12:45 PM

Relevant practice area(s): Wider Fields

Suggested audience knowledge level: Foundational

 

We read a lot about how technology is changing the financial industry and introducing innovation and efficiencies.

However, efficiency is a mind-set, not a tool. If you apply Artificial Intelligence in a wrong process you will simply be replicating your mistakes faster. So, artificial: Yes. Intelligence: not so much.

For a long time, we have allowed wrong and inefficient processes to spread across the organisation. Instead of teaching our employees how to improve them, we have left them to perform non-value adding activities day-in and day-out.

Are their jobs at risk? Yes, they are. But it is our fault as leaders.

It is a misconception to think (or hope) that just the adoption of technology will drive efficiencies in the organisation.

If you want to drive efficiencies, first focus on people and focus on change.

“It is not the strongest species that survive, nor the most intelligent, but the ones most responsive to change”.


CONSTRUCTING OPTIMAL INVESTMENT STATEGIES IN RETIREMENT BY BLENDING DEFERRED ANNUITIES WITH OTHER TRADITIONAL ASSET CLASSES

John Anderson

Presentation with Paper

October 25, 2018, 10:00 AM – 11:15 AM

Relevant practice area(s): Retirements Matters and Investments

Suggested audience knowledge level: Intermediate

 

This is a follow-up paper in relation to the design of optimal investment strategies at and after retirement.   The previous paper co-authored by myself and Steven Empedocles was titled “The retirement income frontier and its application in constructing investment strategies at retirement”.  This introduced a framework for evaluating different strategies at retirement.

We intend to look at the retirement income frontier, using parameters and assumptions consistent with the previous paper, and specifically analyse the impact of blending deferred annuities within the investment strategy and whether this leads to more optimal outcomes in terms of legacy potential and income sustainability compared to strategies only blending living and traditional immediate annuities.

The model, using assumptions specific to the South African environment as an example, is constructed to determine the lifetime spending needs being met and the expected financial reserve upon death using various strategies that include both draw-down accounts with varying asset allocations (or “living annuities”) and deferred annuities.

The results are then compared to the previous paper to determine the impact of deferred annuities and the properties (in terms of lifetime spending needs and expected financial reserve upon death) of including them into an investment strategy in retirement as opposed to typical guaranteed annuities.

We are aiming to within this also look at the impact of varying the income needs profile of a pensioner over time instead of this being fixed in real terms (to model the effect of a pensioner’s needs reducing in the early years but then increasing in later years as medical and the cost of care being a lot more significant at advanced ages).

Our overall aim is to determine whether the inclusion of deferred annuities in varying proportions result in more optimal outcomes for pensioners and whether the results differ according to whether a pensioner’s needs varies over time as suggested by international literature.


ACTUARIES AND MICROINSURANCE: PROFESSIONALISM, FINANCIAL SOUNDNESS AND VALUE TO CUSTOMERS

Janice Angove, William Melville, David Kirk, Seth Eshun

Presentation

October 25, 2018, 10:00 AM – 11:15 AM

Relevant practice area(s): Microinsurance

Suggested audience knowledge level: Foundational and Intermediate

 

This panel explores the role of actuaries in microinsurance and the implications of the South African microinsurance regulations for actuarial work in this area. Professional responsibilities need to be considered in the light of balancing the simplified technical requirements with challenges and risks of the business and the objective of ensuring an appropriate standard of actuarial services to entities that may have less experience with formal insurance. An issue of particular relevance to actuaries is the measurement of the financial soundness of microinsurers. The session will compare the capital requirements for a typical funeral insurer under the framework for financial soundness of microinsurers and the standard formula under the Solvency Assessment and Management regime and test the robustness of each method by investigating the financial soundness of the insurer under various stress scenarios. Monitoring the viability of microinsurance initiatives and value to customers is important in a market where customers are more financially vulnerable and may have less experience with insurance. This session will conclude with a global perspective on the performance of microinsurance initiatives in terms of viability, client value and fair treatment of customers, with insights from the Ghanaian insurance regulator on the experience of microinsurance in Ghana and interventions from the regulator to address concerns around fair treatment of customers.

Outcomes

  • Understand the expectations regarding actuarial input for microinsurance from a regulatory and professional perspective
  • Understand how these requirements coincide with the experience of actuaries working in microinsurance in SA
  • Consider the key questions for the implementation of SAM for microinsurance in SA
  • Identify the differences in the solvency modelling approach for SAM and the microinsurance regulations
  • Compare the minimum solvency requirement for a funeral insurer under SAM and the microinsurance regulations
  • Consider key questions about the fair treatment of customers and viability when assessing or giving advice on microinsurance initiatives
  • Consider the regulator’s perspective on fair treatment of customers and professional questions when assessing or giving advice on microinsurance initiatives

SAM RISK MARGIN SIMPLIFICATIONS: HOW THINGS CAN GO WRONG

Peter Carswell, Ross Farnaby, Nikhara Nirghin

Presentation with Paper

October 24, 2018, 2:15 PM – 3:15 PM

Relevant practice area(s): Life Insurance

Suggested audience knowledge level: Intermediate

 

