Actuarial Judgement Research Articles

Dependency Modeling Approach of Cause-Related Mortality and Longevity Risks: HIV/AIDS

Disaggregation of mortality by cause has advanced the development of life tables for life insurance and pension purposes. However, the assumption that the causes of death are independent is a challenge in reality. Furthermore, models that determine relationships among causes of death such as HIV/AIDS and their impact on mortality and longevity risks seem trivial or inflexible. To address these problems, we aim to determine and build an appropriate copula dependence model for HIV/AIDS against other causes of death in the presence of age, gender, and time. A bivariate copula model is proposed to capture the dependence structure of HIV/AIDS on life expectancy. This approach allows the fitting of flexible and interpretable bivariate copulas for a two-dimensional case. The dataset was derived from the World Health Organization database that constituted annualized death numbers, causes, age, gender, and years (2000 to 2019). Using Kendall’s tau and Pearson linear coefficient values, the survival Joe copulas proved to be a suitable model. The contribution and implication of this research are the quantification of the impact of HIV/AIDS on a life table, and, thus, the establishment of an alternative to the subjective actuarial judgment approach.

Special Report: Measurement-Based Care—Using Data to Augment Psychotherapy

Special Report: Measurement-Based Care—Using Data to Augment Psychotherapy

Stochastic loss reserving with mixture density neural networks

In recent years, new techniques based on artificial intelligence and machine learning in particular have been making a revolution in the work of actuaries, including in loss reserving. A particularly promising technique is that of neural networks, which have been shown to offer a versatile, flexible and accurate approach to loss reserving. However, applications of neural networks in loss reserving to date have been primarily focused on the (important) problem of fitting accurate central estimates of the outstanding claims. In practice, properties regarding the variability of outstanding claims are equally important (e.g., quantiles for regulatory purposes).In this paper we fill this gap by applying a Mixture Density Network (“MDN”) to loss reserving. The approach combines a neural network architecture with a mixture Gaussian distribution to achieve simultaneously an accurate central estimate along with flexible distributional choice. Model fitting is done using a rolling-origin approach. Our approach consistently outperforms the classical over-dispersed model both for central estimates and quantiles of interest, when applied to a wide range of simulated environments of various complexity and specifications.We further extend the MDN approach by proposing two extensions. Firstly, we present a hybrid GLM-MDN approach called “ResMDN“. This hybrid approach balances the tractability and ease of understanding of a traditional GLM model on one hand, with the additional accuracy and distributional flexibility provided by the MDN on the other. We show that it can successfully improve the errors of the baseline ccODP, although there is generally a loss of performance when compared to the MDN in the examples we considered. Secondly, we allow for explicit projection constraints, so that actuarial judgement can be directly incorporated into the modelling process.Throughout, we focus on aggregate loss triangles, and show that our methodologies are tractable, and that they out-perform traditional approaches even with relatively limited amounts of data. We use both simulated data—to validate properties, and real data—to illustrate and ascertain practicality of the approaches.

Longevity trend risk over limited time horizons

AbstractWe consider various aspects of longevity trend risk viewed through the prism of a finite time window. We show the broad equivalence of value-at-risk (VaR) capital requirements at a p-value of 99.5% to conditional tail expectations (CTEs) at 99%. We also show how deferred annuities have higher risk, which can require double the solvency capital of equivalently aged immediate anuities. However, results vary considerably with the choice of model and so longevity trend-risk capital can only be determined through consideration of multiple models to inform actuarial judgement. This model risk is even starker when trying to value longevity derivatives. We briefly discuss the importance of using smoothed models and describe two methods to considerably shorten VaR and CTE run times.

Fair dynamic valuation of insurance liabilities: Merging actuarial judgement with market- and time-consistency

In this paper, we investigate the fair valuation of insurance liabilities in a dynamic multi-period setting. We define a fair dynamic valuation as a valuation which is actuarial (mark-to-model for claims independent of financial market evolutions), market-consistent (mark-to-market for any hedgeable part of a claim) and time-consistent, extending the work of Dhaene et al. (2017) and Barigou and Dhaene (2019). We provide a complete hedging characterization for fair dynamic valuations. Moreover, we show how to implement fair dynamic valuations through a backward iterations scheme combining risk minimization methods from mathematical finance with standard actuarial techniques based on risk measures.

Surgical Intuition.

Surgical Intuition.

