The Demand for Life Insurance: Dynamic Ecological Systemic Theory Using Machine Learning Techniques
By Wookjae Heo
()
About this ebook
This book, adopting machine learning techniques for the financial planning field, explores the demand for life insurance as seen in previous literature and both estimates and predicts the demand for the adoption of life insurance using these techniques. Previous studies used diverse perspectives, like actuarial and life span, in order to understand the demand for life insurance, though these approaches have shown inconsistent findings. Employing two theoretical backgrounds—ecological systemic theory and artificial intellectual methodology—this book explores a better estimation and a prediction of the demand for life insurance and will be of interest to academics and students of insurance, financial planning, and risk management.
Related to The Demand for Life Insurance
Related ebooks
Systems Biomedicine: Concepts and Perspectives Rating: 0 out of 5 stars0 ratingsClinical Prediction Models: A Practical Approach to Development, Validation, and Updating Rating: 0 out of 5 stars0 ratingsApplied Survival Analysis: Regression Modeling of Time-to-Event Data Rating: 4 out of 5 stars4/5Clinical Decision Support Systems: Theory and Practice Rating: 3 out of 5 stars3/5Data and the American Dream: Contemporary Social Controversies and the American Community Survey Rating: 0 out of 5 stars0 ratingsMultiple Imputation and its Application Rating: 0 out of 5 stars0 ratingsMass-Action Law Dynamics Theory and Algorithm for Translational and Precision Medicine Informatics Rating: 0 out of 5 stars0 ratingsPrinciples of Biomedical Informatics Rating: 0 out of 5 stars0 ratingsCommon Errors in Statistics (and How to Avoid Them) Rating: 0 out of 5 stars0 ratingsThe Skeffington Perspective of the Behavioral Model of Optometric Data Analysis and Vision Care Rating: 0 out of 5 stars0 ratingsWorking with Dynamic Crop Models: Methods, Tools and Examples for Agriculture and Environment Rating: 0 out of 5 stars0 ratingsPractical Biostatistics: A Friendly Step-by-Step Approach for Evidence-based Medicine Rating: 5 out of 5 stars5/5Statistical Bioinformatics: For Biomedical and Life Science Researchers Rating: 0 out of 5 stars0 ratingsAnalyzing the Large Number of Variables in Biomedical and Satellite Imagery Rating: 0 out of 5 stars0 ratingsIntroduction to Biostatistics with JMP (Hardcover edition) Rating: 1 out of 5 stars1/5Linear and Generalized Linear Mixed Models and Their Applications Rating: 0 out of 5 stars0 ratingsStatistical Methods for Overdispersed Count Data Rating: 0 out of 5 stars0 ratingsBayesian Population Analysis using WinBUGS: A Hierarchical Perspective Rating: 3 out of 5 stars3/5Systems Design for Remote Healthcare Rating: 0 out of 5 stars0 ratingsIntegrated Population Models: Theory and Ecological Applications with R and JAGS Rating: 0 out of 5 stars0 ratingsMethods and Applications of Sample Size Calculation and Recalculation in Clinical Trials Rating: 0 out of 5 stars0 ratingsComputational and Statistical Methods for Analysing Big Data with Applications Rating: 0 out of 5 stars0 ratingsNeural Networks in Finance: Gaining Predictive Edge in the Market Rating: 3 out of 5 stars3/5Computational Modeling in Biomedical Engineering and Medical Physics Rating: 0 out of 5 stars0 ratingsTheories to Inform Superior Health Informatics Research and Practice Rating: 0 out of 5 stars0 ratingsQuantitative Decisions in Drug Development Rating: 0 out of 5 stars0 ratingsQuantitative Analysis and Modeling of Earth and Environmental Data: Space-Time and Spacetime Data Considerations Rating: 0 out of 5 stars0 ratingsCollaborative Genomics Projects: A Comprehensive Guide Rating: 0 out of 5 stars0 ratingsAnalytics in Healthcare: A Practical Introduction Rating: 0 out of 5 stars0 ratingsAnalysis of Clinical Trials Using SAS: A Practical Guide, Second Edition Rating: 0 out of 5 stars0 ratings
Industries For You
All the Beauty in the World: The Metropolitan Museum of Art and Me Rating: 4 out of 5 stars4/5Uncanny Valley: A Memoir Rating: 4 out of 5 stars4/5Excellence Wins: A No-Nonsense Guide to Becoming the Best in a World of Compromise Rating: 5 out of 5 stars5/5YouTube Secrets: The Ultimate Guide to Growing Your Following and Making Money as a Video I Rating: 5 out of 5 stars5/5Weird Things Customers Say in Bookstores Rating: 5 out of 5 stars5/5YouTube 101: The Ultimate Guide to Start a Successful YouTube channel Rating: 5 out of 5 stars5/5Grocery: The Buying and Selling of Food in America Rating: 4 out of 5 stars4/5Music Law: How to Run Your Band's Business Rating: 0 out of 5 stars0 ratingsBad Pharma: How Drug Companies Mislead Doctors and Harm Patients Rating: 4 out of 5 stars4/5Pharma: Greed, Lies, and the Poisoning of America Rating: 5 out of 5 stars5/5Becoming Trader Joe: How I Did Business My Way and Still Beat the Big Guys Rating: 5 out of 5 stars5/5The Best Story Wins: How to Leverage