New regulation requires South African insurers to follow the Solvency Assessment Management (SAM) framework and requires insurers to calculate a risk margin as part of the technical provisions that they must hold. The full calculation of the risk margin can be computationally expensive and consequently many insurers may make use of simplifications.  This paper aims to assess the appropriateness of the simplifications specifically for a life insurer offering annuity products, [term assurance and whole of life products] to assess whether there are biases within the simplifications. There are systemic biases resulting in understatements of the risk margin for annuities and the factors affecting the variability of the bias are explored. [Risk product result when it is complete]. Where the simplifications are found to introduce bias, some preliminary analysis of that bias is provided. Lastly, a solution is proposed that provides a more accurate method to estimate the risk margin.


BEYOND THE HYPE: A RISK MANAGEMENT FRAMEWORK FOR INSURANCE FIRMS IN THE ERA OF GRAND TECHNOLOGICAL PROMISES

Shasmi Choudree, Ope Bobade

Presentation

October 25, 2018, 11:45 AM – 12:45 PM

Relevant practice area(s): Enterprise and Financial Risk Management (ERM)

Suggested audience knowledge level: Foundational

 

Topic description

  • New technologies and associated methodological developments open up numerous opportunities for the actuarial profession
  • From the use of deep learning techniques (such as using Google’s Tensorflow for the evaluation of P&C claims) to the application of distributed ledgers for payout processes, actuarial practitioners can benefit from a great array of tools and supporting methodologies
  • Alas, no benefit comes without a risk, and the actuarial profession needs to pay close attention to the risks associated with the use of emerging technologies
  • The objective of the topic is to present the foundations for an actuarial risk management framework that addresses issues from reserving to cyber risk

Practical outcomes

  • How can the actuarial profession make use of the latest developments in technology (from analytics, to process, to product) without compromising our role in financial system stability?
  • What are the components of an enterprise risk management framework that takes into account 1) the new nature of existing risks, given technologies; and 2) new risks arising from technology itself
  • A practical understanding of cyber risks and ways to mitigate them (including the underwriting of cyber insurance)
  • Opportunities for the more strategic involvement of actuaries in the management of risks and the definition of business strategy

EV IS DEAD, LONG LIVE EV?

Anelda Dicks, Peter Tripe, Yura Kaliazin

Presentation

October 24, 2018, 12:40 PM – 1:40 PM

Relevant practice area(s): Data Analytics

Suggested audience knowledge level: Foundational

 

This will be a debate between a proponent and opponent about the future of EV given the transition from the SVM to the SAM basis and the future implementation of IFRS17. Topics covered include:

  • Will there still be a need for EV?
  • If we are not using EV, then what will replace it?
  • Is the CSM an adequate substitute for EV? Is it a consistent and realistic enough measure?

The debaters will consider a range of perspectives, including company and shareholder.

Outcomes:

  • Education regarding change in latest APN107 version
  • Understanding EV as we move from SVM to SAM and IFRS17
  • Robust discussion of topic

THE RIPPLE EFFECT OF SOLVENCY II: 2 YEARS IN

Colin Dutkiewicz

Presentation

October 25, 2018, 10:00 AM – 11:15 AM

Relevant practice area(s): Life Insurance

Suggested audience knowledge level: Foundational and Intermediate

 

This presentation will cover the following:

  • Balance sheet management activity since the launch of Solvency II in 2016, including case studies of actual transactions
  • How this interacts with the developments of other Risk Based Capital regimes globally
  • What lessons South Africa can learn with the implementation of SAM in 2018

Outcomes:

  • A broader understand of capital regimes and the direction of travel from the financial crisis through 2018 and beyond
  • An understanding of the interconnectedness of global life insurance capital regimes
  • Insights into particular solutions to capital problems
  • – Ideas of areas of exploration relative to SAM implementation

MACHINE LEARNING AND ARTIFICIAL INTELLIGENCE: TIPS AND TRAPS FOR INSURANCE PRICING AND BEYOND

Lukas Ehlers, Karl Meissner-Roloff

Presentation

October 24, 2018, 2:15 PM – 3:15 PM

Relevant practice area(s): Short Term

Suggested audience knowledge level: Foundational

 

Over the past decade Machine Learning techniques and Artificial Intelligence have become increasingly popular with various insurers now starting to utilise these techniques to price their insurance risks and automate their processes.

During our presentation we will discuss the pricing journey from data collection to model implementation, touching on how these models will impact the stakeholders at each step.  We will also demonstrate the pitfalls of using these models blindly without following generally accepted actuarial techniques and applying actuarial judgment throughout the process.

Outcomes from the presentation:

  • Understand how the most common machine learning techniques work and where they can be implemented
  • Understand the pricing modelling process from data collection to implementation and the stakeholder requirements at each stage
  • Understand the benefit of following the rigorous actuarial methodology and the importance of actuarial judgement
  • Understand the potential shortfalls when using these techniques and how to avoid them.

