Actuarial
Demography (ACTDEM)



Plan:
o Scope
of the Team
o Team
Members
o Algerian
Mortality Data
o Algerian
Fertility Data
o Population
Forecast
o Organized
Conferences & Workshops
o Achieved
and ongoing Research
o References
The
team ActDem is interested in quantitative demography and actuarial
science issues implied by economic insurance or social security. In other
words, the team aims to investigate the demographic behavior of the Algerian
population and to model the different risks in the intention to define the
actuarial balances.
Scope of the Team The
team "Actuarial Demography" focusses on studying the demographic
changes and the risks supported by the population and their impact on actuarial
calculations. Firstly, population dynamics needs to be measured and its
future trend needs to be expected. Public planning as the orientation of
public policies requires the understanding of the trend that the population
dynamics will undertake in the future. Such an element returns to study the
demographic phenomena as mortality, nuptiality and
fertility. Then, the effect of aging and longevity on the sustainability of
the social security system can be evaluated. Secondly, the social risks that
people can be exposed to are numerous and need to be covered. This coverage
can be considered from two points of view: Economic Insurance and Social
Security. Both require the evaluation of the covered risk. When it concern
social security, premium can be shared by the insured themselves and the
state in the framework of its social policy, when it concerns economic
insurance, it should be equivalence between the supported risk and the payed
premium. Actuarial
Demography can be managed into two mean categories: Demographic modeling and
Actuarial tools for life insurance. The first part consists on
the quantitative methods aiming to model the different demographic phenomena:
o Mortality & Longevity, o Fertility, o Population growth, By studying the phenomena cited above, we
aim to quantify the effect of population change on the social security
systems in its different components: o Direct pensions which tie directly to
aging and longevity, o Survivors benefits (spouses) which tie
to the sex differential mortality and the age gap between spouses, o Survivors benefits (orphans) which is
related to the fertility behavior, In addition, many other
social risks need to be studied and quantified either from the insurance of
the social security points of view: o Life annuities / fully funded pension plans, o Disability insurance, o Jobless insurance, o Illness and Health insurance, 
Team Members Farid
FLICI (Team Leader) – Mortality & Longevity Modeling, Pensions Farid
is an actuary, graduated in 2006 from the National High School of Statistics
and Applied Economics of Algiers. In 2011, he obtained his “Magister degree”
in quantitative finance and his PhD in Statistics in 2017 from the same
institution. He has dedicated his two dissertations respectively for "Provisionnement des Rentes Viagères en Algérie entre Approche Statique et Approche Prospective" and “ Longevity and Pension Plan
Sustainability in Algeria: Taking the retirees mortality experience into
account”. He has been teaching “Demographic
analysis” and “Micro economics of insurance” at the school of Statistics
between 2009 and 2012 before he joins the Center for Research in Applied
Economics in the mid2012. His task focused on Demographic modeling and Life
Actuarial Science and more precisely Mortality & Longevity Modeling and
Pension plan sustainability. He has some publications in peer reviewed
journals and conferences proceeding of the International Association of
Actuaries. He is a member of the Mortality Working Group of the IAA. Ibrahim ZAIMEN (Junior
Researcher)  Social Statistics, Fertility and Nuptiality Ibrahim
graduated in statistics and applied economics in 2013 from the National High
School of Statistics and Applied Economics (ENSSEA). In 2016, he obtained his
Magister Degree in Statistics and actually he is enrolled in PhD program at
the same institution. He dedicated his
Magister dissertation to “Missing data imputation using Regression Methods”.