Hollywood Storytelling in Business & Beyond Rating: 5 out of 5 stars5/5INSPIRED: How to Create Tech Products Customers Love Rating: 5 out of 5 stars5/5Setting the Table: The Transforming Power of Hospitality in Business Rating: 5 out of 5 stars5/5Fast Food Nation: The Dark Side of the All-American Meal Rating: 0 out of 5 stars0 ratingsSweet Success: A Simple Recipe to Turn your Passion into Profit Rating: 5 out of 5 stars5/5Artpreneur: The Step-by-Step Guide to Making a Sustainable Living From Your Creativity Rating: 2 out of 5 stars2/5Shopify For Dummies Rating: 0 out of 5 stars0 ratingsThe House of Gucci: A True Story of Murder, Madness, Glamour, and Greed Rating: 4 out of 5 stars4/5How We Do Harm: A Doctor Breaks Ranks About Being Sick in America Rating: 4 out of 5 stars4/5Bottle of Lies: The Inside Story of the Generic Drug Boom Rating: 4 out of 5 stars4/5Disney's Land: Walt Disney and the Invention of the Amusement Park That Changed the World Rating: 4 out of 5 stars4/5
Reviews for The Demand for Life Insurance
0 ratings0 reviews
Book preview
The Demand for Life Insurance - Wookjae Heo
© The Author(s) 2020
W. HeoThe Demand for Life Insurancehttps://doi.org/10.1007/978-3-030-36903-3_1
1. Introduction: A Need of New Framework in Financial Planning with the Case of Life Insurance Demand
Wookjae Heo¹
(1)
South Dakota State University, Brookings, SD, USA
Wookjae Heo
Email: Wookjae.heo@sdstate.edu
Abstract
This chapter is a general introduction to the issue of understanding life insurance and demand for it in the market. Unlike the actuarial science and lifespan-related economic approaches, this book was based on a different analytical framework to understand the demand for life insurance. By using a machine learning technique and the complexities among influential determinant factors, it may be possible to predict the demand for life insurance more accurately. To justify the usage of machine learning in financial planning, a new theoretical framework is briefly introduced in this chapter.
Keywords
Life insuranceDemand for life insuranceComplexity in insurance marketDynamic nonlinear systemic approach
1.1 Introduction and Statement of the Problem
The life insurance market is large, dynamic, and somewhat fragmented. Consider the following facts from the Insurance Information Institute (2015) and LIMRA (2014):
Life insurance has been sold in the United States for over 200 years;
Consumers paid nearly $164 billion in life insurance premiums in 2013;
The five largest insurers (i.e., MetLife, Prudential, New York Life, TIAA-CREF, and Northwestern Mutual) each have revenues that exceed $24,000 million;
Only 49% of consumers age 25–64 own individual life insurance;
Fifty percent of the adult US population say they need life insurance;
Nearly 90% of consumers believe life insurance is too expensive to purchase;
Approximately 10% of the US population plans to purchase a life insurance policy within the next year; and
Forty percent of consumers report that a life event (e.g., death of family member or close friend, getting married or divorced, etc.) prompted them to purchase life insurance.
These insurance details highlight the diverse nature of the life insurance marketplace. On the one hand, the insurance industry has a long history in the United States and is today a multi-billion dollar business. On the other hand, life insurance, as an important financial planning product, has limited market penetration. This helps explain why much of the existing life insurance research has been devoted to understanding the pricing mechanisms of life insurance and sales delivery and purchasing trends among consumers. The life insurance industry has been preoccupied with identifying pricing strategies that will attract new consumers to the insurance marketplace. A quick glance at the insurance industry’s leading Web sites and trade publications shows that industry groups spend a great deal of time, effort, and resources attempting to document life insurance ownership patterns and possibilities.
As will be discussed in this chapter, nearly all previous studies that have focused on the demand for life insurance have been conceptualized from a supply-side perspective. This study attempts to reframe the discussion of life insurance demand by conceptualizing the demand for life insurance as being shaped by many interrelated factors. Further, this study is premised on the notion that these factors can best be modeled using advanced statistical techniques that rely on artificial neural network techniques. This chapter provides a broad overview of the models most often used to predict and explain demand for life insurance among consumers, the factors most often hypothesized to be useful in predicting demand, and the conceptualization of this research. Later chapters provide a review of the relevant literature, a discussion of the research methodology, results from the statistical analyses, and an applied discussion.