DEFAULT REGULATIONS AND IN FUND LIVING ANNUITIES – ISSUES TO CONSIDER WHEN RETIREMENT FUNDS IMPLEMENT IN FUND ANNUITY STRATEGIES AND THE EXPERIENCE AND LEARNINGS TO DATE

David Gluckman, Danie van Zyl

Presentation

October 24, 2018, 10:10 AM – 11:25 AM

Relevant practice area(s): Retirement

Suggested audience knowledge level: Intermediate

 

Sanlam Employee Benefits has experience in designing and administering In Fund living annuities for some of our retirement fund clients, and the aim would be to share what we have learned to date on this topic.

Practical Outcomes

  • Learning about pros and cons of In Fund Living Annuities
  • Could these work as a soft default for retiring members in terms of the default regulations?
  • What has been our experience with retirees who opted for these solutions?
  • Issues to consider as a trustee

INSURANCE AND ITS DISCONTENTS: CAN INSURTECHS SOLVE THE INFORMATION ASYMMETRY AND INCENTIVE PROBLEMS IN SHORT-TERM INSURANCE?

Sumarie Greybe, Alex Thomson

Presentation

October 25, 2018, 10:00 AM – 11:15 AM

Relevant practice area(s): Short Term

Suggested audience knowledge level: Foundational

 

Startup insurtechs regularly cite insurance’s many supposed problems and dis-functions, even going so far as to call insurance ‘broken’.  Is this simply the wishful thinking of new businesses trying to attract customers, or is there some truth to the claims?  And to the extent insurance isn’t delivering the goods, can a new breed of provider not just design better solutions but also compete effectively with the incumbents?

In the context of personal lines short-term insurance, this presentation explores:

  • How the structure of insurance products introduces specific information asymmetries that can impair market functioning
  • The areas where consumers are most vulnerable to the effects of these asymmetries, and their behavioural drivers
  • Circumstances in which provider self-interest can lead to insurance straying most egregiously from its economic function
  • Non-regulatory solutions, from brokers to the blockchain, their effectiveness in addressing these market failures and their prospects going forward

Practical outcomes:

  • Understanding how human biases and limitations present specific opportunities for personal lines insurers
  • Understanding how capitalising on these behaviours affects customer outcomes and market effectiveness
  • Insights into the prospects for insurance models such as peer to peer
  • How to re-appraise actuarial activities such as price optimisation

CAPITAL MANAGEMENT FOR SHAREHOLDER VALUE

Idelia Hoberg

Presentation

October 25, 2018, 10:00 AM – 11:15 AM

Relevant practice area(s): Life Insurance

Suggested audience knowledge level: Advanced

 

With the arrival of SAM, optimisation of capital under a risk-based framework will received increased focus. This presentation will discuss Idelia’s experiences from Europe, in particular:

  • Are there any lessons we can learn from the Solvency II implementation in Europe with regards to capital management,
  • What should firms be focusing on now in order to gain a competitive advantage in the future

The aim of the presentation will be to leave the audience with the following:

  • What are the areas of focus which were potentially not expected prior to the arrival of SAM with regards to capital management
  • Potential impact on business strategy
  • Potential areas of low hanging fruit stemming from lessons learned from Europe
  • Real “war stories” of areas of focus that were not successful

OPENSOURCE GEOCODING TOOLS AND DATASETS

Clive Hogarth

Workshop

October 24, 2018, 10:10 AM – 11:25 AM and October 25, 2018, 1:45 PM – 2:45 PM

Relevant practice area(s): Data Analytics

Suggested audience knowledge level: Foundational and Intermediate

 

The session will discuss open source datasets (and tools) available to assist with:

  • The geocoding of exposure information.
  • Augmentation of pricing datasets with external information.
  • Graphical summaries of exposure information.

Some R tools will be made available to the audience to assist with the implementation after the session.


HOW TO TALK ABOUT TARGET CAPITAL LEVELS

David Kirk

Presentation

October 25, 2018, 1:45 PM – 2:45 PM

Relevant practice area(s): Life Insurance

Suggested audience knowledge level: Foundational

 

Insurers are expected to hold capital above the regulatory minimum.  This “Target Capital” level has key implications for protection of policyholders and confidence in the industry as a whole. However, it also raises costs, dilutes returns to shareholders, diverts capital away from other opportunities and can act as a barrier to entry.

Regulatory feedback on ORSAs invariably asks for more clarity on why a particular capitalisation level (frequently 1.5x) was selected, and the answers are often poor.

The presentation will cover:

  • Why is capital above regulatory minimums necessary?
  • How do capitalisation levels compare across life insurance and short-term insurance in South Africa, and against select international markets.
  • What are some of the determinants of target capital level

A practical way of determining an appropriate target level of capital, or understanding what a specific target level of capital implies about risks in a way that management, the Board and the regulator find useful


WHAT IS FAIR AND WHO GETS TO DECIDE?

David Kirk

Presentation

October 25, 2018, 11:45 AM – 12:45 PM

Relevant practice area(s): Professional Matters

Suggested audience knowledge level: Foundational and Intermediate

 

We can easily agree that fairness is a good thing. It’s just that agreeing on what counts as “fair” has proved to be extremely difficult.