Currently, he is a Junior Researcher at the Centre for Research in Applied
Economics for Development (CREAD). His research focusses on modeling Fertility , Nuptiality and
related areas. Mohamed
SADOUN (Research Engineer) – Financial Modeling, Time Series Mohamed is a PhD student in Applied Mathematics more
oriented to Time Series Modeling at the University of Science and Technology Houari Boumedien (USTHB). He
graduated in Statistical Probability in 2010 and holds a Master's degree in Financiel Mathematics (2012) from USTHB. He integrated
the Centre for Research in Applied Economics for Development (CREAD) in
November 2017. His interest fields are: Probability and Statistical
Inference, Modeling Financial Markets, Time series analysis and algorithmic
programming. Also, he participated in several conferences on Mathematics and
related fields: Journées de l’économétrie
et de la finance (Maroc, Novembre
2016), The African Conference of the International Econometrics
Society (Algeria, June 2017), The
30^{th} International conference of the Jangjeon
Mathematical Society (USTHB, July 2017). Meryem
CHINOUNE (Research Engineer) – Survival Analysis, Disability Meryem is a PhD student (2^{nd} year)
in Applied Statistics at the National High School
of Statistics and Applied Economics (ENSSEA), her research focuses on “Survival Analysis” . She’s currently occupied as an
engineer for research support at the Centre for Research in Applied Economics
for Development (CREAD). She obtained a master degree in 2016 in Applied
Statistics and Econometrics, Her Master's dissertation was entitled:
Analytical Study and ShortTerm Forecast of air passenger’s traffic
(20062015). Her skills revolve mostly around Statistics, Data Analysis
(XLSTAT, SPSS) and Time Series (EVIEWS). Soumia
BOUCHOUK (Research Engineer) – Fertility, Nuptiality Soumia holds the diploma of engineer in statistics and applied Economics /
option: management and decision models from the High National School of
Statistics and Applied Economics (ENSSEA) in 2011. Since 2014, she has been
working as research engineer at the Centre for Research in Applied Economics
for Development (CREAD). Actually, she’s preparing a Master degree in
Demography, option: “Health & Population” and she is
preparing her dissertation on “the regional differences in fertility
behaviour of the Algerian Population”. 
Algerian
Mortality Data
Usually,
the practice of mortality modeling distinguishes mainly two types of life tables:
Demographic life tables and Actuarial Life tables. The first type is concerned
with the mortality of the global population while the second focusses on the
mortality of the insured population. Here, data can be provided by the social
security system (including retirement) or from the private insurance companies
(internal models or experience life tables) or from the whole insurance market
(Market’s life tables). In some situations, when the data of the insured
population is lacked, the global population life tables can be used for
insurance products pricing and reserving.
Implementing
Mortality forecasting Models requires to make
available a continuous mortality surface as long and as consistent as possible.
Mortality data in Algeria are mainly provided by the Office for National
Statistics (ONS: www.ons.dz) in the framework of its annual
publications. The first life table based on the civil registration data has
been published in 1977. The ONS has tried to publish this kind of statistics
with an annual frequency, but it become possible only starting from 1998.
Accordingly, some years of the period [19771997] are still with no data.
Similarly, some life tables of the same period were closed out at early ages
(below the age 80).
The
Mortality Rates as published by the ONS can be downloaded through the following
links:
(Data Source : ONS annual publications
Period (1977
2011): OFFICE NATIONAL DES STATISTIQUES ONS. (2012a). Rétrospective de données
Statistiques 1962  2011 :
Démographie. www.ons.dz/IMG/pdf/CH1DEMOGRAPHIE.pdf
2012: OFFICE
NATIONAL DES STATISTIQUES ONS. (2013). Démographie Algérienne 2012. http://www.ons.dz/IMG/pdf/demographie_algerienne2012.pdf
2013: OFFICE
NATIONAL DES STATISTIQUES ONS. (2014, n° 658). Démographie Algérienne 2013. http://www.ons.dz/img/pdf/demographie_algerienne2013.pdf
2014: OFFICE
NATIONAL DES STATISTIQUES ONS. (2015, n° 690). Démographie Algérienne 2014. http://www.ons.dz/IMG/pdf/Demographie_algerienne_2014.pdf
2015: OFFICE
NATIONAL DES STATISTIQUES ONS. (2016, n° 740). Démographie Algérienne 2015. http://www.ons.dz/IMG/pdf/Demographie_algerienne_2015.pdf
2016: OFFICE
NATIONAL DES STATISTIQUES ONS. (2017, n° 779). Démographie Algérienne 2016. http://www.ons.dz/IMG/pdf/Demographie_algerienne_2016.pdf
)
National
life tables are published following five ages description. Actuarial
calculations (longevity pricing among others) need to make available continuous
historical mortality surfaces under a single age description. Also, mortality
rates need to be extended until the survival age limit or nearly. In a previous
work (Flici, 2014), we have tried to estimate the
missing data in the Algerian mortality surface.
