JSS Journal of Statistical Software

March 2008, Volume 25, Issue 7. http://www.jstatsoft.org/

actuar: An R Package for Actuarial Science

Christophe Dutang Vincent Goulet Mathieu Pigeon

Université Claude Bernard Lyon 1 Université Laval Université Laval

Abstract
actuar is a package providing additional Actuarial Science functionality to the R sta-
tistical system. The project was launched in 2005 and the package is available on the
Comprehensive R Archive Network since February 2006. The current version of the pack-
age contains functions for use in the fields of loss distributions modeling, risk theory
(including ruin theory), simulation of compound hierarchical models and credibility the-
ory. This paper presents in detail but with few technical terms the most recent version of
the package.

Keywords: insurance, loss distributions, risk theory, ruin theory, credibility theory, distribu-
tions, phase-type, matrix exponential, simulation, hierarchical, R.

1. Introduction
actuar is a package providing additional Actuarial Science functionality to the R statistical
system (R Development Core Team 2008). Various packages on the Comprehensive R Archive
Network (CRAN, http://CRAN.R-project.org/) provide functions useful to actuaries, e.g.,
copula (Yan 2007), Rmetrics (Wuertz 2007), SuppDists (Wheeler 2008) or the R recommended
packages nlme (Pinheiro, Bates, DebRoy, Sarkar, and the R Development Core Team 2007)
and survival (Lumley 2008) just to name a few. However, actuar aims to serve as a central
location for more specifically actuarial functions and data sets. The project was officially
launched in 2005 and is under active development.
The feature set of the package can be split in four main categories: loss distributions modeling,
risk theory (including ruin theory), simulation of compound hierarchical models and credibility
theory. As much as possible, the developers have tried to keep the “user interface” of the
various functions of the package consistent. Moreover, the package follows the general R
philosophy of working with model objects.
This paper reviews the various features of version 0.9-7 of actuar. We provide enough actuarial
2 actuar: An R Package for Actuarial Science

background where needed to make the paper self-contained, but otherwise give numerous
references for the reader interested to dive more into the subject.
Future versions of the package can be obtained from CRAN at http://CRAN.R-project.
org/package=actuar.

2. Loss distributions modeling

One important task of actuaries is the modeling of claim amount distributions for ratemaking,
loss reserving or other risk evaluation purposes. Package actuar offers many functions for loss
distributions modeling. The present section details the following actuar features:

1. introduction of 18 additional probability laws and utility functions to get raw moments,
limited moments and the moment generating function;

2. fairly extensive support of grouped data;

3. calculation of the empirical raw and limited moments;

4. minimum distance estimation using three different measures;

5. treatment of coverage modifications (deductibles, limits, inflation, coinsurance).

2.1. Probability laws

R already includes functions to compute the probability density function (pdf), the cumulative
distribution function (cdf) and the quantile function of a fair number of probability laws, as
well as functions to generate variates from these laws. For some root foo , the utility functions
are named dfoo , pfoo , qfoo and rfoo , respectively.
The actuar package provides d, p, q and r functions for all the probability laws useful for
loss severity modeling found in Appendix A of Klugman, Panjer, and Willmot (2004) and
not already present in base R, excluding the inverse Gaussian and log-t but including the
loggamma distribution (Hogg and Klugman 1984). Table 1 lists the supported distributions
as named in Klugman et al. (2004) along with the root names of the R functions.
In addition to the d, p, q and r functions, the package provides m, lev and mgf functions to
compute, respectively, theoretical raw moments

mk = E[X k ], (1)

theoretical limited moments

E[(X ∧ x)k ] = E[min(X, x)k ] (2)

and the moment generating function

MX (t) = E[etX ], (3)

when it exists. Every probability law of Table 1 is supported, plus the following ones: beta,
exponential, chi-square, gamma, lognormal, normal (no lev), uniform and Weibull of base R
 Journal of Statistical Software 3

Family Distribution Root

Transformed beta Transformed beta trbeta
Burr burr
Loglogistic llogis
Paralogistic paralogis
Generalized Pareto genpareto
Pareto pareto
Inverse Burr invburr
Inverse Pareto invpareto
Inverse paralogistic invparalogis
Transformed gamma Transformed gamma trgamma
Inverse transformed gamma invtrgamma
Inverse gamma invgamma
Inverse Weibull invweibull
Inverse exponential invexp
Other Loggamma lgamma
Single parameter Pareto pareto1
Generalized beta genbeta

Table 1: Probability laws supported by actuar classified by family and root names of the R
functions.

