- Original article
- Open Access
Model confidence sets and forecast combination: an application to age-specific mortality
- Han Lin Shang^{1}Email authorView ORCID ID profile and
- Steven Haberman^{2}
https://doi.org/10.1186/s41118-018-0043-9
© The Author(s) 2018
- Received: 27 March 2018
- Accepted: 9 October 2018
- Published: 21 November 2018
Abstract
Background
Model averaging combines forecasts obtained from a range of models, and it often produces more accurate forecasts than a forecast from a single model.
Objective
The crucial part of forecast accuracy improvement in using the model averaging lies in the determination of optimal weights from a finite sample. If the weights are selected sub-optimally, this can affect the accuracy of the model-averaged forecasts. Instead of choosing the optimal weights, we consider trimming a set of models before equally averaging forecasts from the selected superior models. Motivated by Hansen et al. (Econometrica 79(2):453–497, 2011), we apply and evaluate the model confidence set procedure when combining mortality forecasts.
Data and methods
The proposed model averaging procedure is motivated by Samuels and Sekkel (International Journal of Forecasting 33(1):48–60, 2017) based on the concept of model confidence sets as proposed by Hansen et al. (Econometrica 79(2):453–497, 2011) that incorporates the statistical significance of the forecasting performance. As the model confidence level increases, the set of superior models generally decreases. The proposed model averaging procedure is demonstrated via national and sub-national Japanese mortality for retirement ages between 60 and 100+.
Results
Illustrated by national and sub-national Japanese mortality for ages between 60 and 100+, the proposed model-averaged procedure gives the smallest interval forecast errors, especially for males.
Conclusion
We find that robust out-of-sample point and interval forecasts may be obtained from the trimming method. By robust, we mean robustness against model misspecification.
Keywords
- Equal predictability test
- Japanese human mortality database
- Mean interval score
- Model averaging
- Root mean square forecast error
Introduction
Because of declining mortality rates in mainly developed countries, the improvement in human survival probability contributes significantly to an aging population. As a consequence, pension funds and insurance companies face longevity risk. The longevity risk is a potential risk attached to the increasing life expectancy of policyholders, which can eventually result in a higher payout ratio than expected (Crawford et al. 2008). The concerns about longevity risk have led to a surge in interest among pension funds and insurance companies in accurately modeling and forecasting age-specific mortality rates (or death counts or survival probabilities). Any improvement in the forecast accuracy of mortality rates will be beneficial for determining the allocation of current and future resources at the national and sub-national levels (see, e.g., Koissia (2006); Denuit et al. (2007); Hanewald et al. (2011)).
Many different models for forecasting age-specific mortality rates have been proposed in the literature (see Booth (2006); Booth and Tickle (2008); Currie et al. (2004); Girosi and King (2008); Shang et al. (2011); Tickle and Booth (2014), for reviews). Of these, a significant milestone in demographic modeling and forecasting was the work by Lee and Carter (1992). They implemented a principal component method to model age-specific mortality and extracted a single time-varying index representing the trend in the level of mortality rates, from which the forecasts are obtained by a random walk with drift. While the Lee-Carter method is simple and robust in situations where age-specific log mortality rates have linear trends (Booth 2006), it has the limitation of attempting to capture the patterns of mortality rates using only one principal component and its associated scores. To rectify this deficiency, the Lee-Carter method has been extended and modified. For example, from a discrete data matrix perspective, Booth et al. (2002), Renshaw and Haberman (2003a), and Cairns et al. (2006); Cairns et al. (2009) proposed the use of more than one component in the Lee-Carter method to model age-specific mortality rates; Renshaw and Haberman (2006) proposed an age-period-cohort extension to the Lee-Carter model under a Poisson error structure, while Plat (2009b) extended this model by incorporating the dependence between ages. Cairns et al. (2006) used a logistic transformation to model the relationship between the death probability and age observed over time, while Cairns et al. (2009) extended this model by incorporating the cohort effect. Girosi and King (2008) and Wiśniowski et al. (2015) considered a Bayesian paradigm for the Lee-Carter model estimation and forecasting. Hatzopoulos and Haberman (2009) followed a generalized linear model approach which leads to models that have a similar structure to the Lee-Carter model but with a generalized error structure. Hunt and Blake (2014) presented a general structural form of mortality models. From a continuous function perspective, Hyndman and Ullah (2007) proposed a functional data model that utilizes nonparametric smoothing and higher-order principal components, while Shang et al. (2011) and Shang (2016) considered a multilevel functional data model to model mortality rates jointly for multiple populations.
There exist many papers on comparing the forecast accuracy among several mortality forecasting methods. However, the most accurate forecasting method has been determined based on an aggregate loss function. Instead of identifying the most accurate method, Shang (2012) considered a model averaging approach to combine forecasts from a range of methods, such as the Lee-Carter and functional time-series methods. In this paper, we propose a new model averaging approach motivated by Samuels and Sekkel (2017). The proposed model averaging method uses the model confidence set procedure to select a set of superior models and combines the forecasts by assigning equal weights to the set of superior models. In contrast with Shang (2012), the problem is centered not on the selection of optimal weights, but on the selection of superior models. In the “A competing model averaging method” section, we compare the forecast accuracy between the existing and proposed model average methods.
With the aim of evaluating and comparing the forecast accuracy of different forecasting methods, forecast competition has a long history. The “M” competition originated from Makridakis et al. (1979) was the first attempt at a large empirical comparison of forecasting methods. In that “M” competition, there were 1001 time series for which participants were invited to submit their forecasts. Later, the results were published in Makridakis et al. (1982). The “M” competition has progressed slowly over the years, with the most recent M4 competition taken place in 2018. The top-performing teams all combine forecasts from a range of statistical and machine learning methods via some model averaging to improve the point and interval forecast accuracies.
Despite the popularity of model averaging in statistical and forecasting literature (see, e.g., Bates and Granger (1969); Dickinson (1975); Clemen (1989)), model averaging has not received increasing attention in the demographic literature with the noticeable exceptions of Shang (2012); Shang (2015) in the context of mortality forecasting, ((Bijak 2011), Chapter 5) in the context of migration forecasting, and Abel et al. (2013) and Shang et al. (2014) for the overall population growth rate. Shang (2012) revisited many statistical methods and combined their forecasts based on two weighting schemes, one of which has been adapted for comparison in the “A competing model averaging method” section. Both weighting schemes determine the weights by using either in-sample forecast accuracy or in-sample goodness-of-fit. Because of the finite sample, the weight assigned to the worst model is often small but not zero. In turn, this may lead to inferior forecast accuracy than the one based on oracle weights. This motivates us to consider an alternative model averaging idea. Instead of assigning weights to the forecasts from all models, we trim out the worse performing models based on a statistical significance test, such as the model confidence set procedure of Hansen et al. (2011) (see also Samuels and Sekkel (2017)).
While most attention has been paid to selecting model combination weights see, (e.g., Fischer and Harvey (1999); Genre et al. (2013)), Aiolfi et al. (2010) pointed out that there has been little research focusing on which models to include in the model pool. Graefe (2015) asserted that a simple average of the forecasts produced by individual models is a benchmark and commonly outperforms more complicated weighting schemes that rely on the estimation of theoretically optimal weights. The simple average of the forecasts performs well when the model fits the data poorly: when the sample number per predictor is low and when the predictors are highly correlated (Graefe 2015).