The
completed mortality surfaces (nQx) can be downloaded
here:
Data Source:
Annual publications of the Office for
National Statistics ONS (Fiveages mortality rates) ;
missing data has been estimated by Flici (2014).
Then,
we have interpolated the detailed ages mortality rates
and extrapolated them to the age beyond 80 in Flici (2016). Thus, we obtained a
detailed continuous mortality surface extended till the age 120.
Figure
1 shows the obtained mortality surface in logarithmic scale. Figure 2 show the
same surface in a 3D scale.
Figure
1: 2D mortality surface (log mortality rates),
male population
Figure
2: Algerian Mortality Surface, 19772014,
detailed ages 0 to 120.
Source: Annual publications of the Office for
National Statistics ONS (Fiveages mortality rates) ; missing data has
been estimated by Flici (2014), Single ages mortality rates interpolated and
old age mortality rates have been extrapolated by Flici (2016b).
The Data used to draw plots 1
and 2 can be downloaded following the links:
Data Source:
Annual publications of the Office for
National Statistics ONS (Fiveages mortality rates) ; missing data has
been estimated by Flici (2014), Single ages mortality rates interpolated and old
age mortality rates have been extrapolated by Flici (2016b).
Mortality
forecasting:
Mortality
forecasting is an essential exercise since it allows to
consider the future improvement in mortality. This last, can be used as a
Population Health Index when more sophisticated tools are lacked (Health Life
Expectancy, Disability Free Life Expectancy). Also, Population projections
require projecting the age specific mortality rates in the future. In
insurance, mortality forecasting allows to take the future improvement of
mortality wellknown as “Longevity Risk” when pricing and reserving life
annuities and pensions.
Algerian
Fertility Data
Fertility
data in Algeria are mainly provided by the Office for National Statistics on
the basis of the civil registration data. This principal source is completed by
census data and specific surveys data. Our database was constituted by the five
ages fertility rates published by the Office for National Statistics (ONS) over
the periods [19641969], [19761991], [19942000] and [20082014]. In concern
of population censuses, only the last three ones (1987, 1998 and 2008) gave the
five ages fertility rates of the female population ages [1549].
This
data allowed us to constitute a discontinuous fertility surface in the sense
that some years of the observation periods are still with no data. Missing data
were estimated in FLICI(2016d), and the Age Specific
Fertility Rates ASFR’s have been interpolated by the same work. Detailed data
are presented in Figure 3.
Figure
3: Algerian Fertility Surface, detailed ages (15
– 49 years old), period: 1964 – 2014
Source: Annual publications of the Office for
National Statistics ONS (Fiveages Fertility rates):
civil registration and census data. missing data has been estimated by Flici (2016d).
The
initial database containing the five ages fertility
rates as provided by civil registration or census data can be downloaded from
the link:
The
completed Fertility surface containing the fiveages fertility
rates for the period 1964 to 2016 can be downloaded from the link:
The database containing the
ASFR’s for ages going from 15 to 49 and the period from 1964 to 2014 can be
downloaded from the link:
Data Source:
Annual publications of the Office for
National Statistics ONS (Fiveages fertility rates) ;
missing data has been estimated and Age specific Fertility rates have been
interpolated by Flici (2016d).
Population
Forecast
One of
the most important exercises in the field of quantitative demography is
population forecasting. Public planning, social policies designing and
actuarial calculations need to project the population structure by age and sex
in the future. In this sense, and in order to evaluate the financial
sustainability of the Algerian retirement system, the population of Algeria has
been projected to the horizon of 2070. Further details about the used technics
and methodology can be found in Flici (2017).