and the inverse Gaussian distribution of the package SuppDists (Wheeler 2008). The m and
lev functions are especially useful with estimation methods based on the matching of raw or
limited moments; see Section 2.4 for their empirical counterparts. The mgf functions come in
handy to compute the adjustment coefficient in ruin theory; see Section 3.5.
In addition to the 17 distributions of Table 1, the package provides support for a family
of distributions deserving a separate presentation. Phase-type distributions (Neuts 1981) are
defined as the distribution of the time until absorption of continuous time, finite state Markov
processes with m transient states and one absorbing state. Let
 
T t
Q= (4)
0 0

be the transition rates (or intensity) matrix of such a process and let (π, πm+1 ) be the initial
probability vector. Here, T is an m × m non-singular matrix with tii < 0 for i = 1, . . . , m and
tij ≥ 0 for i 6= j, t = −T e and e is a column vector with all components equal to 1. Then
the cdf of the time until absorption random variable with parameters π and T is
(
πm+1 , x=0
F (x) = T x
(5)
1 − πe e, x > 0,

where
∞
X Mn
eM = (6)
n!
n=0
is the matrix exponential of matrix M .
4 actuar: An R Package for Actuarial Science

The exponential, the Erlang (gamma with integer shape parameter) and discrete mixtures
thereof are common special cases of phase-type distributions.
The package provides functions {d,p,r,m,mgf}phtype for phase-type distributions. Function
pphtype is central to the evaluation of ruin probabilities; see Section 3.6.
The core of all the functions presented in this subsection is written in C for speed. The matrix
exponential C routine is based on expm() from the package Matrix (Bates and Maechler 2008).

2.2. Grouped data

What is commonly referred to in Actuarial Science as grouped data is data represented in an
interval-frequency manner. In insurance applications, a grouped data set will typically report
that there were nj claims in the interval (cj−1 , cj ], j = 1, . . . , r (with the possibility that
cr = ∞). This representation is much more compact than an individual data set — where
the value of each claim is known — but it also carries far less information. Now that storage
space in computers has almost become a non issue, grouped data has somewhat fallen out of
fashion.
Still, grouped data remains in use in some fields of actuarial practice and also of interest
in teaching. For this reason, actuar provides facilities to store, manipulate and summarize
grouped data. A standard storage method is needed since there are many ways to represent
grouped data in the computer: using a list or a matrix, aligning the nj s with the cj−1 s or
with the cj s, omitting c0 or not, etc. Moreover, with appropriate extraction, replacement and
summary methods, manipulation of grouped data becomes similar to that of individual data.
First, function grouped.data creates a grouped data object similar to — and inheriting from
— a data frame. The input of the function is a vector of group boundaries c0 , c1 , . . . , cr and
one or more vectors of group frequencies n1 , . . . , nr . Note that there should be one group
boundary more than group frequencies. Furthermore, the function assumes that the intervals
are contiguous. For example, the following data

Group Frequency (Line 1) Frequency (Line 2)

(0, 25] 30 26
(25, 50] 31 33
(50, 100] 57 31
(100, 150] 42 19
(150, 250] 65 16
(250, 500] 84 11

is entered and represented in R as

R> x <- grouped.data(Group = c(0, 25, 50, 100, 150, 250,

+ 500), Line.1 = c(30, 31, 57, 42, 65, 84), Line.2 = c(26,
+ 33, 31, 19, 16, 11))

Object x is stored internally as a list with class

R> class(x)

[1] "grouped.data" "data.frame"

 Journal of Statistical Software 5

With a suitable print method, these objects can be displayed in an unambiguous manner:

R> x

Group Line.1 Line.2

1 (0, 25] 30 26
2 (25, 50] 31 33
3 (50, 100] 57 31
4 (100, 150] 42 19
5 (150, 250] 65 16
6 (250, 500] 84 11

Second, the package supports the most common extraction and replacement methods for
"grouped.data" objects using the usual [ and [<- operators. In particular, the following
extraction operations are supported.

(i) Extraction of the vector of group boundaries (the first column):

R> x[, 1]

[1] 0 25 50 100 150 250 500

(ii) Extraction of the vector or matrix of group frequencies (the second and third columns):

R> x[, -1]

Line.1 Line.2
1 30 26
2 31 33
3 57 31
4 42 19
5 65 16
6 84 11

(iii) Extraction of a subset of the whole object (first three lines):

R> x[1:3, ]

Group Line.1 Line.2

1 (0, 25] 30 26
2 (25, 50] 31 33
3 (50, 100] 57 31

Notice how extraction results in a simple vector or matrix if either of the group boundaries
or the group frequencies are dropped.
As for replacement operations, the package implements the following.