This paper focuses on a statistical significance approach to select the models to be included in the forecast combination. Via the model confidence set procedure of Hansen et al. (2011), we determine a set of statistically superior models, conditional on the model’s in-sample performance for forecasting age-specific mortality rates. By equally averaging the forecasts from the superior models, we evaluate and compare point and interval forecast accuracies, as measured by the root mean square forecast error and mean interval score, respectively.
The outline of this paper is described as follows: From the Japanese national and sub-national age-specific mortality data in the “Japanese national and sub-national age-specific mortality data” section, we first visualize the heterogeneity in age-specific mortality rates among 47 prefectures (Additional file 1). Then, we revisit some commonly used multivariate and functional time-series extrapolation methods for forecasting age-specific mortality rates in the “Time-series extrapolation models” section. Using the model confidence set procedure of (Hansen et al. 2011) described in the “Model confidence set” section, we select a set of superior models based on their point or interval forecast accuracy and demonstrate the robust accuracy of the proposed model averaging method in the “Forecast results” section. In the “A competing model averaging method” section, we present an adaption of an existing model averaging method where optimal weights are estimated based on in-sample forecast accuracy and assigned to the forecasts from all models. In the “Discussion” section, we conclude and outline how the methodology presented here can be further extended.
Japanese national and sub-national age-specific mortality data
We study the Japanese age-specific mortality rates from 1975 to 2015, obtained from the Japanese Mortality Database (2017). We consider ages from 60 to 99 in single years of age, while the last age group contains all ages at and beyond 100 (abbreviated as 100+). We consider modeling mortality at older ages, as the mortality forecasts are an important input for calculating annuity prices for retirees and the corresponding reserves held by insurance companies and pension funds. Some of the models considered were designed for modeling mortality at older ages, such as the Cairns-Blake-Dowd models.
We split the Japanese mortality rates by sex and prefecture. We are also interested in the mortality data at the sub-national (i.e., prefecture) level. The mortality forecasts at the prefecture level are more useful than the mortality forecasts at the national level for local policy making and planning.
In the supplement, we have plotted the geographic locations (from North to South) of the 47 prefectures within eight regions of Japan (Additional file 2). Also, we present the names of prefecture within each of the eight regions of Japan. Shang and Haberman (2017) and Shang and Hyndman (2017) present plots of the ratio of mortality between each prefecture and Japan by age or year.
Time-series extrapolation models
We study some time-series extrapolation methods for modeling and forecasting age-specific mortality rates. The models that we have considered are subjective and far from extensive, but they suffice to serve as a test bed for demonstrating the performance of forecast combination. From actuarial science, we consider a family of Renshaw-Haberman (RH) models (see, e.g., Renshaw and Haberman (2003a, b, 2006, 2008); Haberman and Renshaw (2008, 2009) and a family of Cairns-Blake-Dowd (CBD) models (Cairns et al. 2006). These two models perform well for mortality at higher ages, such as between 60 and 100+. From demography, we consider a family of Lee-Carter (LC) models (see, e.g., Lee and Carter (1992); Booth et al. (2006); Zhao (2012); Zhao et al. (2013)). From statistics, we consider a family of functional time-series models (see, e.g., Hyndman and Booth (2008); Hyndman and Shang (2009); Hyndman et al. (2013)). For implementation, we use the StMoMo package of Villegas et al. (2018) for the RH and CBD models; we use the demography package of Hyndman (2017) for the LC models for the Gaussian error setting and functional time-series models.
Notations
Let the random variable D_{x,t} be the number of death counts in a population at age x and year t. A rectangular data array (d_{x,t},e_{x,t}) is available for data analysis where d_{x,t} is the observed number of deaths and e_{x,t} is the corresponding exposure to risk (Hatzopoulos and Haberman 2009). The force of mortality and central mortality rates are given by μ_{x,t} and m_{x,t}=d_{x,t}/e_{x,t}, respectively. Cross-classification is by individual calendar year t∈[t_{1},t_{n}] (range n) and by age x∈[x_{1},x_{k}], either grouped into k (ordered) categories or by individual year (range k), in which case year-of-birth or cohort year z=t−x∈[t_{1}−x_{k},t_{n}−x_{1}] (range n+k−1) is defined (see also Hatzopoulos and Haberman (2009)).
Lee-Carter model under a Gaussian error setting
Note that α_{x} is the age pattern of the log mortality rates averaged across years; \(\beta _{x}^{(1)}\) is the first principal component capturing relative change in the log mortality rate at each age x; \(\kappa _{t}^{(1)}\) is the first set of principal component scores measuring general level of the log mortality rate at year t; bilinear terms \(\beta _{x}^{(1)}\kappa _{t}^{(1)}\) incorporating the age-specific period trends (Pitacco et al. (2009), Section 6.2); and ε_{x,t} is the model residual at age x and year t.
In the demographic forecasting literature, the LC model adjusts κ_{t} by refitting to the total number of deaths (see Lee and Carter (1992)). In Lee and Miller’s (2001) method, the adjustment of κ_{t} involves fitting life expectancy at birth in the year t. In Booth et al.’s (2006) method, the adjustment of κ_{t} involves fitting to the age distribution of deaths rather than to the total number of deaths.
where \(b_{x}^{2}\) is the variance of the first principal component, calculated as the square of the β_{x} in (1); u_{n+h|n}=var(κ_{n+h}|κ_{1},…,κ_{n}) can be obtained from the univariate time-series model; and the model residual variance v_{x} is estimated by averaging the residual squares \(\left \{\epsilon _{x,1}^{2},\dots,\epsilon _{x,n}^{2}\right \}\) for each x in (1).
Renshaw-Haberman model under a Poisson error setting
where α_{x} is an age function capturing the general shape of mortality by age; a time index \(\kappa _{t}^{(1)}\) specifies the mortality trend, and \(\beta ^{(1)}_{x}\) modulates its effect across ages; and γ_{t−x} denotes a random cohort effect as a function of the birth-year (t−x) (see also Villegas et al. (2018)). To estimate the parameters in (3), Renshaw and Haberman (2006) assume a Poisson distribution of deaths and use a log-link function targeting the force of mortality.
Age-period-cohort (APC) model
Cairns-Blake-Dowd (CBD) model
where \(\bar {x}\) is the average age in the sample range. While \(\kappa _{t}^{(1)}\) can be viewed as a time-varying interpret, \(\kappa _{t}^{(2)}\) can be viewed as a time-varying slope. Cairns et al. (2006) produce mortality forecasts by projecting \(\kappa _{t}^{(1)}\) and \(\kappa _{t}^{(2)}\) jointly using a bivariate random walk with drift.
M7: quadratic CBD model with cohort effects
where x_{c} is a constant parameter to be estimated. Equations (5) and (6) are referred to as M6 and M8, respectively.
Plat model
In the families of the RH and CBD models, Cairns et al. (2006, 2011), Haberman and Renshaw (2011), and Villegas et al. (2018) assume that the period indexes follow a multivariate random walk with drift. For the cohort index, they assume it follows a univariate autoregressive integrated moving average model.
The functional time-series models for one population
Functional principal component analysis
Hyndman and Ullah (2007) consider a functional time-series model for forecasting age-specific mortality rate, where age is treated as a continuum. The functional time-series model allows one to smooth the observed data points, in order to reduce or eliminate measurement error. To smooth data, Hyndman and Ullah (2007) suggest the use of a penalized regression spline with monotonic constraints applied to age-specific log mortality rates denoted by lnm_{t}(x) (see Hyndman and Ullah (2007), for detail). Here, we propose to smooth age-specific mortality rates from ages 0 to 100+ and then truncate the smoothed mortality rates from 60 to 100+.