To
project the Age Specific Mortality Rates to the future, we proceeded by a
coherent mortality forecasting methodology. Throughout the application process,
it turned out that the use of the Lee Carter model (Lee and Carter, 1992) to
project mortality rates independently for males and females leads to some
incoherence regarding the sex mortality ratio by the horizon of the forecast.
The use of the coherent mortality forecasting approach allows to avoid this kind of incoherence.
In
another work (Flici, 2016c), we compared two coherent mortality models: The
product Ratio method proposed by Hyndman et al. (2013) and the LeeCarter model
with an additive common component for males and females proposed by Li and Lee
(2005). It turned out that the first model leads to better results regarding
the Goodness of fit and the coherence.
To fix
the future scenarios of the fertility rates evolution in Algeria during the
upcoming half century, we consider the views of experts rather than a
probabilistic forecast. We aim to keep the population forecast results obtained
in the present work within the framework of the National official projection
methodology. Currently, the final report has not been published yet. But the
framework of the projection was discussed and fixed in the meeting of the
population forecasts subcommittee of the National Committee for Population
(NCP) which depends from the National Health Ministry.
Discussions
have mainly focused on a comparison between the probabilistic modeling approach
and the view of experts approach. The first approach works well in the case of
mortality, but not as well for fertility. The use of the LeeCarter model as it
was adapted to forecast fertility rates (Lee, 1993) led approximately to a
constant Total Fertility Rate (TFR) at around 3 births / women by the horizon
of the forecast. Experts of the NCP judged that this scenario can be used as a
High Scenario rather than a middle one. Fertility rates are expected to keep
slightly decreasing in the coming years. The low scenario was defined in order
to respect the minimum TFR required for the population regeneration (2.1). In
final, a level of 2.5 was kept as a medium scenario by 2030. Since the
projection results was greatly depending of these underlying hypothesis which
present a kind of weakness over time, the projection's horizon was limited to
only 2030 for the National Population forecast. In order to keep working within
these hypothesis while extending the projection's
horizon until 2070 to suit the objective of the present work, we keep the same
level to be expected at 2070 rather than 2030 as in the national projection.
Figure
4: Forecasting Life expectancy at birth and
Total Fertility Rate, and Population structure ,
Algeria: 20182070
The
projection results as the projected life expectancy and total fertility rates can
be downloaded from the links:
Data Source:
Mortality forecast (FLICI, 2016c)
Fertility forecast (FLICI, 2017)
Population Projection (FLICI, 2017)
First, we have forecasted two
components: Mortality and Fertility. Because immigration's data are missed or
not available in the required format, we suppose that the immigration flow is
equal to 0. The combination of the age specific mortality and fertility rates
applied to the population structure given by the population census of 2008, we
obtained the global population evolution and its age and sex structure for the
whole period 20152075. When we observe the transformation of the population
pyramid from 2020 to 2070, we can deduce that the top of the pyramid is
expected to enlarge, compared to the basis of the pyramid. That reflects the
aging process of the Algerian population during the 50 upcoming years. The part
of the population aged 60 and over is expected to increase from around 8% in
2015 to keep around 20% starting from 2050.
Figure 7.22 lets appear the
expected decrease in the report of the population at working age, on the
population at retirement age. In 2015, we have more than 7 individuals at
working age corresponding to 1 individual at retirement age. This report is
expected to fall to 4.5 in 2030 and to stabilize around 2.7 for the period
[20502070]. This funding means that, if it is difficult to keep equilibrium
between incomes and outcomes of the pension plan under a theoretical report of
7, it will be more difficult to keep it under a value of 2.7. The main
challenge will be to make a great part of the population at working age in
occupation and then within the social security system. In the following parts,
we will try to consider these two last elements in order to define the long
term sustainability of the pension plan.