(i) Replacement of one or more group frequencies:

6 actuar: An R Package for Actuarial Science

R> x[1, 2] <- 22

R> x

Group Line.1 Line.2

1 (0, 25] 22 26
2 (25, 50] 31 33
3 (50, 100] 57 31
4 (100, 150] 42 19
5 (150, 250] 65 16
6 (250, 500] 84 11

R> x[1, c(2, 3)] <- c(22, 19)

R> x

Group Line.1 Line.2

1 (0, 25] 22 19
2 (25, 50] 31 33
3 (50, 100] 57 31
4 (100, 150] 42 19
5 (150, 250] 65 16
6 (250, 500] 84 11

(ii) Replacement of the boundaries of one or more groups:

R> x[1, 1] <- c(0, 20)

R> x

Group Line.1 Line.2

1 (0, 20] 22 19
2 (20, 50] 31 33
3 (50, 100] 57 31
4 (100, 150] 42 19
5 (150, 250] 65 16
6 (250, 500] 84 11

R> x[c(3, 4), 1] <- c(55, 110, 160)

R> x

Group Line.1 Line.2

1 (0, 20] 22 19
2 (20, 55] 31 33
3 (55, 110] 57 31
4 (110, 160] 42 19
5 (160, 250] 65 16
6 (250, 500] 84 11

It is not possible to replace the boundaries and the frequencies simultaneously.

 Journal of Statistical Software 7

Histogram of x[, −3]

0.003
0.002
Density

0.001
0.000

0 100 200 300 400 500

x[, −3]

Figure 1: Histogram of a grouped data object

The package defines methods of a few existing summary functions for grouped data objects.
Computing the mean
r  
X cj−1 + cj
nj (7)
2
j=1

is made simple with a method for the mean function:

R> mean(x)

Line.1 Line.2
188.0 108.2

Higher empirical moments can be computed with emm; see Section 2.4.
The R function hist splits individual data into groups and draws an histogram of the fre-
quency distribution. The package introduces a method for already grouped data. Only the
first frequencies column is considered (see Figure 1 for the resulting graph):

R> hist(x[, -3])

R has a function ecdf to compute the empirical cdf of an individual data set,
n
1X
Fn (x) = I{xj ≤ x}, (8)
n
j=1
8 actuar: An R Package for Actuarial Science

where I{A} = 1 if A is true and I{A} = 0 otherwise. The function returns a "function"
object to compute the value of Fn (x) in any x.
The approximation of the empirical cdf for grouped data is called an ogive (Klugman, Panjer,
and Willmot 1998; Hogg and Klugman 1984). It is obtained by joining the known values of
Fn (x) at group boundaries with straight line segments:



0, x ≤ c0
 (c − x)F (c ) + (x − c )F (c )
j n j−1 j−1 n j
F̃n (x) = , cj−1 < x ≤ cj (9)

 cj − cj−1

1, x > cr .

The package includes a function ogive that otherwise behaves exactly like ecdf. In particular,
methods for functions knots and plot allow, respectively, to obtain the knots c0 , c1 , . . . , cr of
the ogive and a graph (see Figure 2):

R> Fnt <- ogive(x)

R> knots(Fnt)

[1] 0 20 55 110 160 250 500

R> Fnt(knots(Fnt))

[1] 0.00000 0.07309 0.17608 0.36545 0.50498 0.72093 1.00000

R> plot(Fnt)

2.3. Data sets

This is certainly not the most spectacular feature of actuar, but it remains useful for illus-
trations and examples: the package includes the individual dental claims and grouped dental
claims data of Klugman et al. (2004):

R> data("dental")
R> dental

[1] 141 16 46 40 351 259 317 1511 107 567

R> data("gdental")
R> gdental

cj nj
1 (0, 25] 30
2 ( 25, 50] 31
3 ( 50, 100] 57
4 (100, 150] 42
5 (150, 250] 65
 Journal of Statistical Software 9

ogive(x)

1.0
●

0.8

●
0.6
F(x)

●
0.4

●
0.2

●
0.0

0 100 200 300 400 500

Figure 2: Ogive of a grouped data object

6 (250, 500] 84
7 (500, 1000] 45
8 (1000, 1500] 10
9 (1500, 2500] 11
10 (2500, 4000] 3

2.4. Calculation of empirical moments

The package provides two functions useful for estimation based on moments. First, function
emm computes the kth empirical moment of a sample, whether in individual or grouped data
form:

R> emm(dental, order = 1:3)

[1] 3.355e+02 2.931e+05 3.729e+08

R> emm(gdental, order = 1:3)

[1] 3.533e+02 3.577e+05 6.586e+08

Second, in the same spirit as ecdf and ogive, function elev returns a function to compute
the empirical limited expected value — or first limited moment — of a sample for any limit.
Again, there are methods for individual and grouped data (see Figure 3 for the graphs):
10 actuar: An R Package for Actuarial Science

elev(x = dental) elev(x = gdental)

● ● ●
300

300
●
250

●
●
Empirical LEV

Empirical LEV
200

200
●
●

● ●
150

100
100

●
●
●

50
●
50

●
● ●
● ●

0
0 500 1000 1500 0 1000 2000 3000 4000

x x

Figure 3: Empirical limited expected value function of an individual data object (left) and a
grouped data object (right)

R> lev <- elev(dental)

R> lev(knots(lev))

[1] 16.0 37.6 42.4 85.1 105.5 164.5 187.7 197.9 241.1 335.5

R> plot(lev, type = "o", pch = 19)

R> lev <- elev(gdental)
R> lev(knots(lev))

[1] 0.00 24.01 46.00 84.16 115.77 164.85 238.26 299.77 324.90
[10] 347.39 353.34

R> plot(lev, type = "o", pch = 19)

2.5. Minimum distance estimation

Two methods are widely used by actuaries to fit models to data: maximum likelihood and
minimum distance. The first technique applied to individual data is well covered by function
fitdistr of the package MASS (Venables and Ripley 2002). The second technique minimizes
a chosen distance function between theoretical and empirical distributions. Package actuar
provides function mde, very similar in usage and inner working to fitdistr, to fit models
according to any of the following three distance minimization methods.

1. The Cramér-von Mises method (CvM) minimizes the squared difference between the
theoretical cdf and the empirical cdf or ogive at their knots:
n
X
d(θ) = wj [F (xj ; θ) − Fn (xj ; θ)]2 (10)
j=1
 Journal of Statistical Software 11

for individual data and

r
X
d(θ) = wj [F (cj ; θ) − F̃n (cj ; θ)]2 (11)
j=1

for grouped data. Here, F (x) is the theoretical cdf of a parametric family, Fn (x) is
the empirical cdf, F̃n (x) is the ogive and w1 ≥ 0, w2 ≥ 0, . . . are arbitrary weights
(defaulting to 1).

2. The modified chi-square method (chi-square) applies to grouped data only and mini-
mizes the squared difference between the expected and observed frequency within each
group:
Xr
d(θ) = wj [n(F (cj ; θ) − F (cj−1 ; θ)) − nj ]2 , (12)
j=1
Pr
where n = j=1 nj . By default, wj = n−1
j .

3. The layer average severity method (LAS) applies to grouped data only and minimizes the
squared difference between the theoretical and empirical limited expected value within
each group:
Xr
d(θ) = ˜ n (cj−1 , cj ; θ)]2 ,
wj [LAS(cj−1 , cj ; θ) − LAS (13)
j=1

˜ n (x, y) = Ẽn [X ∧ y] − Ẽn [X ∧ x] and Ẽn [X ∧ x]

where LAS(x, y) = E[X ∧ y] − E[X ∧ x], LAS
is the empirical limited expected value for grouped data.

The arguments of mde are a data set, a function to compute F (x) or E[X ∧ x], starting values
for the optimization procedure and the name of the method to use. The empirical functions
are computed with ecdf, ogive or elev.
The expressions below fit an exponential distribution to the grouped dental data set, as per
Example 2.21 of Klugman et al. (1998):

R> mde(gdental, pexp, start = list(rate = 1/200), measure = "CvM")

rate
0.003551

distance
0.002842

R> mde(gdental, pexp, start = list(rate = 1/200), measure = "chi-square")

rate
0.00364

distance
13.54
12 actuar: An R Package for Actuarial Science

Coverage modification Per-loss variable (Y L ) Per-payment variable (Y P )

(
0, X≤d n
Ordinary deductible (d) X − d, X>d
X − d, X > d
(
0, X ≤ d n
Franchise deductible (d) X, X>d
X, X > d
( (
X, X ≤ u X, X≤u
Limit (u)
u, X > u u, X>u

Coinsurance (α) αX αX
Inflation (r) (1 + r)X (1 + r)X

Table 2: Coverage modifications for the per-loss and per-payment variables as defined in
Klugman et al. (2004).