This presentation will explore different interpretations of “fairness” in a product development and pricing perspective, and the implications of the limited guidance for actuaries on these issues.

  • I will cover a summary of other research / presentations in this space, including criticism of actuaries in this space
  • I will cover the current drivers of expectations of actuaries of have a view on “fairness” (and distinguish this from premium adequacy, actuarial justification for differences in experience etc.)
  • Brief summary of requirements on actuaries around fairness from select global markets.
  • I will report back on brief survey of actuaries, statutory actuaries, management and NED views on whether and how they assess fairness. Differences in views by area of work, experience and role will be highlighted.

IFRS 17 – GETTING TO GRIPS WITH PROFIT AND REVENUE

Jacques Kritzinger, Christiaan Nel, Carolyn Clarke, Emce van Zyl

Presentation

October 24, 2018, 10:10 AM – 11:25 AM

Relevant practice area(s): Life Insurance

Suggested audience knowledge level: Foundational and Intermediate

 

Description:

  • IFRS 17 aims to recognise profit and revenue as insurance services are provided
  • Prior to its development no attempt had been made to adjust premiums recognised for long-term contracts to those earned in the period, rather than written
  • The mechanics to achieve this are described over a 300-page IFRS standard
  • The presentation will aim to explain the concepts in a simple and accessible manner and as far as possible, visually

Outcomes:

  1. Gain a robust understanding of the IFRS 17 definitions of profit and revenue through understanding the basic underlying concepts as well as the practical mechanics;
  2. Understand the drivers of profit recognition;
  3. Understand whether and how premiums are recognised as services are provided;
  4. Know and understand key interpretation issues affecting the mechanics which have been identified through developing implementation plans and the latest industry views on how they should be addressed. These include the treatment of premium experience adjustments and how the impact of inflation on each type of cash flow, both past and future, should be treated;
  5. Understand the potential impacts of the standard on profit and revenue, as well as other metrics; and
  6. Be empowered to effectively convey this understanding to your stakeholders within and outside of the business (boards, shareholders, analysts etc.).

BITCOIN DECRYPTED: WHAT DOES IT MEAN FOR FINANCIAL SERVICES?

Imran Lorgat

Presentation

October 25, 2018, 1:45 PM – 2:45 PM

Relevant practice area(s): Wider Fields

Suggested audience knowledge level: Foundational and Intermediate

 

Bitcoin is often in the headlines, but few understand it as a new technology. Few realize that Bitcoin was the first Blockchain and started 9 years ago, or that it’s already changing the way we do international business. In this talk I’ll demystify the technical aspects of Bitcoin for a financially-literate audience, and I’ll discuss what impact it’s having, and could have in the future, on the Insurance and Financial Services sector.

I’ll explain and discuss

  • What is Bitcoin, what it does, and where it came from
  • How Bitcoin Transactions work
  • How Blockchain works
  • How Mining works
  • What drives Bitcoin’s value (as a payment network, and as a currency asset)
  • How Bitcoin functions as an asset
  • What Bitcoin is currently doing in Financial Services (currency, insurance cover, smart contracts, remittances, asset management)
  • What Bitcoin could be doing in the future.

Outcomes for the audience

  • Understand the different components of ‘Bitcoin’ (software, payment network, digital asset on a Blockchain)
  • Understand how the Bitcoin network and Bitcoin transactions work, as well as the advantages of the technology
  • Be able to make an informed decision on Bitcoin’s value as an asset
  • Be aware of the many current and potential applications of this new technology in the Financial Services space
  • Be entertained. I’d prefer to keep the talk at a low technical level so that anyone with a financial background can enjoy it

DRIVERLESS VEHICLES: HOW THEY WORK AND HOW THEY WILL BE PRICED

Sarvesh Maharaj, Lukas Ehlers

Presentation

October 25, 2018, 10:00 AM – 11:15 AM

Relevant practice area(s): Short Term

Suggested audience knowledge level: Foundational

 

With self-driving vehicles currently being prototyped on our roads, tech giants are promising that motor insurance will all but disappear in the near future.

During our presentation, we will examine what the real impact of these vehicles will have on the cost of the current insurance risks as well as the cost of the new risks these vehicles will introduce and how these could be allowed for in the insurers pricing models.  We will specifically consider how these vehicles will change the nature of liability claims.

Outcomes from the presentation:

  • Broaden thinking around what the future motor insurance landscape will look like under self-driving vehicles.
  • Understand how self-driving vehicles manage to “see” the road.
  • Understand how much the current premium rating practices may need to adapt.
  • Understand what components of the comprehensive motor insurance premium will change, how they could potentially change and what will remain as is.