Fair Dynamic Valuation of Insurance Liabilities: Merging Actuarial Judgement With Market- and Time-Consistency

In this paper, we investigate the fair valuation of insurance liabilities in a dynamic multi-period setting. We define a fair dynamic valuation as a valuation which is actuarial (mark-to-model for claims independent of financial market evolutions), market-consistent (mark-to-market for any hedgeable part of a claim) and time-consistent, extending the work of Dhaene et al. (2017) and Barigou and Dhaene (2018). We provide a complete hedging characterization for fair dynamic valuations. Moreover, we show how to implement fair dynamic valuations through a backward iterations scheme combining risk minimization methods from mathematical finance with standard actuarial techniques based on risk measures.

Induced seismicity closed-form traffic light system for actuarial decision-making during deep fluid injections

The rise in the frequency of anthropogenic earthquakes due to deep fluid injections is posing serious economic, societal, and legal challenges to many geo-energy and waste-disposal projects. Existing tools to assess such problems are still inherently heuristic and mostly based on expert elicitation (so-called clinical judgment). We propose, as a complementary approach, an adaptive traffic light system (ATLS) that is function of a statistical model of induced seismicity. It offers an actuarial judgement of the risk, which is based on a mapping between earthquake magnitude and risk. Using data from six underground reservoir stimulation experiments, mostly from Enhanced Geothermal Systems, we illustrate how such a data-driven adaptive forecasting system could guarantee a risk-based safety target. The proposed model, which includes a linear relationship between seismicity rate and flow rate, as well as a normal diffusion process for post-injection, is first confirmed to be representative of the data. Being integrable, the model yields a closed-form ATLS solution that is both transparent and robust. Although simulations verify that the safety target is consistently ensured when the ATLS is applied, the model from which simulations are generated is validated on a limited dataset, hence still requiring further tests in additional fluid injection environments.

Fair valuation of insurance liabilities: Merging actuarial judgement and market-consistency

In this paper, we investigate the fair valuation of liabilities related to an insurance policy or portfolio in a single period framework. We define a fair valuation as a valuation which is both market-consistent (mark-to-market for any hedgeable part of a claim) and actuarial (mark-to-model for any claim that is independent of financial market evolutions). We introduce the class of hedge-based valuations, where in a first step of the valuation process, a ‘best hedge’ for the liability is set up, based on the traded assets in the market, while in a second step, the remaining part of the claim is valuated via an actuarial valuation. We also introduce the class of two-step valuations, the elements of which are very closely related to the two-step valuations which were introduced in Pelsser and Stadje (2014). We show that the classes of fair, hedge-based and two-step valuations are identical.

New all oral therapy for chronic hepatitis C virus (HCV): a novel long-term cost comparison.

BackgroundIn the US, the prevalence of hepatitis C virus (HCV) has surpassed the prevalence of human immunodeficiency virus (HIV), with about 3.3 million people chronically infected with the disease. Given the aging of the Baby Boomer generation and the subsequent implementation of age-based screening recommendations, HCV diagnoses are expected to increase. Utilization of anti-viral pharmacotherapy is also expected to increase as more effective and tolerable all-oral therapies for HCV become available in the United States. This research allows payors to assess the disease burden and treatment impact of HCV in their member group.MethodsA set of three integrated economic models was developed to estimate the disease and cost burden of HCV based on existing literature, wholesale acquisition costs, industry standards, and actuarial judgment. Model 1 estimates the HCV antibody prevalence of HCV in a payer’s member group based on population size and the age, sex, and region distribution of the members. Model 2 predicts the number of uncured chronic HCV members who represent the future treatment and medical cost burden for the payer over the next 14 years. Model 3 contrasts the pharmacy, medical, and overall costs for treatment and medical care over 14 years for three therapeutic scenarios: interferon-based standard of care (SOC), all oral therapy, and natural course of disease progression, while accounting for the frequency of HCV genotype within the member population.ResultsIn a payer population of 100,000 members with an age, sex, and region distribution matching the United States, the seroprevalence of HCV was estimated to be 1.26 %. Combined pharmacy and medical costs for uncured chronic HCV positive members was least expensive for all oral therapy. The per patient with HCV cost savings for all oral therapy compared to SOC were about $3000 per year over 14 years. In a sensitivity analysis, the 12-week all oral therapy for genotype 1 provided overall cost savings vs. a 24-week interferon-based SOC regimen until all oral therapy costs exceeded $99,000.ConclusionsIn most modeled scenarios, the all-oral therapeutic scenario was less costly than SOC, even in sensitivity analyses.