R> mde(gdental, levexp, start = list(rate = 1/200), measure = "LAS")

rate
0.002966

distance
694.5

2.6. Coverage modifications

Let X be the random variable of the actual claim amount for an insurance policy, Y L be the
random variable of the amount paid per loss and Y P be the random variable of the amount
paid per payment. The terminology for the last two random variables refers to whether or
not the insurer knows that a loss occurred. Now, the random variables X, Y L and Y P will
differ if any of the following coverage modifications are present for the policy: an ordinary or
a franchise deductible, a limit, coinsurance or inflation adjustment (see Klugman et al. 2004,
Chapter 5 for precise definitions of these terms). Table 2 summarizes the definitions of Y L
and Y P .
Often, one will want to use data Y1L , . . . , YnL (or Y1P , . . . , YnP ) from the random variable Y L
(Y P ) to fit a model on the unobservable random variable X. This requires expressing the
pdf or cdf of Y L (Y P ) in terms of the pdf or cdf of X. Function coverage of actuar does
just that: given a pdf or cdf and any combination of the coverage modifications mentioned
above, coverage returns a function object to compute the pdf or cdf of the modified random
variable. The function can then be used in modeling like any other dfoo or pfoo function.
For example, let Y P represent the amount paid by an insurer for a policy with an ordinary
deductible d and a limit u − d (or maximum covered loss of u). Then the definition of Y P is
(
P X − d, d ≤ X ≤ u
Y = (14)
u − d, X ≥ u
 Journal of Statistical Software 13

and its pdf is 



0, y=0

 f X (y + d)
, 0<y <u−d


1 − F (d)

fY P (y) = X (15)
1 − FX (u)
, y =u−d


1 − FX (d)




y > u − d.

0,

Assume X has a gamma distribution. Then an R function to compute the pdf (15) in any y
for a deductible d = 1 and a limit u = 10 is obtained with coverage as follows:

R> f <- coverage(pdf = dgamma, cdf = pgamma, deductible = 1,

+ limit = 10)
R> f

function (x, shape, rate = 1, scale = 1/rate)

ifelse(x == 0, 0, ifelse(0 < x & x < 9, do.call("dgamma", list(
x + 1, shape = shape, rate = rate, scale = scale))/do.call("pgamma",
list(1, shape = shape, rate = rate, scale = scale, lower.tail = FALSE)),
ifelse(x == 9, do.call("pgamma", list(10, shape = shape,
rate = rate, scale = scale, lower.tail = FALSE))/do.call("pgamma",
list(1, shape = shape, rate = rate, scale = scale,
lower.tail = FALSE)), 0)))
<environment: 0x1bab75c>

R> f(0, shape = 5, rate = 1)

[1] 0

R> f(5, shape = 5, rate = 1)

[1] 0.1343

R> f(9, shape = 5, rate = 1)

[1] 0.02936

R> f(12, shape = 5, rate = 1)

[1] 0

Note how function f is built specifically for the coverage modifications submitted and contains
as little useless code as possible. For comparison purpose, the following function contains no
deductible and no limit:

R> g <- coverage(dgamma, pgamma)

R> g
14 actuar: An R Package for Actuarial Science

function (x, shape, rate = 1, scale = 1/rate)

ifelse(x == 0, 0, ifelse(0 < x & x < Inf, do.call("dgamma", list(
x, shape = shape, rate = rate, scale = scale)), ifelse(x ==
Inf, 0, 0)))
<environment: 0x15b8b64>

The vignette "coverage" contains more detailed pdf and cdf formulas under various combi-
nations of coverage modifications.

3. Risk theory
Risk theory refers to a body of techniques to model and measure the risk associated with a
portfolio of insurance contracts. A first approach consists in modeling the distribution of total
claims over a fixed period of time using the classical collective model of risk theory. A second
input of interest to the actuary is the evolution of the surplus of the insurance company over
many periods of time. In ruin theory, the main quantity of interest is the probability that the
surplus becomes negative, in which case technical ruin of the insurance company occurs.
The interested reader can find more on these subjects in Klugman et al. (2004); Gerber (1979);
Denuit and Charpentier (2004); Kaas, Goovaerts, Dhaene, and Denuit (2001), among others.
The current version of actuar contains four visible functions related to the above problems:
two for the calculation of the aggregate claim amount distribution and two for ruin probability
calculations.
We briefly expose the underlying models before we introduce each set of functions.