With smoothed age-specific log mortality curves (lnm_{t}(x)), we obtain a mean function denoted by μ(x). With the de-centered smoothed data, we apply a functional principal component analysis to reduce dimensionality to some functional principal components (i.e, ϕ_{k}(x)) and their associated scores (β_{k}=(β_{1,k},…,β_{n,k})). The functional principal component is constructed by sample variance of discretized functional data. Conditioning on the observed data and the estimated principal components, the point forecast of future mortality curves can be obtained by forecasting estimated principal component scores via a univariate time-series method. The prediction interval can be constructed similarly to the way of constructing prediction intervals for the Lee-Carter method in the “Lee-Carter model under a Gaussian error setting” section.
Robust functional principal component analysis
The integrated squared error provides a measure of estimation accuracy for the functional principal component approximation of the functional data. Outliers are those years that have a larger integrated squared error than a critical value calculated from a χ^{2} distribution (see Hyndman and Ullah (2007), for details). By assigning zero weight to outliers, we can again apply the functional time-series method to model and forecast age-specific mortality rates.
The functional time-series models for multiple subpopulations
When forecasting age-specific mortality for multiple subpopulations, it is advantageous to use a model that can capture correlation among subpopulations as the covariance of the multiple subpopulations often exhibits cross-correlation. By modeling the cross-correlation, it may improve forecast accuracy. Here, we consider the problem of jointly modeling and forecasting female and male mortality in order to produce coherent forecasts. Note that coherent forecasts can also be achieved by jointly modeling female or male mortality for all 47 prefectures.
Product-ratio method of Hyndman et al. (2013)
Instead of modeling female and male mortality data, we model the product and ratio functions. The advantage of this approach is that the product and ratio functions tend to behave roughly independently of each other, provided that the multiple subpopulations have approximately equal variances. On the logarithmic scale, these are sums and differences that are nearly uncorrelated. The functional time-series method in “The functional time-series models for one population” section can be applied to forecast the product and ratio functions (see Hyndman et al. (2013), fore details).
Multivariate functional time-series method
We consider data where each observation consists of w≥2 functions, [ lnm^{(1)}(x),…, lnm^{(w)}(x)]^{⊤}∈R^{w}. These multivariate functions are defined over the same domain \(\mathcal {I}\) (e.g., Jacques and Preda (2014); Chiou et al. (2014); Shang and Yang (2017)).
We follow Shang and Yang (2017) and consider the stacking of multiple subpopulations into a long vector of functions, i.e., we stack the discretized data points of each sub-population together for the same year. Then, we perform a multivariate functional principal component analysis to reduce dimensionality and summarize the main mode of information. With the extracted principal components and their scores, a functional time-series method in “The functional time-series models for one population” section can be applied again.
Multilevel functional time-series method
The multilevel functional data model has a strong resemblance to a two-way functional analysis of variance model studied by Morris et al. (2003), Cuesta-Albertos and Febrero-Bande (2010), and Zhang (2014, Section 5.4). It is a special case of the general “functional mixed model” proposed in Morris and Carroll (2006). In the case of two subpopulations, the basic idea is to decompose age-specific log mortality curves among different subpopulations into a sex-specific average μ^{j}(x), a common trend across subpopulations R_{t}(x), a sex-specific residual trend \(U_{t}^{j}(x)\), and measurement error \(e_{t}^{j}(x)\) with finite variance (σ^{2})^{j} (see, e.g., Hatzopoulos and Haberman (2013); Shang (2016)). The common and sex-specific residual trends are modeled by projecting them onto the eigenvectors of covariance operators of the aggregate and population-specific centered stochastic processes, respectively.
Model confidence set
The model confidence set procedure proposed by Hansen et al. (2011) consists of a sequence of tests permitting the construction of a set of “superior” models, where the null hypothesis of equal predictive ability (EPA) is not rejected at a specified confidence level. The EPA test statistic can be evaluated for any arbitrary loss function, such as the square or absolute loss function.
where t_{ρξ} and t_{ρ.} are defined in (10) and (11), respectively. While T_{R,M} uses the loss differential between models ρ and ξ, T_{max,M} uses the aggregated loss differential between models ρ and ξ over ξ. Oftentimes but not always, the models selected on the basis of T_{R,M} form a subset of the models selected on the basis of T_{max,M}.
Forecast results
Point forecast evaluation
An expanding window analysis of a time-series model is commonly used to assess model and parameter stabilities over time. It assesses the constancy of a model’s parameter by computing parameter estimates and their corresponding forecasts over an expanding window of a fixed size through the sample size (see Zivot and Wang 2006, Chapter 9 for details). Using the first 21 observations from 1975 to 1995 in the Japanese age-specific mortality rates, we produce one-step-ahead point forecasts. Through an expanding window approach, we re-estimate the parameters in the time-series forecasting models using the first 22 observations from 1975 to 1996. Forecasts from the estimated models are then produced for one-step-ahead. We iterate this process by increasing the sample size by one year until reaching the end of the training data period in 2005. This process produces ten one-step-ahead forecasts in the validation data period from 1996 to 2005. We compare these forecasts with the holdout samples to determine the point and interval forecast accuracies. By using the MCS procedure, we identify a superior set of models for averaging. Through the expanding window approach, we evaluate the out-of-sample point and interval forecast accuracies for the testing data from 2006 to 2015.
Interval forecast evaluation
where \(S_{\alpha, \xi }\left [\widehat {\mathcal {Y}}_{n+\xi }^{l}(x_{j}), \widehat {\mathcal {Y}}_{n+\xi }^{u}(x_{j}); \mathcal {Y}_{n+\xi }(x_{j})\right ]\) denotes the interval score at the ξth curve of the forecasting period.