WHAT DO YOU WANT? MANAGING INVESTMENTS FOR BETTER OUTCOMES WHEN YOU RETIRE

Warren Matthysen

Presentation with Paper

October 25, 2018, 10:00 AM – 11:15 AM

Relevant practice area(s): Retirements Matters and Investments

Suggested audience knowledge level: Intermediate

 

The paper intends to have a broad discussion about the provision of benefits in retirement. the main premise is that actuaries in the past advised Defined Benefit (DB) schemes, where a number of experience risks were pooled (at sponsoring employer level) between different generations of employees/members. in a Defined Contribution (DC) environment the risks are all ‘managed’ by the single member. It has been largely managed by the concept of maximising investment return to build up enough funds to cover all risks.

This is potentially an inefficient way to manage the benefit risks. I hope to position a way to ‘pool’ risks, charge for them ‘efficiently’, in terms of costs for pensioners/annuitants, and manage investments to obtain better benefit outcomes in retirement.

The paper will hopefully cover the following:

  • Provide a perspective on the decision-making process for retirees
  • An investment decision that balances a desire for a life benefit (an income benefit on survival) and a death benefit (legacy benefit on passing away)
  • The behavioural desire to ‘control’ the asset and not to pay it away to an insurer, e.g., for a guaranteed annuity
  • The benefits of different investment strategies to maximise expected return outcomes
  • The benefits of different investment risk management techniques to manage the risk associated with meeting benefits expectations
  • The benefits of pooling risks, e.g., mortality/longevity to provide the benefits when they are required – death/survival

Combining findings from this risk management to propose a framework for the retirement investment decision


SLAYING THE DRAGON – DEMYSTIFYING EXAM RESULTS & IMPROVING INDIVIDUAL AND COMPANY RETURNS ON INVESTMENT IN ACTUARIAL EDUCATION

Mike McDougall

Presentation with Paper

October 24, 2018, 12:40 PM – 1:40 PM

Relevant practice area(s): Education

Suggested audience knowledge level: Intermediate

 

Investigating factors impacting performance in actuarial exams including:

  1. Demographic factors
  2. University
  3. Employer
  4. Interventions (courses, counselling, SAADP)
  5. Prior attempts
  6. Exemptions

Also, an analysis of study benefit policies and their impact on qualification rate.

Target audience: actuarial managers, students & academics

Practical outcomes:

  1. Managing students for exam success
  2. Successful study leave policies
  3. Successful interventions
  4. Optimising the cost of qualification
  5. Return on study investment

REINSURANCE IN THE IFRS17 OPERATING ENVIRONMENT

Michael Meerkotter, Theuns Botha

Presentation

October 24, 2018, 10:10 AM – 11:25 AM

Relevant practice area(s): Life Insurance

Suggested audience knowledge level: Intermediate

 

IFRS17, the accounting standard for insurance contracts has introduced a number of changes to the financial operating environment for insurers. These changes start with measurement and work their way through the entirety of insurers’ reporting and data systems.

To date the focus has been largely on the measurement challenges related to insurance contracts issued. However, as time has progressed, other facets such as dealing with the measurement of reinsurance contracts has begun receiving attention.

Under IFRS17 a number of changes to the measurement of reinsurance contracts held by an insurer have been introduced. These changes in some cases go well beyond the changes to insurance contracts issued and will result in operational complexity when measuring these contracts.

Due to the significance that reinsurance plays in the day to day operations of many insurers, the complexity arising from the standard’s requirements will likely see many insurers reviewing their existing arrangements in order to simplify the operational demands placed on them.

This presentation will focus on the practical challenges arising from the requirements of the new standard: namely what the requirements mean from a technical and industry point of view. In doing so, we will illustrate the measurement differences between the prevailing regimes and those arising from IFRS17 and provide an industry perspective on what the impact of these requirements will mean going forwards.


IS THERE ROOM FOR ACTUARIES IN PUBLIC POLICY?

Lusani Mulaudzi

Presentation

October 24, 2018, 2:15 PM – 3:15 PM

Relevant practice area(s): Professional Matters

Suggested audience knowledge level: Foundational, Intermediate and Advanced

 

The Actuarial Society of South Africa is busy grappling with how to engage in Public Policy.  At the moment, the Society comments on green papers, white papers and any other important directives from regulatory authorities, parliament and government departments.  This is deemed to be reactive and there is a growing realization that the Society needs to be more proactive.  There are general perceptions that the Society serves the needs of corporate South Africa – especially the insurance industry.  On the contrary, the Actuarial Society views itself as an independent voice on public matters and seeks to cement this view.  To achieve this, one of the options being currently considered is to enhance the level and frequency of research into public policy issues.

The research paper will:

  1. Explore the effectiveness of historical examples of public policy commentaries by the Society.
  2. Analyse examples of effective public policy interventions by similar organisations around the world and in South Africa.
  3. Identify characteristics for effective public policy engagement.
  4. Make recommendations.