3.1. The collective risk model

Let random variable S represent the aggregate claim amount (or total amount of claims) of
a portfolio of independent risks over a fixed period of time, random variable N represent the
number of claims (or frequency) in the portfolio over that period, and random variable Cj
represent the amount of claim j (or severity). Then, we have the random sum

S = C1 + · · · + CN , (16)

where we assume that C1 , C2 , . . . are mutually independent and identically distributed random
variables each independent of N . The task at hand consists in evaluating numerically the cdf
of S, given by

FS (x) = P[S ≤ x]
X∞
= P[S ≤ x|N = n]pn
n=0
X∞
= FC∗n (x)pn , (17)
n=0

where FC (x) = P[C ≤ x] is the common cdf of C1 , . . . , Cn , pn = P[N = n] and FC∗n (x) =
 Journal of Statistical Software 15

P[C1 + · · · + Cn ≤ x] is the n-fold convolution of FC (·). If C is discrete on 0, 1, 2, . . . , one has


I{x ≥ 0},
 k=0
∗k
FC (x) = FC (x), k=1 (18)
Px ∗(k−1)

y=0 FC (x − y)fC (y), k = 2, 3, . . .

3.2. Discretization of claim amount distributions

Some numerical techniques to compute the aggregate claim amount distribution (see Sec-
tion 3.3) require a discrete arithmetic claim amount distribution; that is, a distribution defined
on 0, h, 2h, . . . for some step (or span, or lag) h. The package provides function discretize
to discretize a continuous distribution. (The function can also be used to modify the support
of an already discrete distribution, but this requires additional care.)
Let F (x) denote the cdf of the distribution to discretize on some interval (a, b) and fx denote
the probability mass at x in the discretized distribution. Currently, discretize supports the
following four discretization methods.

1. Upper discretization, or forward difference of F (x):

fx = F (x + h) − F (x) (19)

for x = a, a + h, . . . , b − h. The discretized cdf is always above the true cdf.

2. Lower discretization, or backward difference of F (x):

(
F (a), x=a
fx = (20)
F (x) − F (x − h), x = a + h, . . . , b.

The discretized cdf is always under the true cdf.

3. Rounding of the random variable, or the midpoint method:

(
F (a + h/2), x=a
fx = (21)
F (x + h/2) − F (x − h/2), x = a + h, . . . , b − h.

The true cdf passes exactly midway through the steps of the discretized cdf.

4. Unbiased, or local matching of the first moment method:

E[X ∧ a] − E[X ∧ a + h]


 + 1 − F (a), x=a

 h
2E[X ∧ x] − E[X ∧ x − h] − E[X ∧ x + h]

fx = , a<x<b (22)
 h
 E[X ∧ b] − E[X ∧ b − h] − 1 + F (b),



x = b.
h
The discretized and the true distributions have the same total probability and expected
value on (a, b).
16 actuar: An R Package for Actuarial Science

Upper Lower

● ●
● ●
● ●
0.8

0.8
● ●
F(x)

F(x)
● ●
0.4

0.4
0.0

0.0
●

0 1 2 3 4 5 0 1 2 3 4 5

x x

Rounding Unbiased

● ● ●
● ●
0.8

0.8

● ●

● ●
F(x)

F(x)
0.4

0.4

● ●
0.0

0.0

0 1 2 3 4 5 0 1 2 3 4 5

x x

Figure 4: Comparison of four discretization methods

Figure 4 illustrates the four methods. It should be noted that although very close in this
example, the rounding and unbiased methods are not identical.
Usage of discretize is similar to R’s plotting function curve. The cdf to discretize and,
for the unbiased method only, the limited expected value function are passed to discretize
as expressions in x. The other arguments are the upper and lower bounds of the discretiza-
tion interval, the step h and the discretization method. For example, upper and unbiased
discretizations of a Gamma(2, 1) distribution on (0, 17) with a step of 0.5 are achieved with,
respectively,

R> fx <- discretize(pgamma(x, 2, 1), method = "upper",

+ from = 0, to = 17, step = 0.5)
R> fx <- discretize(pgamma(x, 2, 1), method = "unbiased",
+ lev = levgamma(x, 2, 1), from = 0, to = 17, step = 0.5)

Function discretize is written in a modular fashion making it simple to add other discretiza-
tion methods if needed.
 Journal of Statistical Software 17

3.3. Calculation of the aggregate claim amount distribution

Function aggregateDist serves as a unique front end for various methods to compute or ap-
proximate the cdf of the aggregate claim amount random variable S. Currently, five methods
are supported.