It is important to start any discussion about life insurance demand by clarifying the role of life insurance within a consumer’s financial plan. Essentially, life insurance serves as a precaution for an unexpected event like the premature death of a family’s breadwinner, providing protection for the loss of a member of a household is the main reason people should, and most often do, purchase life insurance (Rejda, 2008; Thoyts, 2010). Losing a household’s breadwinner critically increases the financial vulnerability of a household, since the major source of income disappears with the death of the breadwinner. Even if the person who dies is not the household’s breadwinner, the loss of a related household member generates various economic burdens on the other members in the household. For instance, the premature death of a household member could leave unpaid medical bills and possible debt to be paid by others in the household (Rejda, 2008). These examples illustrate how life insurance works as a tool for managing the risk of premature death in a household.
When attempting to understand how life insurance works as a financial buffer against financial disaster, it is important to predict who will be more likely to purchase life insurance and the factors that lead people to purchase life insurance. Specifically, financial planners provide consulting advice on the purchase of appropriate life insurance policies based, in part, on their clients’ socioeconomic situation. Financial planners need valid, reliable tools to predict which people would like to purchase life insurance and the kinds of factors that are associated with purchasing behavior. In addition, policy makers have a need to understand the reason people purchase insurance and the manner in which people decide to purchase. This interest is premised on the notion that political issues, such as Social Security funding, are strongly related to the purchase of life insurance (Black & Skipper, 2000). Policy makers need to enforce market controls to help the market work efficiently and effectively. Furthermore, educators and researchers need information about consumers and purchase decision factors because predicting the demand for life insurance is one possible way to increase consumers’ financial well-being (Lynch, 2010).
There have been two major approaches used by researchers to understand the demand for life insurance: (a) actuarial science and (b) lifespan-related economic perspectives (e.g., human capital theory and the life cycle hypothesis). Actuarial science and lifespan-related economics are well developed and easily adapted for the use in understanding and explaining the life insurance market. For instance, these two major approaches explain well the price of life insurance and the quantity of selling in the real market. However, these approaches have serious limitations. These approaches tend to focus on finding and explaining the macro-level equilibrium in the real market rather than identifying and predicting influential factors in the decision-making process at the household level. In other words, actuarial science and lifespan-related economics are premised on different analytical purposes to estimate the equilibrium between demand and supply. These analytical approaches are not as efficient when predicting influential demand factors (i.e., which people tend to purchase life insurance and the factors associated with life insurance purchase decisions).
Specifically, actuarial life insurance theories, like financial models, mortality models, net single premium functions, and multiple discretion models, use limited specific factors to estimate the demand for life insurance. Typical inputs include factors such as present value, longevity, death rates, and cancellation ratios. Since actuarial life insurance theories were developed from an industrial perspective (i.e., supply-side), actuarial models focus on appropriate premium rates for life insurance. It is worth exploring this approach. The actuarial approach tends to rely on one of the following functions: (a) present value of time capital from financial models, (b) survivor function, and (c) net single premium function. The formula for each method is shown below:
$${\text{AC}} = c_{1} v^{t1} + \, c_{2} v^{t2} + \cdots + c_{k} v^{tk}$$(1.1)
$$s\left( \chi \right) = 1{-}F_{X} \left( \chi \right) = { \Pr }\left[ {X > \chi } \right]$$(1.2)
$$E\left[ Z \right] = \int\limits_{0}^{\infty } {b_{t} v_{t}^{t} p_{\chi } \mu_{\chi } \left( t \right){\text{dt}}}$$(1.3)
For function (1.1), c denotes the present value, v means the discount factor associated with time (t), and t is the time period. The present value for time capital (AC) is the same as the sum of the time series’ values. For function (1.2), X denotes a distribution of lifetimes among the population, where χ means people who survive at the age of χ. The survivor function indicates the probability a person survives until the age of χ. For function (1.3), v is the discount factor associated with time (t), b is the insurance benefit, pχ is the surviving probability until the age of χ, and μχ is the mortality probability at the age of χ. Expected net single premium (E[Z]) is the integration of insurance benefits, surviving and mortality probabilities, and the discount factor. As shown in these three functions, actuarial science is focused on projecting longevity and associated costs. These are appropriate tools to estimate the price of life insurance. However, from the consumer’s perspective, the actuarial science method is not an efficient tool when answering the following two questions: (a) Which people tend to purchase life insurance and (b) what factors are associated with the decision to purchase life insurance?