Organized
Conferences
Even
before the creation of the team of “Actuarial Demography” properly saying,
Farid FLICI initiated in January 2016 a series of conferences on actuarial
methods applied to life insurance known as JAVA (Journée
d’Actuariat Vie Algérie).
Until December 2017, 5 editions of JAVA have been organized.
JAVA
I: January 27^{th},2016 , CREAD
·
Les Tables de mortalité Comme outils de tarification et de provisionnement
en assurance vie– Farid FLICI (CREAD)
·
Construction Tables de mortalité d’expérience pour une compagnie d’assurance – Moussa MEHIRECHE
(Ministère des Finances)
·
Tables de mortalité Adaptée à l’expérience de mortalité des retraité –
Farid FLICI (CREAD) ;
JAVA II: March 7^{th,} 2016, Univ. Bouzareah.
·
Construction Tables de mortalité d’expérience pour une compagnie d’assurance – Moussa
MEHIRECHE (Ministère des Finances) ;
·
Longevity and life annuities reserving
in Algeria : comparison of mortality models – Farid FLICI (CREAD) ;
JAVA III: 7 December 7th, 2016, ENSSEA,
Kolea.
·
Santé / prévoyance collective et individuelle Garantie/règles de
souscriptions/tarification (Hamza Rabehi, Institut
des Actuaires FRANCE)
·
Financial Sustainability of the
Algerian Pension Plan (Farid FLICI, CREAD)
JAVA IV : May 25^{th}, 2017, CREAD
·
On the
heterogeneity of human population as reected by the
mortality dynamics. (Séverine ARNOLD, UNIL,
Switzerland, Via Skype)
·
Closingout
the Algerian Life tables: For more Accuracy and Adequacy at old ages (Farid
FLICI, CREAD)
JAVA V, 26 Novembre 2017, CREAD.
·
Tables de mortalité en assurance vie en Afrique Subsaharienne (Aymric Kamega, PDG d’ACAMVie, Cameroun,
Via Skype)
·
Tables de
mortalité d’expérience incorporant une échelle de projection :
l’expérience des retraités algériens (Khadidja SENOUCI, Doctorante, ENSSEA)
JAVA VI, 27 Juin 2018, CREAD.
·
A new inference strategy for general population mortality tables (Alexandre BOUMEZOUED, Milliman Paris,
Via Skype)
·
Mortality forecasting using a multifunctional LeeCarter Model :
l’expérience des retraités algériens (Mohamed D. SADOUN, CREAD and USTHB)
Achieved and ongoing research
FLICI, F ; SENOUCI, K. et HANNANI, Y. 2017. “Tables de mortalité d'expérience incorporant une échelle de projection : adaptation au cas des retraites en Algérie”, Bulletin Français d’Actuariat, 17 (34) : 532.
Les données de mortalité des retraités algériens ne sont pas disponibles pour des périodes suffisamment longues pour pouvoir y appliquer des modèles prospectifs de mortalité. Le positionnement de la mortalité d’expérience sur une référence externe est l’une des solutions techniques permettant de contourner le problème de données. Néanmoins elle est un peu compliquée du fait qu’elle nécessite de trouver une référence externe adéquate et d’en faire des projections avant de pouvoir projeter la mortalité d’expérience. Dans ce travail, nous proposons une méthode plus simple et aussi performante que la première méthode. Partant d’une table d’expérience périodique de retraités à laquelle on applique une échelle de projection, nous parviendrons à projeter la mortalité d’expérience. Les résultats montrent que les hommesretraités ne présentent aucun avantage significatif comparativement au reste de la population. Par contre, à 50 ans, les femmes retraitées peuvent espérer vivre 3 ans plus longtemps que les femmes du reste de la population.
Motsclés: mortalité, retraite, expérience, échelle d’améli oration, Algérie.
FLICI, F.
and HAMMOUDA, NE. 2014. “Analysis of halfcentury of mortality
evolution in Algeria: 1992 2012”. Conference of the Middle
East Economics Association MEEA. June 2014. Tlemcen,
Algeria.