1. Recursive calculation using the algorithm of Panjer (1981). This requires the severity
distribution to be discrete arithmetic on 0, 1, 2, . . . , m for some monetary unit and the
frequency distribution to be a member of either the (a, b, 0) or (a, b, 1) family of distri-
butions (Klugman et al. 2004). (These families contain the Poisson, binomial, negative
binomial and logarithmic distributions and their extensions with an arbitrary mass at
x = 0.) The general recursive formula is:
Pmin(x,m)
(p1 − (a + b)p0 )fC (x) + y=1 (a + by/x)fC (y)fS (x − y)
fS (x) = ,
1 − afC (0)
with starting value fS (0) = PN (fC (0)), where PN (·) is the probability generating func-
tion of N . Probabilities are computed until their sum is arbitrarily close to 1.
The recursions are done in C to dramatically increase speed. One difficulty the program-
mer is facing is the unknown length of the output. This was solved using a common,
simple and fast technique: first allocate an arbitrary amount of memory and double this
amount each time the allocated space gets full.

2. Exact calculation by numerical convolutions using (17) and (18). This also requires
a discrete severity distribution. However, there is no restriction on the shape of the
frequency distribution. The package merely implements the sum (17), the convolu-
tions being computed with R’s function convolve, which in turn uses the Fast Fourier
Transform. This approach is practical for small problems only, even on today’s fast
computers.

3. Normal approximation of the cdf, that is

 
x − µS
FS (x) ≈ Φ , (23)
σS

where µS = E[S] and σS2 = VAR[S]. For most realistic models, this approximation is
rather crude in the tails of the distribution.

4. Normal Power II approximation of the cdf, that is

s !
3 9 6 x − µS
FS (x) ≈ Φ − + +1+ , (24)
γS γS2 γS σ S

3/2
where γS = E[(S −µS )3 ]/σS . The approximation is valid for x > µS only and performs
reasonably well when γS < 1. See Daykin, Pentikäinen, and Pesonen (1994) for details.

5. Simulation of a random sample from S and approximation of FS (x) by the empirical

cdf (8). The simulation itself is done with function simul (see Section 4). This func-
tion admits very general hierarchical models for both the frequency and the severity
components.
18 actuar: An R Package for Actuarial Science

Here also, adding other methods to aggregateDist is simple due to its modular conception.
The arguments of aggregateDist differ depending on the calculation method; see the help
page for details. One interesting argument to note is x.scale to specify the monetary unit of
the severity distribution. This way, one does not have to mentally do the conversion between
the support of 0, 1, 2, . . . assumed by the recursive and convolution methods and the true
support of S.
The function returns an object of class "aggregateDist" inheriting from the "function"
class. Thus, one can use the object as a function to compute the value of FS (x) in any x.
For illustration purposes, consider the following model: the distribution of S is a compound
Poisson with parameter λ = 10 and severity distribution Gamma(2, 1). To obtain an approx-
imation of the cdf of S we first discretize the gamma distribution on (0, 22) with the unbiased
method and a step of 0.5, and then use the recursive method in aggregateDist:

R> fx <- discretize(pgamma(x, 2, 1), from = 0, to = 22,

+ step = 0.5, method = "unbiased", lev = levgamma(x, 2, 1))
R> Fs <- aggregateDist("recursive", model.freq = "poisson",
+ model.sev = fx, lambda = 10, x.scale = 0.5)
R> summary(Fs)

Aggregate Claim Amount Empirical CDF:

Min. 1st Qu. Median Mean 3rd Qu. Max.
0.0 14.5 19.5 20.0 25.0 71.0

Hence, object Fs contains an empirical cdf with support

R> knots(Fs)

[1] 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
[13] 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5
[25] 12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5 16.0 16.5 17.0 17.5
[37] 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 23.5
[49] 24.0 24.5 25.0 25.5 26.0 26.5 27.0 27.5 28.0 28.5 29.0 29.5
[61] 30.0 30.5 31.0 31.5 32.0 32.5 33.0 33.5 34.0 34.5 35.0 35.5
[73] 36.0 36.5 37.0 37.5 38.0 38.5 39.0 39.5 40.0 40.5 41.0 41.5
[85] 42.0 42.5 43.0 43.5 44.0 44.5 45.0 45.5 46.0 46.5 47.0 47.5
[97] 48.0 48.5 49.0 49.5 50.0 50.5 51.0 51.5 52.0 52.5 53.0 53.5
[109] 54.0 54.5 55.0 55.5 56.0 56.5 57.0 57.5 58.0 58.5 59.0 59.5
[121] 60.0 60.5 61.0 61.5 62.0 62.5 63.0 63.5 64.0 64.5 65.0 65.5
[133] 66.0 66.5 67.0 67.5 68.0 68.5 69.0 69.5 70.0 70.5 71.0