HOW MUCH IS ENOUGH: BRINGING ACTUARIAL MODELLING TO BEAR ON MEDICAL SPECIALIST PLANNING IN A CHANGING WORLD

Shivani Ranchod, Dave Strugnell

Presentation

October 24, 2018, 10:10 AM – 11:25 AM

Relevant practice area(s): Healthcare

Suggested audience knowledge level: Foundational

 

We describe the modelling approach taken to estimate both the future supply of medical specialists and the demand for their services in South Africa. This requires working with multiple, imperfect data sets, and bringing together demographic, epidemiological, actuarial, economic and public health perspectives. We highlight the importance of this work, given the potential overall shortage of and severe maldistribution of medical specialists in South Africa and illustrate the ways in which the model can facilitate long-term planning and resource allocation.

Outcomes

  1. An understanding of how actuarial modelling techniques can be applied to solving public health problems;
  2. Practical insight into the combining of data sets, dealing with weaknesses of data sets and methods to deal with de-identification of personal information;
  3. Insights into dealing with the absence of centralised data – there is neither a national repository of data on medical specialists nor a census of medical specialist data.
  4. Practical insight into the need for multi-disciplinary approaches to solving complex problems – in this case ensuring that there is allowance for notions of fiscal constraint, structural changes in the population, changes in the burden of disease, technological changes in the way medicine is practiced, feminisation of the medical professions, productivity changes and changes to the ways of work (including vertical and horizontal substitution);
  5. An awareness of ensuring model usability (i.e. how to build end-user understanding into model design); and
  6. Illustration of the importance of a rigorous research-based approach to modelling to leverage off extensive experience in other countries.

AI IN ACTUARIAL SCIENCE

Ronald Richman

Presentation with Paper

October 25, 2018, 11:45 AM – 12:45 PM

Relevant practice area(s): Short Term

Suggested audience knowledge level: Intermediate

 

Rapid advances in Artificial Intelligence and Machine Learning are creating products and services with the potential not only to change the environment in which actuaries operate, but also to provide new opportunities within actuarial science. These advances are based on a modern approach to designing, fitting and applying neural networks, generally referred to as “Deep Learning”. This paper investigates how actuarial science may adapt and evolve in the coming years to incorporate these new techniques and methodologies.  After providing some background on machine learning and deep learning and providing a heuristic for where actuaries might benefit from applying these techniques, the paper surveys emerging applications of AI in actuarial science, with examples from mortality modelling, claims reserving, non-life pricing and telematics.  For some of the examples, code has been provided on GitHub so that the interested reader can experiment with these techniques for themselves. The paper concludes with an outlook on the potential for actuaries to integrate deep learning into their activities.

The audience will take away the following:

  • Understand how actuaries are starting to use modern AI and machine learning tools for classical actuarial problems
  • Become aware of the possibility of applying these approaches to their own work
  • Gain a view of the common tools used for machine learning in wider industries
  • Be introduced to some practical examples they can take home and experiment with

THE BOTS ARE HERE! – ANALYTICS & AUTOMATION IN LIFE INSURANCE PRICING

Louis Rossouw

Presentation

October 24, 2018, 10:10 AM – 11:25 AM

Relevant practice area(s): Data Analytics

Suggested audience knowledge level: Foundational

 

In an exponential world advances in analytics may change the way in which actuaries approach pricing.  In such a world we need more accurately priced business in shorter amount of time often involving minimal or no human intervention.

This talk will cover:

  • Underlying technological trends affecting pricing.
  • Examples of how tech can automate actuarial experience analysis and pricing or even removing the need for human involvement.
  • Role of Machine Learning and AI.

Practical Outcomes

  • Attendees should get an overview of technical trends affecting pricing
  • Attendees would get a sense of what might be possible in future with pricing.
  • Understand the potential impact of machine learning and AI on pricing.

MACHINE LEARNING ACTUARIES

Louis Rossouw

Workshop

October 25, 2018, 10:00 AM – 11:15 AM

Relevant practice area(s): Data Analytics

Suggested audience knowledge level: Foundational

 

This workshop will cover the basics of machine learning and cover the main basics of the main techniques of machine learning explaining in lay terms how they work together with advantages and disadvantages of these approaches.

  • Regression
  • Decision Trees
  • Random Forrest
  • Gradient Boosting
  • Support Vector Machines
  • Neural Networks

The session will also focus on:

  • The importance of the task at hand and the associated data
  • Common pitfalls in machine learning tasks
  • Measuring performance of models

Take-ways would be:

  • Broad understanding of machine learning techniques
  • Examples of the above in R
  • Importance of understanding the data
  • Ways to measure the performance of these models

AUGMENTED REALITY INSURANCE

Adriaan Rowan, Alex Conway

Presentation

October 24, 2018, 10:10 AM – 11:25 AM

Relevant practice area(s): Data Analytics

Suggested audience knowledge level: Foundational

 

The presentation will show a working prototype of Augmented Reality Insurance (ARI). ARI allows your mobile phone to become a real time interface with which to register for insurance, get a quote on each item you want to insure and submit a claim.

One example is car insurance. To insure your car, you just need a photo of your ID, drivers licence, car VIN and take 10 photos of the car. No phone call or forms needed. The tool can automatically value your car, assess damage and produce the quote based on your location.

To claim, all you need to do is take a picture of the damage and it gives you a quote of repair cost and recommends where you can get it repaired. The GPS and time stamp helps reduce fraud.