Since the independence, the Algerian
population has earned almost 30 years in life expectancy at birth. However,
this improvement was accompanied by changes in the estimation methodology and improvement
in vital statistics quality. Thus, we can not simply
analyze mortality evolution in Algeria without addressing all the necessary
caution in regards to the methodological changes effects. Our objective is to
give a summary presentation of these methodological changes and data
imperfections sources before to achieve a relative analysis of the life
expectancy evolution. This analysis considers the data published by the Office
of National Statistics (ONS) from 1962 to 2012.
Keywords: mortality, life expectancy, changepoints, civil records, Algeria.
FLICI, F. 2014. "Estimation
of the missing life tables in the Algerian mortality Surface (1977 1999) by
using the LeeCarter formula”. SMTDA2014.
June 2014, Lisbon. Portugal.
Abstract:
Mortality forecasting needs usually the availability of a continuous mortality surfaces.
In Algeria, the first life table based on the civil registration data has been
constructed in 1977. At that time, the life tables could not be constructed
with an annual
frequency because of the lack of data and experience. It became possible only starting from 1998. No one has tried to estimate the missing life tables of the period [19771998]. Today, it is necessary to make this data available and that’s the
main objective of the present paper. The idea is to use the LeeCarter model (Lee and Carter, 1992) to estimate this missing data. For this, the model parameters
were firstly estimated on the available
data. Then, the time component will be fitted; the fitting function will be
used to estimate the missing values. Combined with the age parameters, that
leads to estimate the missing mortality rates for males, females and both sex
populations while improving the adequacy compared to the original data.
Keywords: Mortality Rates, Missing data, polynomial
fit, Lee Carter, Algeria.
FLICI, F. 2016. “Provisionnement des
rentes viagères en Algérie entre approche statique et approche prospective”, Bulletin Français d’Actuariat, 16 (31) : 540.
Résumé :
L’espérance
de vie à la naissance de la population Algérienne ne cesse de réaliser des
améliorations considérables. Cette longévité combinée avec les nouvelles
contraintes du marché algérien des rentes viagères ne fait que réduire
l’utilité des tables de mortalité statiques et accentuent la nécessité
d’utiliser les tables de mortalité prospectives pour la tarification et le
provisionnement. En l’absence de données de mortalité spécifiques au marché des
rentes viagères, nous proposons, dans le présent travail, de nous baser
directement sur les données de la population globale afin de construire une
table de mortalité prospective pour la tarification et le provisionnement des
rentes viagères en Algérie. Les résultats obtenus sous l’approche prospective
sont comparés à ceux obtenus sous l’approche statique des tables de mortalité.
Motsclefs : Provisionnement,
rentes viagères, longévité, table de mortalité statiques, table de mortalité
prospectives, Algérie.
FLICI, F.
2016. “Closing out the Algerian life tables: for more accuracy and adequacy at
old ages’’, Proceeding of the International Association of
Actuaries – ASTIN Colloquium, 2016, LisbonPortugal.
Abstract:
Giving data
unavailability or irregularity beyond a certain age, particularly in developing
countries, the model life tables are an unavoidable solution to estimate the
old ages mortality. As an international standard, this tool provides
approximate estimations which are not specified for a particular country, thus
the obtained results are not always satisfactory especially when these model
life tables are not used in the right way. Thereby, estimating the old ages
mortality by extrapolating the observed trend at the younger ages is assumed to
provide more coherent results. In the present paper, we evaluate and compare
some proposed old ages mortality models to extend mortality rates beyond the
age of 80 for the Algerian population. The comparison will be based on various
criteria: Goodnessoffit, predictive capacity, sex differential coherence,
predicted age limit and single sexes vs bothsexes adequacy. The final results
will be used to correct the historical series of life expectancy at birth for
Algeria between 1977 and 2014.
Keywords: mortality
rates, goodness of fit, predictive capacity, BIC, MSE, extrapolation, life
expectancy, Algeria.
FLICI, F. 2015. “Life annuities calculation in Algeria: Continuous time approach”. Journal
of Statistical Science and Application, April 2015, Vol. 3, No. 56, 85100.