Determining a superior set of models
A list of the 19 models considered
Family of models | Label | Model |
---|---|---|
Renshaw-Haberman | 1 | Lee-Carter model with Poisson error structure |
2 | Renshaw-Haberman model | |
3 | Age-period-cohort model | |
Cairns-Blake-Dowe | 4 | Cairns-Blake-Dowe model |
5 | M6 model | |
6 | M7 model | |
7 | M8 model | |
8 | Plat model | |
Lee-Carter | 9 | Lee-Carter model with Gaussian error structure |
10 | Booth-Maindonald-Smith model | |
11 | Lee-Carter model with adjustment of life expectancy | |
12 | Lee-Carter model with no adjustment to the score | |
Functional time series | 13 | Functional data model |
14 | Robust functional data model | |
15 | Product-ratio model | |
16 | Multivariate functional data model | |
17 | Multilevel functional data model | |
Model averaging | 18 | MCS (T_{max,M}) to select equal-weighted models |
19 | MCS (T_{R,M}) to select equal-weighted models |
MCS procedure using the T_{max,M} test applied to the RMSFE in the validation set from 1996 to 2005 for forecasting the Japanese female and male national and sub-national mortality for ages between 60 and 100+. From the 17 models, below is the selected superior set of the model(s)
Population | Superior models | |
---|---|---|
Female | Male | |
Japan | 1 | 1 |
Hokkaido | 1, 8, 9, 10, 11, 12, 13, 14, 15, 16 | 16 |
Aomori | 7 | 16 |
Iwate | 3, 5, 7, 8, 13, 14, 15, 17 | 13 |
Miyagi | 2, 3, 5, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 | 1, 3, 5, 8, 13, 14, 15, 16, 17 |
Akita | 7 | 14 |
Yamagata | 13, 14, 16 | 5, 7 |
Fukushima | 3 | 13 |
Ibaraki | 1, 2, 3, 5, 7, 9, 10, 11, 12, 13, 14, 15, 17 | 16 |
Tochigi | 2, 7, 14, 15 | 8, 17 |
Gunma | 8 | 2, 3, 5, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 |
Saitama | 8 | 14 |
Chiba | 3, 8, 9, 10, 11, 15 | 16 |
Tokyo | 8 | 13, 15, 16 |
Kanagawa | 8 | 14 |
Niigata | 3, 5, 7, 8, 14, 15 | 3, 5, 7, 8, 13, 17 |
Toyama | 3, 5, 7, 8 | 13, 17 |
Ishikawa | 2, 3, 4, 5, 7, 8, 13, 14, 15, 17 | 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 |
Fukui | 3 | 3, 4, 5, 7, 8, 13, 14, 15, 16, 17 |
Yamanashi | 7 | 5, 13, 14, 15, 16, 17 |
Nagano | 8 | 14 |
Gifu | 13 | 3, 5, 7, 13, 14, 15, 16, 17 |
Shizuoka | 1, 8, 10 | 13, 14, 15, 16 |
Aichi | 1, 3, 8, 9, 10, 11, 12, 14 | 13, 14, 15, 17 |
Mie | 9 | 8 |
Shiga | 8 | 13 |
Kyoto | 8, 13, 14, 15, 17 | 14 |
Osaka | 8 | 17 |
Hyogo | 8 | 13, 14, 15, 16, 17 |
Nara | 3, 5, 7, 8, 13, 14 | 13, 14, 15, 16, 17 |
Wakayama | 8 | 17 |
Tottori | 7 | 3, 5, 8, 13, 14, 15, 16, 17 |
Shimane | 13, 15 | 1, 3, 5, 7, 8, 13, 14, 15, 16, 17 |
Okayama | 10 | 1, 3, 8, 13, 14, 15, 16 |
Hiroshima | 10 | 15 |
Yamaguchi | 8, 15 | 14 |
Tokushima | 7 | 7 |
Kagawa | 14 | 3, 5, 7, 8, 13, 14, 15, 16, 17 |
Ehime | 7 | 3, 8, 13, 14, 15, 16, 17 |
Kochi | 14 | 17 |
Fukuoka | 1 | 13 |
Saga | 3, 5, 7, 8, 9, 10, 11, 12, 15 | 5, 7, 14 |
Nagasaki | 1, 8, 10, 11, 13, 14, 15, 16, 17 | 17 |
Kumamoto | 15 | 14 |
Oita | 13, 14 | 5, 7, 13, 14, 15, 16 |
Miyazaki | 2, 3, 5, 8, 13, 14, 15, 17 | 16 |
Kagoshima | 1, 3, 8, 9, 10, 11, 12, 13, 14, 15 | 3, 13, 14 |
Okinawa | 9, 10, 11, 12, 13, 14, 15, 17 | 13, 14, 15, 16, 17 |
MCS procedure using the T_{R,M} test applied to the RMSFE in the validation set from 1996 to 2005 for forecasting the Japanese female and male national and sub-national mortality for ages between 60 and 100+
Population | Superior models | |
---|---|---|
Female | Male | |
Japan | 1 | 1 |
Hokkaido | 1 | 16 |
Aomori | 7 | 16 |
Iwate | 8 | 13 |
Miyagi | 2, 3, 7, 8, 9, 10, 12, 14, 15 | 3, 13, 14, 15, 16, 17 |
Akita | 7 | 14 |
Yamagata | 13, 14, 16 | 5, 7, 13, 15, 16 |
Fukushima | 3 | 13 |
Ibaraki | 3, 7, 14, 15 | 16 |
Tochigi | 15 | 8, 17 |
Gunma | 8 | 2, 3, 5, 8, 13, 14 |
Saitama | 8 | 14 |
Chiba | 3, 8, 9 | 16 |
Tokyo | 8 | 16 |
Kanagawa | 8 | 14 |
Niigata | 3, 7, 8, 14 | 3, 5, 8, 13, 17 |
Toyama | 3, 5, 7, 8 | 13, 17 |
Ishikawa | 5, 7, 13 | 3, 5, 7, 13, 17 |
Fukui | 3 | 3, 7, 8, 13, 14, 16, 17 |
Yamanashi | 7 | 15 |
Nagano | 8 | 14 |
Gifu | 13 | 3, 5, 7, 13, 14, 15, 17 |
Shizuoka | 1, 8, 10 | 8, 13, 14, 15, 16 |
Aichi | 8 | 15 |
Mie | 3, 5, 7, 8, 9, 10, 11 | 8 |
Shiga | 8 | 13 |
Kyoto | 14 | 14 |
Osaka | 8 | 17 |
Hyogo | 8 | 15 |
Nara | 3, 5, 7 | 14 |
Wakayama | 8 | 17 |
Tottori | 7 | 3, 5, 8, 16, 17 |
Shimane | 13, 15 | 8, 15, 16, 17 |
Okayama | 10 | 1, 3, 8, 13 |
Hiroshima | 10 | 15 |
Yamaguchi | 8, 15 | 14 |
Tokushima | 3, 5, 7, 8, 15 | 5, 7, 13, 14, 15, 16 |
Kagawa | 14 | 3, 5, 7, 13, 15, 17 |
Ehime | 7 | 1, 3, 8, 13, 14, 15, 16, 17 |
Kochi | 14 | 17 |
Fukuoka | 1, 9, 10 | 13 |
Saga | 3, 7, 8, 11, 15 | 7 |
Nagasaki | 13 | 17 |
Kumamoto | 10, 11, 12, 13, 14, 15, 17 | 14 |
Oita | 13, 14, 17 | 5, 7, 13 |
Miyazaki | 3 | 16 |
Kagoshima | 1, 3, 8, 10, 11, 12, 14, 15 | 3, 13 |
Okinawa | 12, 14 | 15, 16, 17 |
MCS procedure using the T_{max,M} test applied to the mean interval score in the validation set from 1996 to 2005 for forecasting the Japanese female and male national and sub-national mortality rates for ages between 60 and 100+
Population | Superior models | |
---|---|---|
Female | Male | |
Japan | 8 | 8 |
Hokkaido | 8, 13, 14, 15, 16 | 15 |
Aomori | 8 | 14 |