Developing Actuarial Judgement

Explicit actuarial capability frameworks are being developed to measure actuarial quality as a basis for pre- and post-qualification learning goals. These frameworks focus on the development of broader and normative capabilities that can be described as actuarial judgement. This paper explores different perspectives on how actuarial judgement might be identified and developed. It suggests that actuarial students need more help in developing personal and coherent frameworks with which to evaluate their decisions. The actuarial core technical syllabuses should be extended to include more ethical issues and a critique of positive economics. Ethical issues should include some understanding of the rationalizations that accompany unethical and corrupt behaviour. Committed teachers and mentors prepared to address ethical conflicts in practical cases play a role that needs greater emphasis.

Violence: To Predict in Order to Prevent?

Violence is a social problem. Violence is also a public health problem. However, violence can be “cut” several ways; its definition varies by type of violence. As such, the theories, which seek to understand violence, differ—as do the methods advocated for its eradication (or at least its reduction). These points could not be made any more evident by the reading of these two disparate books, one on violence prediction and one on its prevention. In this review each book is first discussed separately, and then directly compared. There remains no consensus on how best to predict violence. Yet mental health professionals are called upon to know the likelihood that violent offenders will commit a new violent offense. Violent Offenders strongly advocates for an actuarial assessment approach to predicting violence recidivism ranging from common assault to multiple murders. Empirical support is provided for such an approach with research findings available from the maximum security Oak Ridge Division of the Mental Health Centre, Penetonguishene. The book, a compilation of 35 years of work, is broken down into four parts. In the first section, the authors discuss history and background in order to point to the necessity of actuarial risk appraisal instruments. Actuarial instruments involve a limited set of static and historical factors (age, marital status, characteristics of the offense) as predictors of violence. After a review of research on three types of offenders (mentally disordered offender, sex offenders, and fire setters) in the second section, the authors describe their development of violence prediction instruments, which are further detailed in the appendices. Such development was in response to their realization that one tool applicable to all offenders for whom courts, clinicians, and parole officials had to make predictive decisions was needed. The final section deals with altering risk of violence and includes concluding comments. This second edition of Violent Offenders moves beyond the clinical recommendations from its first publication. Provocatively, the authors argue that incorporating actuarial methods into existing clinical practice is not enough. Here, they press for the total replacement of existing practice with these actuarial methods. This view challenges traditional clinical training approaches and indeed the reader may feel challenged. The book’s plea is stated in the context of the repeated argument that clinical judgment is overused and faulty, and that by mixing the two approaches (actuarial and clinical judgment) the prediction of violence is worse than with either one alone. This edition does provide an alternate place for clinical judgment by insisting that “circumstances that tempt one to adjust an actuarial score should instead be considered separately in adjusting supervision and providing interventions to reduce risk” (p. 198). This new volume also includes a revised scoring procedure for the actuarial assessment. The actuarial method is not without its critics. For instance, two arguments against these methods hinge on the approach’s exclusion of social forces outside of individuals’ control (e.g., poverty, discrimination) as predictors of Sex Roles (2007) 57:471–473 DOI 10.1007/s11199-007-9265-6