For the case of the lifespan-related economic perspective (e.g., life cycle model and human capital theory), economic factors like price and quantity of life insurance in the market are considered in the demand function and utility function, respectively. Since the lifespan-related economic perspective focuses on the utilities created in a market, optimal price and quantity are the main outcomes associated with economic life insurance research. The following functions represent the life cycle hypothesis and human capital model equation:
$${\text{C}} = {\text{aW}} + {\text{bY}}$$(1.4)
$${\text{P}} = {\text{MP}} + {\text{G}} = {\text{W}} + {\text{C}} = \pi$$(1.5)
For the life cycle hypothesis function (1.4), consumption (C) is the same as the sum of marginal propensity to consume (a) for wealth (W) and the marginal propensity to consume (b) for income (Y). For the human capital model function (1.5), P denotes total labor productivity, MP denotes marginal productivity, G denotes excess of future receipts, W denotes wage, C denotes sum of opportunity costs for training or education, and π denotes the total wage. Equation (1.5) shows that if there is education or training for a worker (P = MP + G), the effect of education or training will generate a return as more productivity and total wage (W + C = π). Similar to the actuarial science approach, classical economic approaches can be used to understand the macro-level equilibrium in a market. However, from the perspective of consumer behavior, economic approaches are less than optimal in showing which people tend to purchase life insurance and the factors that are associated with the purchase of life insurance.
Although these two perspectives focus on finding a market-oriented equilibrium, many researchers have attempted to investigate the determinants of life insurance demand based upon these two perspectives (e.g., Anderson & Nevin, 1975; Babbel, 1985; Berekson, 1972; Browne & Kim, 1993; Burnett & Palmer, 1984; Chen, Wong, & Lee, 2001; Duker, 1969; Ferber & Lee, 1980; Fitzgerald, 1987; Gandolfi & Miners, 1996; Gutter & Hatcher, 2008; Hammond, Houston, & Melander, 1967; Hau, 2000; Lewis, 1989; Liebenberg, Carson, & Dumm, 2012; Mantis & Farmer, 1968; Showers & Shotick, 1994; Williams, 1986; Zietz, 2003). Using these empirical and practical approaches, researchers have noted diverse possible determinant factors, such as age, gender, family structure, and wealth on the demand for life insurance. Results from the literature illustrate that there are a number of diverse determinants associated with the demand for life insurance.
However, there is still a limitation associated with the common approaches used in most widely used empirical models; namely, there has been no consistent, uniform specific framework for selecting predictor variables. With the lack of a consistent framework, models of the determinant factors change with each empirical study. Zietz (2003) noted that this can create some confusing and controversial results. Table 1.1 shows how previous findings from the literature tend to be contradictory and at times opposite of traditional assumptions.
Table 1.1
Controversial research findings related to life insurance demand
Note Table adapted from An examination of the demand for life insurance,
by E. N. Zietz, 2003, Risk management and insurance review, 6, pp. 159–191. Copyright 2003 by Wiley
As shown in Table 1.1, many variables exhibit contradictory and controversial associations with the demand for life insurance. Some variables (e.g., education, income, and inflation/interest rates) show bipolar directional associations (i.e., positive and negative association) with life insurance demand. A few other variables, including wife working status and family size/family structure, show one of three possible results: positive association, negative association, and non-significant association. Other variables, such as life expectancy, marital status, religion, net worth, and employment, rarely show consistent results.
Besides the associations shown in Table 1.1, the relationships among other variables and the demand for life insurance often lead to conflicting findings. For instance, Showers and Shotick (1994) found age had a positive association with the demand for life insurance. However, Chen et al. (2001) found age was negatively related to the demand for life insurance. It is possible to assume, because of many factor changes, covariance and interaction terms among selected determinant factors become unstable depending on the research model. Therefore, previous research findings on the determinant factors of life insurance should be expected to show inconsistent results. These inconsistencies restrict more accurate predictions of the demand for life insurance.
In order to better predict the demand for life insurance, it is necessary to adopt a more robust, consistent, and conceptual methodological framework that is able to combine a solid theoretical background as well as empirically determined features. This means moving beyond traditional supply-side frameworks to ones that take into account the interrelated nature of demand variables. Based on an understanding of the two traditional classical perspectives and empirical applications, this study will explore a new framework to explain which people tend to purchase life insurance and why people purchase life insurance. The framework is based on a dynamic nonlinear systemic approach, which will be discussed later in this chapter. Within this dynamic, nonlinear, and systemic approach, many determinants for life insurance demand are assumed to be interdependently correlated. The outcomes from this study will provide a better understanding of the factors associated with life insurance demand from a consumer’s perspective. This study will ultimately provide a higher prediction level for the demand for life insurance.
1.2 Purpose and Justification of Study
Compared to the actuarial science and lifespan-related economics approaches, this study was based on a different