Abstract:
The present paper aims to show the impact of continuous time
calculation on life insurance pricing and reserving in the Algerian context.
The discrete time approach allows insurance companies to facilitate calculation
process but with less accuracy. This approach implies constancy of death
quotients during a year. However, the death risk is a continuous function in
time. For more accuracy and equity in pricing, calculation needs to consider
the exact dates of different payments and also a continuous capitalization
process. This gives more adequate premium with fewer hypotheses. This work
shows how insurers can propose more adequate pricing using the same actuarial
life table.
Keywords: Life annuities, life table, continuous
time, fitting, extrapolating, Algeria.
FLICI, F. 2015. "Mortality forecasting for the Algerian population with considering cohort
effect”. Proceeding of the
International Association of Actuaries. IAAC 2015.
OsloNorway.
Mortality forecasting becomes a challenge not
only for demographers but also for actuaries. Different models have been
proposed for this issue ensuring both efficiency and simplicity. In the
beginning, most of the proposed models were based on the time and age
dimensions. The observation of the error’s distribution in the case of some
populations revealed that some cohorts are not well represented by the
AgePeriod approach. The idea was to introduce a cohort component in order to
improve the goodnessoffit and the forecasting capacity. In the present paper,
we propose to forecast the specificage mortality rates for the Algerian
population by comparing the AgePeriod Approach (LeeCarter model) and the
AgePeriodCohort approach (Renshaw and Haberman, 2006; Currie, 2006).
Keywords: Mortality forecasting, Cohort, fitting, Algeria,
life annuities.
FLICI, F. 2016. “Longevity and
Life Annuities Reserving in Algeria: Comparison of mortality models”.
International Actuarial Association – Life Section Conference, Hong Kong, April
2016. (Abstract)
Abstract:
Using prospective life tables for pension plan managing
and life annuities reserving is now more than
essential especially for developing countries for which life expectancy keeps
improving. Algeria takes part of these countries and it is more than important
to provide the life insurance sector a dynamic life table to be used for
actuarial calculations. Here, we aim to achieve this objective by comparing a
set of mortality models, principally LeeCarter and CairnsBlakeDows models
with variants. To do this, we will use the Algerian mortality surfaces; for
males and females, for the period [19772013] and for the age range [50  79].
Models evaluation and selection will be principally based on the
goodnessoffit, and performed by introducing complementary criteria: the
predictive capacity, the male female coherence, and also the expected gap in
life expectancy between males and females.
Keywords: Prospective mortality models, goodnessoffit,
coherence, forecasting.
FLICI, F. 2016. “Coherent
Mortality Forecasting for the Algerian Population”. SAMOS Conference in Actuarial Sciences and Finance ASF2016.
Samos, Greece, May 2016.
Abstract:
Mortality
forecasting is much needed for population projections and actuarial prospective
calculations. The independent single sex’s mortality forecasting leads in
almost cases to some incoherence regarding the expected male female mortality evolution.
To avoid any unrealistic convergence or divergence in this sense, a coherent
mortality forecasting is needed. In the present paper, we do a coherent
mortality forecasting for the male female Algerian populations by comparing the
two main proposed approaches: Li and Lee (2005) and Hyndman and al. (2013).
Keywords: Life expectancy, mortality,
coherent forecast, sex ratio, Algeria.
FLICI,
F. and PLANCHET, F. 2016. “Construction of a dynamic life table based on the Algerian retired
population mortality experience”. Mathematical and Statistical Methods for Actuarial Sciences and
Finance conference MAF2016, Paris, France, April 2016.
The aim of the
present paper is to construct a prospective life table adapted to the mortality
experience of the Algerian retired. Mortality data for the retired population
is only available for the age range [4595[ and for
the period [20042013]. The use of the conventional prospective mortality
models is not supposed to lead to a robust mortality forecast given the
shortness of the lenght of the historical data. To
ensure a robust forecasting, we use the global population mortality as an
external reference. The positionning of the
experience mortality on the reference will allow to project
the Age Specific Death Rates calculated on the basis of the experience of the
retired population.
keywords: Prospective lifetables, Experience lifetables,
reference mortality, retired population, Algeria.