This can be extended to any type of short term insurance and example applications will be demo’ed.

People don’t want to buy insurance, but they do want to own a car. If the process of getting an insurance quote can be seamlessly incorporated into the car search process, the insurance sales process becomes a value-add service.

“The future is already here — it’s just not very evenly distributed” – William Ford Gibson


MEZZANINE DEBT: IS THIS SANDWICH REALLY A FREE LUNCH?

Chris Siriram

Presentation

October 25, 2018, 1:45 PM – 2:45 PM

Relevant practice area(s): Investments

Suggested audience knowledge level: Intermediate

 

  • The introduction of SAM requires insurers to consider economic capital i.e. a risk-based regime
  • Much has been written about the efficacy (or otherwise) of traditional asset classes such as equities and fixed income (including credit) but the discourse has not yet broadened out to private and mezzanine debt
  • Sandwiched between equity and senior debt, mezzanine debt has a unique role to play in the investment strategy of insurance companies
  • We show that in a South African context, both private and mezzanine debt will help insurers achieve their return objectives after allowing for, not only investment risk, but also the cost of capital
  • We compare and contrast the expected investment return of different asset classes against a ‘SAM-aware’ efficient frontier to further the debate on constructing optimal investment portfolios under SAM
  • We focus on the unique contribution of private debt, and specifically mezzanine debt, and illustrate its economic value to an insure
  • We conclude by highlighting some of the work done by actuaries globally in assessing these ‘alternative’ asset classes and the relevance of these studies to the South African market.

PRACTICAL DATA SCIENCE USE CASES IN THE 4TH INDUSTRIAL REVOLUTION

Stefan Steffen

Presentation

October 25, 2018, 1:45 PM – 2:45 PM

Relevant practice area(s): Data Analytics

Suggested audience knowledge level: Foundational and Intermediate

 

This presentation will provide greater context for the impact of the 4th industrial revolution whilst demonstrating practical data science use cases on how to navigate it.

1. Understand the critical context that defines our future landscape

  • Exponential change in data volumes, connectivity, storage and compute drives the 4th industrial revolution
  • Acceleration in the advance of algorithms and machine learning touches every aspect of life and business
  • The implication of “speed of light” limitations on the future of cloud computing and the shift towards the edge

2. Get a sense of practical use cases that point to what you should do now to capitalize on this disruptive change

  • Machine learning application to Telecoms data
  • Computer vision impact on retail customer experience
  • A blend of data science and IOT that is driving smart farming
  • The impact of Cassandra and Confluent Kafka on end-to-end supply chain visibility

You will gain the following by participating in this session:

  1. An overview of the key drivers of the 4th industrial revolution
  2. An understanding of “speed of light” limitations and what that means in the 4th industrial revolution
  3. Practical application of data science techniques in various real business contexts
  4. A layman’s understanding of 4 data science/machine learning/deep learning techniques

A FORMULATION OF ECOLOGICAL VALUE ADDED

Rob Thomson

Presentation

October 24, 2018, 12:40 PM – 1:40 PM

Relevant practice area(s): Wider Fields

Suggested audience knowledge level: Foundational

 

Can human activity as practiced today sustain ecological, social and economic well-being, and if not, how can we change economic activity so that the ecology, society and the economy can sustain it? The measurement of sustainability will assist in identifying the main contributors to the problem. Without ecological sustainability we cannot have economic sustainability. Therefore, the presenters propose a formulation of the effects of an entity’s activities on the biosphere or, as they refer to it as, the entity’s “ecological value added”.

The past five years have seen increasing awareness of the effect that ecological, social and economic unsustainability will have on the actuarial profession. The reality is that the future will not look like the past. We shall see effects on actuarial assumptions, modelling and advice. Unsustainability will potentially redefine the economic paradigm in which actuaries operate to include the ecological domain. Given actuaries’ skills and experience in the modelling of the outcomes of entities’ activities in the financial domain and the eclectic nature of our discipline, the actuarial profession can complement the skills of other professions to help address the challenges of ecological, social and economic sustainability.

On the other hand, the profession has much to learn from other disciplines.

This research will highlight ways in which the actuarial profession can learn from and contribute to the research agenda.

At present there is no methodology for the measurement of an entity’s ecological value added or for the establishment of an entity’s ecological sustainability. The purpose of this paper is to develop and formulate such a methodology.

Whilst the formulation does not include parameterisation at this stage, it demonstrates the possibility of holding any entity accountable for its ecological value added.


GROWING INSURANCE AND CHALLENGES IN EMERGING MARKETS

Jaco van der Merwe

Presentation

October 24, 2018, 12:40 PM – 1:40 PM

Relevant practice area(s): Short Term

Suggested audience knowledge level: Intermediate

 

Insurance penetration rates in emerging markets are still considerably less than in developed countries with the majority of the population remaining uninsured. There are a number of supply and demand side factors, which have led to an underinsured population. We will identify and rank these factors.