Iwate | 13 | 15 |
Miyagi | 15 | 15 |
Akita | 17 | 15 |
Yamagata | 13 | 15 |
Fukushima | 15 | 15 |
Ibaraki | 8, 13, 14, 15, 17 | 15 |
Tochigi | 15 | 15 |
Gunma | 8 | 15 |
Saitama | 8 | 15 |
Chiba | 8 | 17 |
Tokyo | 8 | 15 |
Kanagawa | 8 | 15 |
Niigata | 8 | 15, 17 |
Toyama | 8 | 15 |
Ishikawa | 7, 8, 13, 14, 15, 17 | 13, 14, 15, 17 |
Fukui | 8 | 13 |
Yamanashi | 8 | 15 |
Nagano | 8 | 15 |
Gifu | 15 | 15 |
Shizuoka | 8 | 15 |
Aichi | 8 | 15 |
Mie | 8, 13, 14, 15, 16 | 15 |
Shiga | 8 | 15 |
Kyoto | 14, 15 | 15 |
Osaka | 8 | 15 |
Hyogo | 1, 3, 5, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17 | 17 |
Nara | 8, 13, 14, 15, 16 | 15 |
Wakayama | 8, 15 | 15 |
Tottori | 8 | 15 |
Shimane | 15 | 15 |
Okayama | 15 | 15 |
Hiroshima | 8 | 15 |
Yamaguchi | 8 | 15 |
Tokushima | 15, 16 | 13, 14, 15, 16 |
Kagawa | 8, 13, 14, 15 | 15 |
Ehime | 8 | 15 |
Kochi | 17 | 15 |
Fukuoka | 8 | 15 |
Saga | 15 | 15 |
Nagasaki | 8 | 15 |
Kumamoto | 13, 14, 15, 17 | 15 |
Oita | 15 | 15 |
Miyazaki | 17 | 15 |
Kagoshima | 8, 13, 14, 15, 16 | 15 |
Okinawa | 15 | 15 |
MCS procedure using the T_{R,M} test applied to the mean interval score in the validation set from 1996 to 2005 for forecasting the Japanese female and male national and sub-national mortality rates for ages between 60 and 100+
Population | Superior models | |
---|---|---|
Female | Male | |
Japan | 8 | 8 |
Hokkaido | 8, 13, 15, 16, 17 | 15 |
Aomori | 8 | 14 |
Iwate | 13 | 15 |
Miyagi | 15 | 15 |
Akita | 17 | 15 |
Yamagata | 13 | 15 |
Fukushima | 15 | 15 |
Ibaraki | 8, 13, 14, 17 | 15 |
Tochigi | 15 | 15 |
Gunma | 8 | 15 |
Saitama | 8 | 15 |
Chiba | 8 | 17 |
Tokyo | 8 | 15 |
Kanagawa | 8 | 15 |
Niigata | 8 | 13, 15, 17 |
Toyama | 8 | 15 |
Ishikawa | 13 | 15 |
Fukui | 8 | 13 |
Yamanashi | 8 | 15 |
Nagano | 8 | 15 |
Gifu | 8 | 15 |
Shizuoka | 8 | 15 |
Aichi | 8 | 15 |
Mie | 8, 13, 15 | 15 |
Shiga | 8 | 15 |
Kyoto | 14, 15 | 15 |
Osaka | 8 | 15 |
Hyogo | 8 | 17 |
Nara | 8, 15, 16 | 15 |
Wakayama | 8, 15 | 15 |
Tottori | 8 | 15 |
Shimane | 15 | 15 |
Okayama | 15 | 15 |
Hiroshima | 8 | 15 |
Yamaguchi | 8 | 13, 14, 15 |
Tokushima | 15, 16 | 13, 15, 16 |
Kagawa | 8, 14, 15 | 15, 16, 17 |
Ehime | 8 | 15 |
Kochi | 17 | 15 |
Fukuoka | 8 | 15 |
Saga | 15 | 15 |
Nagasaki | 8 | 15 |
Kumamoto | 13, 15, 17 | 15 |
Oita | 15 | 15 |
Miyazaki | 17 | 15 |
Kagoshima | 15 | 15 |
Okinawa | 15 | 15 |
Point and interval forecast comparison
Point and interval forecast accuracies among the 17 models and two model-averaged methods in the Japanese national data and the average of 47 sub-national populations for ages between 60 and 100+
RMSFE | Mean interval score | ||||
---|---|---|---|---|---|
Series | Method | National data | Sub-national data | National data | Sub-national data |
Female | 1 | 0.54 | 1.11 | 1.81 | 3.88 |
2 | 4.20 | 6.24 | 7.12 | 61.70 | |
3 | 1.34 | 1.61 | 4.91 | 5.41 | |
4 | 1.98 | 2.17 | 7.24 | 7.09 | |
5 | 1.20 | 1.51 | 3.71 | 4.50 | |
6 | 2.05 | 2.27 | 5.95 | 5.84 | |
7 | 0.87 | 1.27 | 1.93 | 2.89 | |
8 | 0.33 | 1.02 | 0.86 | 2.40 | |
9 | 0.67 | 1.26 | 2.10 | 4.26 | |
10 | 0.69 | 1.27 | 2.24 | 4.34 | |
11 | 0.69 | 1.27 | 2.20 | 4.33 | |
12 | 0.69 | 1.28 | 2.22 | 4.30 | |
13 | 0.71 | 1.21 | 1.31 | 2.71 | |
14 | 0.72 | 1.21 | 1.35 | 2.70 | |
15 | 0.61 | 1.21 | 1.32 | 2.94 | |
16 | 0.83 | 1.25 | 1.43 | 3.40 | |
17 | 0.81 | 1.23 | 1.80 | 2.73 | |
18 | 0.54 | 1.24 | 0.87 | 2.57 | |
19 | 0.54 | 1.22 | 0.87 | 2.57 | |
Male | 1 | 0.71 | 2.55 | 2.79 | 9.56 |
2 | 2.27 | 4.02 | 7.97 | 14.57 | |
3 | 1.93 | 3.13 | 8.17 | 11.40 | |
4 | 2.92 | 3.76 | 12.18 | 13.01 | |
5 | 1.78 | 2.96 | 6.42 | 9.45 | |
6 | 2.75 | 3.69 | 9.72 | 10.47 | |
7 | 1.60 | 2.84 | 3.81 | 7.91 | |
8 | 0.65 | 2.47 | 1.37 | 5.77 | |
9 | 0.99 | 3.78 | 3.68 | 12.80 | |
10 | 1.00 | 3.76 | 3.80 | 12.78 | |
11 | 1.02 | 3.83 | 3.91 | 13.05 | |
12 | 1.00 | 3.47 | 3.63 | 10.88 | |
13 | 0.72 | 2.51 | 1.50 | 5.86 | |
14 | 0.78 | 2.51 | 1.49 | 5.87 | |
15 | 0.72 | 2.47 | 1.98 | 5.02 | |
16 | 0.93 | 2.47 | 1.79 | 7.23 | |
17 | 0.76 | 2.44 | 1.79 | 5.49 | |
18 | 0.71 | 2.51 | 1.36 | 5.06 | |
19 | 0.71 | 2.50 | 1.36 | 5.07 |
As measured by the mean RMSFE over 10 years in the forecasting period, the Plat model gives the most accurate point forecasts for the whole of Japan. For the average of 47 prefectures in Japan, the Plat model gives the most accurate point forecasts for females, while the multilevel functional time-series method performs the best for males. Although the model averaging methods are not the best model, in this case, they rank among the top-performing methods. Between the two statistical significance tests, there is a marginal difference in the point forecast accuracy.
As measured by the mean interval scores over ten different horizons, the Plat model produces the most accurate interval forecasts for the whole of Japan and its prefectures for females. For males, the model averaging methods perform the best for the whole of Japan and rank as the second best after the product-ratio method for the average of 47 prefectures. Again, there is a marginal difference between the two statistical significance tests regarding interval forecast accuracy.