Letter

Back to table of contents Previous article Next article LetterFull AccessLetterGregory A. Aarons Ph.D.Gregory A. Aarons Ph.D.Search for more papers by this authorPublished Online:1 Nov 2006https://doi.org/10.1176/ps.2006.57.11.1656AboutSectionsPDF/EPUB ToolsAdd to favoritesDownload CitationsTrack Citations ShareShare onFacebookTwitterLinked InEmail In Reply: Mr. Seitzer's letter regarding my article supports the contention that high-quality leadership is an important consideration in the provision of high-quality mental health services. Establishing a link between leadership and the attitudes of mental health service providers toward adopting innovation and change—evidence-based practices in particular—was the main point of the article. The article does not, however, contend that evidence-based practices are the only means of improving outcomes for those involved in mental health services. There is considerable debate regarding applicability, standards for fidelity, and the need for local adaptation of evidence-based practices. The article takes no stand on the value of evidence-based practices per se. The Evidence-Based Practice Attitudes Scale ( 1 ), which was used in the study, was developed because providers' attitudes toward adopting innovation have been largely ignored in the research literature, even while the impetus to disseminate and implement evidence-based practices has gained momentum. Mr. Seitzer's assertion that outcomes matter more than particular practices raises a complex and important issue. Services might be greatly improved if outcomes were routinely empirically assessed in usual care. Rigorous outcome measurement and use of such outcome data to inform clinical practice are relatively rare. Indeed, research suggests that clinicians' attitudes about using such measures vary greatly, with some believing that clinical knowledge and judgment are at least as accurate as empirically reliable and valid assessment, thus obviating the need for such measures ( 2 ). There is evidence that draws such an assumption into question ( 3 ). Assessments will likely be beneficial to the degree that they can be incorporated into the clinical process both to better understand initial diagnoses and problems and to track clinical progress. Although my study did not directly address outcome assessment, it did establish a link between leadership and providers' attitudes toward adopting evidence-based practices as part of a research program on improving the context for delivery of high-quality mental health services and implementation of innovation ( 4 ). An innovation in mental health care may or may not be evidence based. What is emerging from this line of research is that leadership and organizational context are likely to play an important role in implementing innovation and promoting more positive outcomes for mental health and social service agencies and clinics ( 5 ). Reference1. Aarons GA: Mental health provider attitudes toward adoption of evidence-based practice: the Evidence-Based Practice Attitude Scale (EBPAS). Mental Health Services Research 6:61-74, 2004Google Scholar2. Garland AF, Kruse M, Aarons GA: Clinicians and outcome measurement: what's the use? Journal of Behavioral Health Services and Research 30:393-405, 2003Google Scholar3. Dawes RM, Faust D, Meehl PE: Clinical versus actuarial judgment. Science 243:1668-1674, 1989Google Scholar4. Aarons GA: Measuring provider attitudes toward evidence-based practice: consideration of organizational context and individual differences. Child and Adolescent Psychiatric Clinics of North America 14:255-271, 2005Google Scholar5. Glisson C: The organizational context of children's mental health services. Clinical Child and Family Psychology Review 5:233-253, 2002Google Scholar FiguresReferencesCited byDetailsCited byNone Volume 57Issue 11 November, 2006Pages 1656-1656PSYCHIATRIC SERVICES November 2006 Volume 57 Number 11 Metrics PDF download History Published online 1 November 2006 Published in print 1 November 2006

The Impact of States' Small Group Health Insurance Reforms on Uninsurance Rates

The impact of states' small group market reforms on uninsurance rates was examined. Reform status was quantified on five reform dimensions: Access Improvement, Premium Reduction, Premium Differential Reduction, Continuity of Coverage, and Enhancing Valued Plan Features. These reform indices were calculated based on actuarial judgment of the market impact potential of each regulation. Regression analysis showed no association between uninsurance rates and the depth of reforms on any dimension, while controlling for income, foreign-born population, black population, and employment in the smallest businesses. Possible reasons for the lack of impact are discussed.

Assessing the Likelihood of an Important Clinical Outcome: New Insights from a Comparison of Clinical and Actuarial Judgment

To assess and rank the performance of different methods of predicting the probability of death following a specified surgical procedure. Actuarial estimates of the probability of early mortality for 40 patients were derived from 2 sources: a large published surgical series and a smaller series from the center where surgery was performed. Surgeons and cardiologists also provided probability estimates for these patients. Estimates derived from the published literature were too optimistic and did not differentiate between patients more, or less, likely to die (i.e., failed to discriminate). Doctors' judgments were unbiased but failed to discriminate. Local actuarial estimates (influenced by only 1 or 2 variables) were unbiased, did discriminate, but exhibited more random variation. The preferred source of estimates depends upon which aspect of accuracy is of greatest importance. Differences in patient selection and error in the identification of risk factors mean that published results will not always appropriately predict surgical risk at other institutions. Risk stratification may be more robust when based on a small set of cross-validated predictors rather than a larger set of predictors that includes some whose reliability has not been confirmed.

The recalcitrance of overconfidence and its contribution to decision aid neglect

AbstractThree experiments tested the hypothesis that people's overconfidence in the quality of their intuitive judgment strategies contributes to their reluctance to use helpful actuarial judgment aids. Participants engaged in a judgment task that required them to use five cues to decide whether a prospective juror favored physician‐assisted suicide. Participants had the opportunity to examine the judgments of a statistical equation that correctly classified 77% of the prospective jurors. In all experiments, participants infrequently examined the equation, performed worse than the equation, and were highly overconfident. In Experiments 1 and 2, outcome feedback and calibration feedback failed to reduce overconfidence. In Experiment 3, enhanced calibration feedback reduced overconfidence and increased reliance on the equation, thus leading to improved judgment performance. Copyright © 2005 John Wiley & Sons, Ltd.