FLICI, F. and PLANCHET, F. 2018. “Financial Sustainability of the Algerian Retirement System: a perspective analysis of the 50 coming years”, Presented at The International Pension Workshop IPW, 2324 April 2018, Lisbon, Portugal. (Full Paper). & at the 31st International Congress of Actuaries ICA, 48 June 2018, Berlin Germany (Talk record).
Abstract:
Maintaining retirement systems stability becomes a big challenge
for all countries. Such sustainability is widely related to a set of elements
changing over time as population structure, longevity, employment and
affiliation to social security. For the case of Algeria, public retirement
works according to the payasyougo principle and equilibrium is maintained by
public subsidy. But, population is now aging, and longevity is improving. The
retired population is supposed to grow faster than the population at working age.
Consequently, it will be hard to keep equality between retirement incomes and
outcomes in such conditions especially if we consider the weakness of the
insured population in respect to the global population. The part of this last
depends largely of public employment; the private sector is still less covered.
In the other hand, a set of elements are still defined by the social and
economic context as wages, pension benefit amount and contribution rates. In
the present paper, we present a general review of the different variables
affecting the stability of the Algerian system and then we do a simulation
about the future evolution. In final, we give a set of recommendations in order
to strengthen the stability of the Algerian Retirement System during the 50
coming years.
Keywords:
Pension plan, payasyougo, experience mortality, contribution rates, aging,
longevity.
References
FLICI, F. 2017. FLICI, F. (2017).
“Longevity and pension plan sustainability in Algeria :
Taking the retirees mortality experience into account”, Doctoral thesis, Ecole Nationale Supérieure de Statistique et d'Economie Appliquée (ENSSEA),
Algiers, Algeria.
FLICI, F. 2016a. “Provisionnement
des rentes viagères en Algérie entre approche statique et approche prospective”, Bulletin Français d’Actuariat, 16 (31) : 540.
FLICI, F. 2016b. “Closing out
the Algerian life tables: for more accuracy and adequacy at old ages’’, Proceeding of the International Association of Actuaries
– ASTIN Colloquium, 2016, LisbonPortugal.
FLICI, F. 2016c. “Coherent Mortality Forecasting
for the Algerian Population”. SAMOS Conference in Actuarial Sciences and Finance
ASF2016. Samos, Greece, May 2016.
FLICI, F. 2016d. “Projection des taux de fécondité par âge de la
population Algériennes”. Working paper. Cread.
FLICI, F. and PLANCHET, F. 2016. “Financial Sustainability of the
Algerian Pension System”. 3° Journée d’Actuariat Vie Algérie – JAVA3, Kolea,
Algeria, December 2016.
FLICI, F. 2014. "Estimation of the missing life
tables in the Algerian mortality Surface (1977 1999) by using the LeeCarter
formula”. SMTDA2014. June 2014, Lisbon. Portugal.
FLICI, F. and HAMMOUDA, NE. 2014. “Analysis of halfcentury of mortality evolution in Algeria: 1992
2012”. Conference of the Middle East Economics Association
MEEA. June 2014. Tlemcen, Algeria.
HYNDMAN,
R. BOOTH, H. and YASMEEN, F. (2013). Coherent mortality forecasting: the productratio method with functional time
series models. Demography, 50(1) : 261 283.
LEE, R.
(1993). Modeling and forecasting the time series of US fertility: Age
distribution, ranges, and ultimate level. International Journal of Forecasting,
9: 187202.
LEE,
R. and CARTER, L. (1992).
Modeling and Forecasting U. S. Mortality. Journal of
the American Statistical Association, 87 (419) :
659671.
LI,
N. and LEE, R.D. (2005).
Coherent mortality forecasts for a group of populations: an extension of the Lee Carter method. Demography,
42 (3): 575594.