A competing model averaging method
Point and interval forecast accuracies for an existing model-averaged method averaged across the Japanese female and male national and sub-national mortality rates for ages between 60 and 100+. Forecast errors have been multiplied by 100. The results show its inferior point and interval forecast accuracies compared to the proposed two model averaging methods. This further confirms that one should not average all models, but a subset of all “good” models
Series | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | Mean |
---|---|---|---|---|---|---|---|---|---|---|---|
Japan (RMSFE) | |||||||||||
Female | 0.93 | 0.95 | 1.26 | 0.89 | 1.49 | 1.55 | 1.43 | 1.30 | 1.03 | 1.20 | 1.20 |
Male | 0.95 | 1.07 | 2.01 | 0.98 | 2.10 | 1.81 | 1.71 | 1.03 | 0.80 | 1.20 | 1.37 |
Japan (mean interval score) | |||||||||||
Female | 1.81 | 2.13 | 3.40 | 1.75 | 4.99 | 4.55 | 3.55 | 2.80 | 1.70 | 2.73 | 2.94 |
Male | 2.39 | 3.05 | 6.73 | 2.61 | 7.87 | 6.18 | 5.61 | 2.59 | 1.89 | 3.40 | 4.23 |
Japanese prefectures (RMSFE) | |||||||||||
Female | 1.28 | 1.28 | 1.44 | 1.21 | 1.63 | 1.69 | 1.76 | 1.50 | 1.23 | 1.89 | 1.49 |
Male | 2.72 | 2.71 | 3.00 | 2.73 | 3.26 | 3.20 | 2.93 | 2.60 | 2.20 | 2.22 | 2.76 |
Japanese prefectures (mean interval score) | |||||||||||
Female | 2.79 | 2.71 | 3.35 | 2.64 | 3.94 | 4.12 | 4.34 | 3.17 | 3.07 | 3.05 | 3.32 |
Male | 7.18 | 6.74 | 7.94 | 6.92 | 8.79 | 8.58 | 7.73 | 6.18 | 5.42 | 5.63 | 7.11 |
Compared to our proposal model averaging method, the existing model averaging method assigns different weights to different models. From the in-sample forecast errors, the existing model averaging method assigns higher weights for those more accurate models and lower weights for those less accurate models. By contrast, our proposed model averaging method selects a superior subset of models and assign equal weights. From the results, we find that the proposed model averaging method gives a smaller forecast error than the existing model averaging method.
Discussion
Using the national and sub-national Japanese mortality rates, we evaluate and compare point and interval forecast accuracies, as measured by the root mean squared error and mean interval score, among the 17 time-series extrapolation methods and two model averaging methods considered. From the viewpoint of the point forecast accuracy, the Plat model gives the smallest point forecast errors, followed by the Lee-Carter and model-averaged method for Japanese females and males. In this case, because the superior set of the models was determined from in-sample forecast errors, it is the case where the best model for the in-sample forecasts may not be the best model for out-of-sample forecasts. This affects the forecast accuracy of the model-averaged method. Also, the Lee-Carter method with the Poisson error structure is more accurate than the version with the Gaussian error structure. For modeling sub-national Japanese females, the Plat model also performs the best, while the multilevel functional time-series model performs the best for males.
From the viewpoint of the interval forecast accuracy, the Plat model gives the smallest interval forecast errors, followed by the model-averaged methods for Japanese females. For Japanese males, the model-averaged methods produce the smallest interval forecast errors for Japan and are on a par with the product-ratio method for Japanese sub-national data. To our surprise, the Renshaw-Haberman methods produce relatively worse interval forecast accuracy than that produced by the Lee-Carter and functional time-series methods. The result could be due to the instability of parameter estimation.
From Tables 2, 3, 4, and 5, the best model for producing point forecasts does not necessary the same as the best model for producing interval forecasts, and the best model for producing the national series does not necessarily perform the best for the sub-national series, as the features of the data may be different.
Because different models have their advantages and disadvantages, we apply a model-averaged method to select a set of superior model based on the model confidence set. The model confidence set is a procedure that determines a set of superior models based on the in-sample forecast errors.
For producing point forecasts, the selected superior sets of models are more diverse. For producing interval forecasts, the selected superior set of models often includes the product-ratio method, especially for male series. Given the product-ratio method is a joint modeling and coherent forecasting method, it achieves better forecast accuracy for males with a small sacrifice in terms of the female results. By coherent, we believe within each prefecture, female and male subpopulations share the similar characteristics, such as health facilities. Generally, the joint modeling methods, which also include the multivariate and multilevel functional time-series methods, often but not always outperform the model without incorporating correlation among subpopulations. The model-averaged forecasts may not perform the best for females and males, but they tend to give an aggregate best performance.
We show that potential gains in forecast accuracy can be achieved by discarding the worse performing models before combining the forecasts equally. By contrast, an existing model averaging method assigning different weights for all models performs worse than the proposed method. We find that the proposed model averaging method offers a more robust procedure for selecting the forecasting models based on their in-sample performances. By robustness, the model-averaged methods are protected against model misspecification. The advantage of the model-averaged methods is more apparent for males than females.
The accurate forecasting of mortality at retirement ages is essential to determine life, fixed-term, and delayed annuity prices for various maturities and starting ages (see, e.g., Shang and Haberman (2017)). In the online supplement, we present a study on calculating single-premium fixed-term immediate annuity. To forecast mortality rates, we suggest considering the notion of model averaging.
- 1)
We considered ages from 60 to 100+ to study the mortality pattern of retirees. We could apply the model averaging idea to other age groups, such as ages from 0 to 100+. With different age groups, the point and interval forecast results may be different.
- 2)
Other age-specific mortality forecasting models could be incorporated into the model averaging. We consider only time-series extrapolation methods and did not consider the expectation or explanation methods. For long-term forecasts and the expectation, expectation method has been used by the The Institute and Faculty of Actuaries and the CMI (2018). The inclusion of the expectation or explanation methods may alter the selection of superior models.
- 3)
We modeled age-specific mortality rates, but the focus could also be on death counts or survival probabilities. For example, when we model age-specific death counts, there are other models, such as compositional data analysis, that could be included in the initial model pool.
- 4)
After selecting the superior set of models, one could assign different weights instead of equal weights considered. With different weights, our forecast results may be improved as considered in Shang (2012).
- 5)
In applying the model confidence set procedure of Hansen et al. (2011), we consider the 90% confidence level. Considering other levels of confidence is possible. In general, as the confidence level increases, the number of superior models decreases.
- 6)
We evaluated and compared one-step-ahead, five-step-ahead, and ten-step-ahead point and interval forecast accuracies. Considering other forecast horizons is also possible. For the longer term, the extrapolation methods may not perform well.
- 7)
We evaluated point forecast accuracy by the root mean square error and interval forecast accuracy by the mean interval score and coverage probability deviance, respectively. Considering other forecast error criteria, such as mean absolute percentage error or mean absolute scaled error, is also possible. With different error measures, the point and interval forecast results may be different.
We believe the present work paves the way for the above possible future research directions, and the proposed model averaging method should be a welcome addition to the demographic modeling and forecasting.
Declarations
Acknowledgements
Thanks to the two reviewers and the Guest Editor for the insightful comments and suggestions. We thank the comments and suggestions received from the S3RI seminar participants, especially Professors Peter W. F. Smith and Zudi Lu, at the Southampton Statistical Sciences Research Institute at the University of Southampton. We also thank Dr. Andres Villegas for answering our queries about the StMoMo package. The first author thanks the hospitalities of the Department of Statistics, Colorado State University; Department of Statistical Science, Cornell University; and Cass Business School, City University of London, when preparing this manuscript.
Funding
Not applicable
Availability of data and materials
The data used in this paper were obtained from Human Mortality Database cited in the paper.
Authors’ contributions
HLS proposed the statistical methodology and implemented computation. SH polished the manuscript and added detailed exposition especially in the Renshaw-Haberman models. Both authors read and approved the final version of the manuscript.
