Bootstrapping is a nonparametric approach for making statistical estimation and inference when standard parametric assumptions are questionable, or computationally infeasible.
Applying this process to insurance allows estimates to be made of the distribution of most values of interest: future claims, capital, profit, and market returns. Two important sources of uncertainty in the projected values can be taken into account: for instance, the intrinsic volatility of the claims process, and the uncertainty in our parameter estimates of that process.
There are various approaches to constructing confidence intervals with this estimated sampling distribution that can be then used to make statistical inferences. More accurate standard errors, conﬁdence intervals, and even hypothesis tests for more complex samples can be derived using bootstrapping methods.
Bootstrapping method can also be applied to the work in the areas of life insurance, pensions or annuities such as addressing the challenges due to longevity risk. Not only for insurance companies, but this issue is also important for such things as demographic planning as well.
In measuring longevity risk, it is important to be able to quantify the uncertainty in mortality projections through the computation of prediction intervals. For actuarial pricing and reserving purposes, the mortality table will need to be projected to allow for improvement in mortality to be taken into account and sometimes that mortality development can be challenging to predict.
Rather than making assumptions about the sampling distribution of a statistic, bootstrapping uses the variability within a sample to estimate that sampling distribution empirically. This is done by randomly resampling with replacement from the sample many times in a way that models the original data generation process.
We have invited Dr. Frank Ashe, who has more than 30 years of business experience predominantly in risk management, investments and behavioral finance, to conduct a one-day hands-on workshop to show you how to use bootstrapping method to estimate capital and reserves when the usual methods for ascertaining statistical significance do not apply. Dr Ashe’s first use of this method for insurance reserves was in the early 1980’s. Dr Ashe will also show you how the bootstrap allows you to see the distribution of possible mortality rate for future years, based on cohorts, and to estimate the effect of increasing life expectancy on the value of pensions.
KEY LEARNING OUTCOMES
Understand the reasons why the bootstrap is being adopted in business modelling.
Understand the approach underlying the bootstrap implementation, implicit assumptions, and appropriateness to a particular problem.
Learn to know when to use the bootstrap method.
Be able to undertake a bootstrap process for estimating reserves, and as part of an ORSA/ICAAP.
Able to explain to stakeholders why using bootstrapping is a necessary tool in your work.
Each session in the workshop will focus on a different application of the technique:
Investment Returns, General Insurance, Life Insurance & Capital Adequacy.
0900 – 1030
– What is the bootstrap?
– Why was it developed?
– Variations of the technique including block bootstrap.
– Example using time series of market returns.
1030 – 1045
Coffee/ Tea Break
1045 – 1230
– A quick review of the basic triangle of paid claims, and techniques for reserve estimation.
– Stepping through the bootstrap process, demonstrating changes required for different models of the claims process.
– Analysis of outputs (there is more information than just the reserve value!)
1230 – 1400
1400 – 1530
A Little Theory and Parameter Error
– What justification is there for using the bootstrap?
– When can you use it?
– What is parameter error?
– Estimating the effect of future mortality risk on reserves & pricing.
– Incorporating parameter error into the estimate of reserves.
– Bonus! You now have a pricing tool!
1530 – 1545
Coffee/ Tea Break
1545 – 1700
Model risk, Capital Requirements, and Communication
– Which is the correct model to use for claims estimation?
– What if you can’t decide which model is best?
– Estimating capital requirements – simple model.
– Communication of the results to the Board and regulator.
End of Workshop
28 August 2018, Tuesday
0845 – 1700
73, Jalan Raja Chulan, Bukit Bintang, 50200 Kuala Lumpur, Malaysia.
* Early bird rate offer is valid until 7 August 2018 only.
Closing Registration Date
21 August 2018
The workshop will be a mixture of lectures and hands-on exercises. Participants are required to bring their laptops to the workshop.
Dr Frank Ashe
Founder, Quantitative Strategies Pty Ltd
Dr Frank Ashe has a consulting practice specialising in risk management, investments and behavioural finance. Risk management covers the gamut from technical matters in option risk, to strategy, to comparative corporate governance and risk culture. He is an Honorary Associate Professorship at the Macquarie University Applied Finance Centre where he spent 2002 to 2006 as a full-time Associate Professor.
Dr Ashe has worked in consultancies, insurers, investment management firms, bond dealers, and financial software houses in Australia and Canada. His 30+ years of practical experience have been predominantly in the measurement and management of financial risk and return, with an emphasis on asset-liability management and developing risk measurement and management tools for novel situations.
Dr Ashe has been presenting the 2-day course required for CERA qualification by the Actuaries Institute (Australia) since 2010. He is a regular presenter at industry seminars and colloquia through Asia, teaches financial risk management in East Asia and was President of the Australian Q-Group from 2002 to 2011.
Dr Ashe obtained his PhD in Operations Research from the University of New South Wales.
+603-21610433 ext 204
Led by professional with more than 30 years of experience
Practical & Relevant
Networking Opportunity with Fellow Professionals
Who Should Attend
This programme is suitable for actuarial & risk management professionals wanting to learn more about stochastic bootstrapping methods, particularly for estimating capital and reserves as well as estimating mortality risk.
Suitable also for managers of quantitative staff, risk committee members, auditors, researchers and regulators.