Small Group Health Insurance

States are ranked based on the potential of their small group health insurance reforms to enhance health insurance uptake. Reforms were quantified based on their market impact potential. Five dimensions of reforms were identified, Access Improvement, Premium Reduction, Premium Differential Reduction, Continuity of Coverage, and Enhancing Valued Plan Features. The reform indices representing these dimensions were developed based on document review of state statutes, combined with actuarial judgment to identify and assign scores to market-relevant regulations in line with their impact potential. The distribution of the states' reform scores and rankings show wide variation in the depth and focus of their reforms. Only seven of the top ten states on the Total Reform index had consistently higher scores on two or more reform dimensions. The conceptual linkages between specific regulations and the documented small group market problems lead to normative expectations of an association between the depth of state reforms and the prevalence of uninsurance.

Statistical significance testing, construct validity, and clinical versus actuarial judgment: an interesting (seeming) paradox

What practical implications does Meehl’s attack on significance testing have for clinical psychologists, and doesn’t Meehl contradict himself in placing so much emphasis on scientific theories in such works as “Theoretical Risks and Tabular Asterisks” and yet advocating for actuarial judgement, which is essentially an atheoretical approach to clinical predication? This seemingly paradoxical ideology, when considered within the broader context of Meehl’s work, can be readily aligned.

A Stochastic Approach to Risk Management and Decision Making in Defined Benefit Pension Schemes

ABSTRACTThe trustees and sponsors of defined benefit schemes rely on the advice of the Scheme Actuary to make important decisions concerning the funding of the scheme, the investment of its assets, and the use of surplus assets to improve benefits. These decisions have to be made in the face of considerable uncertainty about financial and demographic factors that will affect the future experience of the scheme and its success in meeting various objectives.The traditional actuarial valuation combined with actuarial judgement has played an important role in guiding decision making; but we argue that stochastic methods can add value in certain crucial areas, in particular the financial risk management of defined benefit schemes. Rather than dealing with risk by incorporating margins in the valuation basis, a stochastic approach allows the actuary to evaluate specific and quantifiable risk and performance measures for alternative funding and investment strategies.This paper recommends a framework that, when combined with a suitable stochastic model, measures the risks inherent in contribution rate and asset allocation decisions, allowing better decisions to be made. In doing this, we suggest and apply various risk and performance measures that may be thought appropriate, although our intention is to illustrate their use rather than prescribe them as objective standards. The framework provides the means to explore the trade-offs involved in possible contribution and asset allocation decisions, and points to decision strategies expected to give improved outcomes for the same level of risk. A feature of the approach that marks it out from current asset/liability techniques is that it examines the funding and investment decisions together. It does not derive a contribution rate in the traditional way, but leaves this as free variable, in the same way that the investment decision is taken to be a free variable. Another distinctive feature of our framework is that it is based on projection rather than on valuation, involving stochastic simulation of the experience of the scheme over a time horizon reflecting the concerns of the trustees and the sponsoring employer.The paper provides a case study (based on a model final salary pension scheme) showing the advantages of the framework, and goes on to explain how the results may practically be communicated to trustees and scheme sponsors.

Specificity of Malingering Detection Strategies in Older Adults Using the CVLT and WCST

Strategies for the detection of possible malingering have largely taken two forms. First is the development and validation of domain-specific measures of malingering designed specifically for the detection of malingering (e.g., Test Of Memory Malingering, Word Memory Test, and Word Completion Memory Test). The second has been the development and evaluation of performance patterns and cutoff scores applicable to commonly used neuropsychological instruments. Two such instruments that have been examined are the Wisconsin Card Sorting Test (WCST) and the California Verbal Learning Test (CVLT). While several studies propose possible cutoff scores and actuarial judgement strategies for these tests, the specificity of these strategies within older, nonimpaired adults has not been established. Without proper evaluation of potential age-related effects within these strategies, the actual utility of the strategies within suspected malingerers who are older adults is unknown. Therefore, the present study was conducted to evaluate the specificity of the proposed strategies for detecting possible malingering with the WCST and the CVLT in a sample of older, community-dwelling, nonimpaired adults. The results suggest that the currently existing WCST formulas may have limited utility for the detection of malingering with older adults while the CVLT strategies do appear to have potential clinical utility. The potential of these formulas for use with older adults is discussed.