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
Authors’ Affiliations
References
- Abel, G.J., Bijak, J., Forster, J.J., Raymer, J., Smith, P.W.F., Wong, J.S.T. (2013). Integrating uncertainty in time series population forecasts: an illustration using a simple projection model. Demographic Research, 29, 1187–1226.View ArticleGoogle Scholar
- Aiolfi, M., Capistrán, C., Timmermann, A.G. (2010). Forecast combinations, Working paper 2010-21, CREATES. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1609530. Accessed 20 June 2018.
- Bates, J.M., & Granger, C.W.J. (1969). The combination of forecasts. Operational Research Quarterly, 20(4), 451–468.View ArticleGoogle Scholar
- Bernardi, M., & Catania, L. (2014). The model confidence set package for R. Technical report, Sapienza University of Rome. https://arxiv.org/abs/1410.8504. Accessed 20 June 2018.
- Bijak, J. (2011). Forecasting international migration in Europe: a Bayesian view. New York: Springer.View ArticleGoogle Scholar
- Booth, H. (2006). Demographic forecasting 1980 to 2005 in review. International Journal of Forecasting, 22(3), 547–581.View ArticleGoogle Scholar
- Booth, H., Hyndman, R.J., Tickle, L., De Jong, P. (2006). Lee-Carter mortality forecasting: a multi-country comparison of variants and extensions. Demographic Research, 15, 289–310.View ArticleGoogle Scholar
- Booth, H., Maindonald, J., Smith, L. (2002). Applying Lee-Carter under conditions of variable mortality decline. Population Studies, 56(3), 325–336.View ArticleGoogle Scholar
- Booth, H., & Tickle, L. (2008). Mortality modelling and forecasting: a review of methods. Annals of Actuarial Science, 3(1-2), 3–43.View ArticleGoogle Scholar
- Cairns, A.J.G., Blake, D., Dowd, K. (2006). A two-factor model for stochastic mortality with parameter uncertainty: theory and calibration. The Journal of Risk and Insurance, 73(4), 687–718.View ArticleGoogle Scholar
- Cairns, A.J.G., Blake, D., Dowd, K., Coughlan, G.D., Epstein, D., Khalaf-Allah, M. (2011). Mortality density forecasts: an analysis of six stochastic mortality models. Insurance: Mathematics & Economics, 48(3), 355–367.Google Scholar
- Cairns, A.J.G., Blake, D., Dowd, K., Coughlan, G.D., Epstein, D., Ong, A., Balevich, I. (2009). A quantitative comparison of stochastic mortality models using data from England and Wales and the United States. North American Actuarial Journal, 13(1), 1–35.View ArticleGoogle Scholar
- Chiou, J.-M., Chen, Y.-T., Yang, Y.-F. (2014). Multivariate functional principal component analysis: a normalization approach. Statistica Sinica, 24(4), 1571–1596.Google Scholar
- Clayton, D., & Schifflers, E. (1987a). Models for temporal variation in cancer rates. I: age-period and age-cohort models. Statistics in Medicine, 6(4), 449–467.View ArticleGoogle Scholar
- Clayton, D., & Schifflers, E. (1987b). Models for temporal variation in cancer rates. II: age-period-cohort models. Statistics in Medicine, 6(4), 469–481.View ArticleGoogle Scholar
- Clemen, R.T. (1989). Combining forecasts: a review and annotated bibliography. International Journal of Forecasting, 5(4), 559–583.View ArticleGoogle Scholar
- Crawford, T., de Haan, R., Runchey, C. (2008). Longevity risk quantification and management: a review of relevant literature. Technical report, Society of Actuaries. https://www.soa.org/research-reports/2009/research-long-risk-quant/. Accessed 20 June 2018.
- Cuesta-Albertos, J.A., & Febrero-Bande, M. (2010). A simple multiway ANOVA for functional data. Test, 19(3), 537–557.View ArticleGoogle Scholar
- Currie, I.D., Durban, M., Eilers, P.H.C. (2004). Smoothing and forecasting mortality rates. Statistical Modelling, 4(4), 279–298.View ArticleGoogle Scholar
- Denuit, M., Devolder, P., Goderniaux, A.C. (2007). Securitization of longevity risk: pricing survivor bonds with Wang transform in the Lee-Carter framework. The Journal of Risk and Insurance, 74(1), 87–113.View ArticleGoogle Scholar
- Dickinson, J.P (1975). Some statistical results in the combination of forecasts. Operational Research Quarterly, 24(2), 253–260.View ArticleGoogle Scholar
- Fischer, I., & Harvey, N. (1999). Combining forecasts: what information do judges need to outperform the simple average?. International Journal of Forecasting, 15(3), 227–246.View ArticleGoogle Scholar
- Genre, V., Kenny, G., Meyler, A., Timmermann, A. (2013). Combining expert forecasts: can anything beat the simple average?. International Journal of Forecasting, 29(1), 108–121.View ArticleGoogle Scholar
- Girosi, F., & King, G. (2008). Demographic forecasting. Princeton: Princeton University Press.Google Scholar
- Gneiting, T., & Katzfuss, M. (2014). Probabilistic forecasting. Annual Review of Statistics and Its Application, 1, 125–151.View ArticleGoogle Scholar
- Gneiting, T., & Raftery, A.E. (2007). Strictly proper scoring rules, prediction and estimation. Journal of the American Statistical Association, 102(477), 359–378.View ArticleGoogle Scholar
- Graefe, A. (2015). Improving forecasts using equally weighted predictors. Journal of Business Research, 68(8), 1792–1799.View ArticleGoogle Scholar
- Haberman, S., & Renshaw, A. (2008). Mortality, longevity and experiments with the Lee-Carter model. Lifetime Data Analysis, 14(3), 286–315.View ArticleGoogle Scholar
- Haberman, S., & Renshaw, A. (2009). On age-period-cohort parametric mortality rate projections. Insurance: Mathematics & Economics, 45(2), 255–270.Google Scholar
- Haberman, S., & Renshaw, A. (2011). A comparative study of parametric mortality projection models. Insurance: Mathematics & Economics, 48(1), 35–55.Google Scholar
- Hanewald, K., Post, T., Gründl, H. (2011). Stochastic mortality, macroeconomic risks and life insurer solvency. The Geneva Papers on Risk and Insurance-Issues and Practice, 36(3), 458–475.View ArticleGoogle Scholar
- Hansen, P.R., Lunde, A., Nason, J.M. (2011). The model confidence set. Econometrica, 79(2), 453–497.View ArticleGoogle Scholar
- Hatzopoulos, P., & Haberman, S. (2009). A parameterized approach to modeling and forecasting mortality. Insurance: Mathematics and Economics, 44(1), 103–123.Google Scholar
- Hatzopoulos, P., & Haberman, S. (2013). Common mortality modeling and coherent forecasts. An empirical analysis of worldwide mortality data. Insurance: Mathematics & Economics, 52(2), 320–337.Google Scholar
- Hunt, A., & Blake, D. (2014). A general procedure for constructing mortality models. North American Actuarial Journal, 18(1), 116–138.View ArticleGoogle Scholar
- Hyndman, R., & Booth, H. (2008). Stochastic population forecasts using functional data models for mortality, fertility and migration. International Journal of Forecasting, 24(3), 323–342.View ArticleGoogle Scholar
- Hyndman, R.J. (2017). Demography: forecasting mortality, fertility, migration and population data. R package version 1.20. https://CRAN.R-project.org/package=demography. Accessed 20 June 2018.