These low penetration rates are on the backdrop of a growing more educated and technologically savvy middle class. This offers tremendous opportunities for insurance companies. However, there are some key challenges that still need to be overcome in order to capitalise on this opportunity. We will consider the outlook for insurance in emerging markets, taking into account local sentiment.

Some key challenges facing insurers in emerging markets are inappropriate distribution channels, products that do not meet customers’ needs and outdated paper-based administration systems. We will consider how these challenges have inhibited growth and what insurance companies can do to overcome them. We will also look at some success stories from companies, which have harnessed the power of technology through adopting new distribution channels and developing more customer friendly products to improve their market share.

The government undoubtedly has a crucial role to play in developing appropriate regulation and ensuring that its citizens have ample access to the benefits of insurance. Some examples include, making insurance compulsory, developing risk-based regulation, micro-insurance and establishing insurance programmes to alleviate the impact of catastrophes or weather-related perils e.g. the Sugar Insurance Fund Board (“SIFB”) in Mauritius and the African Risk Capacity (“ARC”). We will consider these examples in more detail and provide local sentiment on the impact of regulation.

We will conduct a survey on a number of insurance companies operating in emerging markets primarily focused in Africa. Our presentation is based on feedback from survey participants and any conclusions we can draw from analysing the survey results.

Outcomes that the audience will be expected to take out of the presentation will be to understand:

  • The key challenges responsible for the low insurance penetration rate in emerging markets.
  • The challenges facing the distribution of insurance products in emerging markets.
  • What product innovation is required to make products more customer centric.
  • How technology can be harnessed to improve penetration rates.
  • What the government is doing to develop the insurance market and increase access to insurance.

A PROJECTION OF THE SA GOVERNMENT’S SOCIAL SECURITY OBLIGATIONS: 2017-2037

Natalie Van Zyl, Dewald Muller, William Melville

Workshop

October 24, 2018, 12:40 PM – 1:40 PM and October 25, 2018, 11:45 AM – 12:45 PM

Relevant practice area(s): Retirement Matters

Suggested audience knowledge level: Foundational and Intermediate

 

The workshop will provide an overview of the creation of a 20-year social security benefit projection model to be used as a comparison to government grant cost modelling.  Social grant spending encompasses, amongst others, State Old Age Pensions expenditure and child grants.

An initial presentation will include a description of the data and assumptions used in the model as well as challenges experienced when creating and checking the model.  Three different projection methods were used, and their results will be presented and compared.

Workshop participants will then be asked for their input on: assumptions underlying the model, the projection methods used versus possible alternate methods as well as comments on projection outputs and potential future model expansion.

Outcomes – The audience will:

  1. Understand current government social grant spending levels as well as projected future levels.
  2. Understand the challenges faced when creating a SA social grant spending model and possible approaches to overcome these challenges.
  3. Experience satisfaction in having participated in a debate regarding the appropriateness of the model.
  4. Have had an opportunity to influence future development of the model.

CLOUD-BASED MANAGED SERVICES IN BANKING AND AN IFRS 9 CASE STUDY

Matthew Walker

Presentation

October 24, 2018, 10:10 AM – 11:25 AM

Relevant practice area(s): Banking

Suggested audience knowledge level: Foundational

 

Advancements in technology across the continent, combined with ever-increasing regulatory demands, have resulted in cloud-based managed services presenting a great opportunity for African banks to fast track their adoption of advanced modelling techniques. This will assist them in ensuring regulatory compliance as well as with broader risk management.

This session will include:

  • A high-level explanation of cloud computing.
  • An overview of managed services, how they differ from typical outsourcing and the types of operations and conditions that lend themselves to managed services.
  • Some of the practical challenges associated with cloud-based managed services.
  • A case study on how a cloud-based IFRS 9 managed service could work for a bank in Africa.

This session will be of interest to anyone wanting to know more about the application of technology to risk management.


THE LIFE ESIDIMENI ARBITRATION AND THE ACTUARIAL QUANTIFICATION OF CONSTITUTIONAL DAMAGES

Gregory Whittaker

Presentation with Paper

October 25, 2018, 10:00 AM – 11:15 AM

Relevant practice area(s): Damages

Suggested audience knowledge level: Intermediate

 

In the instance of a claim for constitutional damages where an aggrieved party makes a claim against the State for damages resulting from its failure to uphold a constitutional imperative, how are such damages to be quantified? Under South African law there is no formula and extremely limited precedent outlining the calculation of constitutional damages. This paper will consider the Life Esidimeni Arbitration proceedings against the Gauteng Department of Health pursuant to the tragic mass death, torture and disappearance of mental health care users from the perspective of an actuary acting as an expert witness for the families of the deceased. There is no manual for calculating the monetary value of a life. Notwithstanding, this paper will set out the considerations made to reach monetary compensation as argued by the legal representatives of the families, as substantiated by the actuary and as eventually awarded by the Arbitrator.

Outcomes

  • Thoughts on the quantification of constitutional damages.
  • The importance of actuaries performing pro-bono work.
  • Expert witness work in the context of alternative dispute resolution forums.
  • Managing relationships between various inter disciplinary teams.