- Hyndman, R.J., Booth, H., Yasmeen, F. (2013). Coherent mortality forecasting: the product-ratio method with functional time series models. Demography, 50(1), 261–283.View ArticleGoogle Scholar
- Hyndman, R.J., & Shang, H.L. (2009). Forecasting functional time series (with discussions). Journal of the Korean Statistical Society, 38(3), 199–211.View ArticleGoogle Scholar
- Hyndman, R., & Ullah, M. (2007). Robust forecasting of mortality and fertility rates: a functional data approach. Computational Statistics & Data Analysis, 51(10), 4942–4956.View ArticleGoogle Scholar
- Jacques, J., & Preda, C. (2014). Model-based clustering for multivariate functional data. Computational Statistics & Data Analysis, 71, 92–106.View ArticleGoogle Scholar
- Japanese Mortality Database. (2017). National Institute of Population and Social Security Research. Available at http://www.ipss.go.jp/p-toukei/JMD/index-en.html. Accessed 20 June 2018.
- Koissia, M.C. (2006). Longevity and adjustment in pension annuities, with application to Finland. Scandinavian Actuarial Journal, 2006(4), 226–242.View ArticleGoogle Scholar
- Lee, R.D., & Carter, L.R. (1992). Modeling and forecasting U.S. mortality. Journal of the American Statistical Association, 87(419), 659–671.Google Scholar
- Lee, R.D., & Miller, T. (2001). Evaluating the performance of the Lee-Carter method for forecasting mortality. Demography, 38(4), 537–549.View ArticleGoogle Scholar
- Makridakis, S., Andersen, A., Carbone, R., Fildes, R., Hibon, M., Lewandowski, R., Newton, J., Parzen, E., Winkler, R. (1982). The accuracy of extrapolation (time series) methods: results of a forecasting competition. Journal of Forecasting, 1(2), 111–153.View ArticleGoogle Scholar
- Makridakis, S., Hibon, M., Moser, C. (1979). Accuracy of forecasting: an empirical investigation. Journal of the Royal Statistical Society: Series A, 142(2), 97–145.View ArticleGoogle Scholar
- Morris, J.S., & Carroll, R.J. (2006). Wavelet-based functional mixed models. Journal of the Royal Statistical Society: Series B, 68(2), 179–199.View ArticleGoogle Scholar
- Morris, J.S., Vannucci, M., Brown, P.J., Carroll, R.J. (2003). Wavelet-based nonparametric modeling of hierarchical functions in colon carcinogenesis. Journal of the American Statistical Association, 98(463), 573–583.View ArticleGoogle Scholar
- Pitacco, E., Denuit, M., Haberman, S., Olivieri, A. (2009). Modelling longevity dynamics for pensions and annuity business. Oxford: Oxford University Press.Google Scholar
- Plat, R. (2009a). On stochastic mortality modeling. Insurance: Mathematics & Economics, 45(3), 393–404.Google Scholar
- Plat, R. (2009b). Stochastic portfolio specific mortality and the quantification of mortality basis risk. Insurance: Mathematics and Economics, 45(1), 123–132.Google Scholar
- Renshaw, A.E., & Haberman, S. (2003a). Lee-Carter mortality forecasting with age-specific enhancement. Insurance: Mathematics & Economics, 33(2), 255–272.Google Scholar
- Renshaw, A.E., & Haberman, S. (2006). A cohort-based extension to the Lee-Carter model for mortality reduction factors. Insurance: Mathematics and Economics, 38(3), 556–570.Google Scholar
- Renshaw, A.E., & Haberman, S. (2008). On simulation-based approaches to risk measurement in mortality with specific reference to Poisson Lee-Carter modelling. Insurance: Mathematics & Economics, 42(2), 797–816.Google Scholar
- Renshaw, A., & Haberman, S. (2003b). Lee-Carter mortality forecasting: a parallel generalized linear modelling approach for England and Wales mortality projections. Journal of the Royal Statistical Society: Series C (Applied Statistics), 52(1), 119–137.View ArticleGoogle Scholar
- Samuels, J.D., & Sekkel, R.M. (2017). Model confidence sets and forecast combination. International Journal of Forecasting, 33(1), 48–60.View ArticleGoogle Scholar
- Shang, H.L. (2012). Point and interval forecasts of age-specific life expectancies: a model averaging approach. Demographic Research, 27, 593–644.View ArticleGoogle Scholar
- Shang, H.L. (2015). Statistically tested ccomparison of the accuracy of forecasting methods for age-specific and sex-specific mortality and life expectancy. Population Studies, 69(3), 317–335.View ArticleGoogle Scholar
- Shang, H.L. (2016). Mortality and life expectancy forecasting for a group of populations in developed countries: a multilevel functional data method. The Annals of Applied Statistics, 10(3), 1639–1672.View ArticleGoogle Scholar
- Shang, H.L., Booth, H., Hyndman, R.J. (2011). Point and interval forecasts of mortality rates and life expectancy: a comparison of ten principal component methods. Demographic Research, 25, 173–214.View ArticleGoogle Scholar
- Shang, H.L., & Haberman, S. (2017). Grouped multivariate and functional time series forecasting: an application to annuity pricing. Insurance: Mathematics & Economics, 75, 166–179.Google Scholar
- Shang, H.L, & Hyndman, R.J (2017). Grouped functional time series forecasting: an application to age-specific mortality rates. Journal of Computational and Graphical Statistics, 26(2), 330–343.View ArticleGoogle Scholar
- Shang, H.L., Wiśniowski, A., Bijak, J., Smith, P.W.F., Raymer, J. (2014). Bayesian functional models for population forecasting. In: Marsili, M., & Capacci, G. (Eds.) In Proceedings of the Sixth Eurostat/UNECE Work Session on Demographic Projections. Istituto Nazionale di Statistica, Rome.Google Scholar
- Shang, H.L., & Yang, Y. (2017). Grouped multivariate functional time series method: an application to mortality forecasting. In: Aneiros, G., Bongiorno, E.G., Cao, R., Vieu, P. (Eds.) In Functional Statistics and Related Fields. Springer, Cham, (pp. 233–241).View ArticleGoogle Scholar
- The Institute and Faculty of Actuaries and the CMI. (2018). Continuous Mortality Investigation. https://www.actuaries.org.uk/learn-and-develop/continuous-mortality-investigation. Accessed 15 Sept 2018.
- Tickle, L., & Booth, H. (2014). The longevity prospects of Australian seniors: an evaluation of forecast method and outcome. Asia-Pacific Journal of Risk and Insurance, 8(2), 259–292.View ArticleGoogle Scholar
- Villegas, A.M., Kaishev, V.K., Millossovich, P. (2018). StMoMo: an R package for stochastic mortality modeling. Journal of Statistical Software, 84(3), 1–38.View ArticleGoogle Scholar
- Wiśniowski, A., Smith, P.W.F., Bijak, J., Raymer, J., Forster, J. (2015). Bayesian population forecasting: extending the Lee-Carter method. Demography, 52(3), 1035–1059.View ArticleGoogle Scholar
- Zhang, J.-T. (2014). Analysis of variance for functional data. Boca Raton: Chapman & Hall.Google Scholar
- Zhao, B.B. (2012). A modified Lee-Carter model for analysing short-base-period data. Population Studies, 66(1), 39–52.View ArticleGoogle Scholar
- Zhao, B.B., Liang, X., Zhao, W., Hou, D. (2013). Modeling of group-specific mortality in China using a modified Lee-Carter model. Scandinavian Actuarial Journal, 2013(5), 383–402.View ArticleGoogle Scholar
- Zivot, E., & Wang, J. (2006). Modeling financial time series with S-PLUS. New York: Springer.Google Scholar