A nice graph of this function is obtained with a method of plot (see Figure 5):

R> plot(Fs, do.points = FALSE, verticals = TRUE, xlim = c(0, 60))

The package defines a few summary methods to extract information from "aggregateDist"
objects. First, there are methods of mean and quantile to easily compute the mean and
obtain the quantiles of the approximate distribution:
 Journal of Statistical Software 19

Aggregate Claim Amount Distribution

Recursive method approximation

1.0
0.8
0.6
FS(x)

0.4
0.2
0.0

0 10 20 30 40 50 60

Figure 5: Graphic of the empirical cdf of S obtained with the recursive method

R> mean(Fs)

[1] 20

R> quantile(Fs)

25% 50% 75% 90% 95% 97.5% 99% 99.5%

14.5 19.5 25.0 30.5 34.0 37.0 41.0 43.5

R> quantile(Fs, 0.999)

99.9%
49.5

Second, the package introduces the generic functions VaR and CTE with methods for objects
of class "aggregateDist". The former computes the value-at-risk VaRα such that
P[S ≤ VaRα ] = α, (25)
where α is the confidence level. Thus, the value-at-risk is nothing else than a quantile. As for
the method of CTE, it computes the conditional tail expectation
CTEα = E[S|S > VaRα ]. (26)
Here are examples using object Fs obtained above:
20 actuar: An R Package for Actuarial Science

Aggregate Claim Amount Distribution

1.0
0.8
0.6
FS(x)

0.4

recursive + unbiased
recursive + upper
recursive + lower
0.2

simulation
normal approximation
0.0

0 10 20 30 40 50 60

Figure 6: Comparison between the empirical or approximate cdf of S obtained with five
different methods

R> VaR(Fs)

90% 95% 99%

30.5 34.0 41.0

R> CTE(Fs)

90% 95% 99%

35.42 38.55 45.01

To conclude on the subject, Figure 6 shows the cdf of S using five of the many combinations
of discretization and calculation method supported by actuar.

3.4. The continuous time ruin model

We now turn to the multi-period ruin problem. Let U (t) denote the surplus of an insurance
company at time t, c(t) denote premiums collected through time t, and S(t) denote aggregate
claims paid through time t. If u is the initial surplus at time t = 0, then a mathematically
convenient definition of U (t) is

U (t) = u + c(t) − S(t). (27)

 Journal of Statistical Software 21

As mentioned previously, technical ruin of the insurance company occurs when the surplus
becomes negative. Therefore, the definition of the infinite time probability of ruin is

ψ(u) = P[U (t) < 0 for some t ≥ 0]. (28)

We define some other quantities needed in the sequel. Let N (t) denote the number of claims
up to time t ≥ 0 and Cj denote the amount of claim j. Then the definition of S(t) is analogous
to (16):
S(t) = C1 + · · · + CN (t) , (29)
assuming N (0) = 0 and S(t) = 0 as long as N (t) = 0. Furthermore, let Tj denote the
time when claim j occurs, such that T1 < T2 < T3 < . . . Then the random variable of the
interarrival (or wait) time between claim j − 1 and claim j is defined as W1 = T1 and

Wj = Tj − Tj−1 , j ≥ 2. (30)

For the rest of this discussion, we make the following assumptions:

1. premiums are collected at a constant rate c, hence c(t) = ct;

2. the sequence {Tj }j≥1 forms an ordinary renewal process, with the consequence that
random variables W1 , W2 , . . . are independent and identically distributed;

3. claim amounts C1 , C2 , . . . are independent and identically distributed.

3.5. Adjustment coefficient

The quantity known as the adjustment coefficient ρ hardly has any physical interpretation,
but it is useful as an approximation to the probability of ruin since we have the inequality

ψ(u) ≤ e−ρu , u ≥ 0.

The adjustment coefficient is defined as the smallest strictly positive solution (if it exists) of
the Lundberg equation
h(t) = E[etC−tcW ] = 1, (31)
where the premium rate c satisfies the positive safety loading constraint E[C − cW ] < 0. If C
and W are independent, as in the most common models, then the equation can be rewritten
as
h(t) = MC (t)MW (−tc) = 1. (32)

Function adjCoef of actuar computes the adjustment coefficient ρ from the following argu-
ments: either the two moment generating functions MC (t) and MW (t) (thereby assuming
independence) or else function h(t); the premium rate c; the upper bound of the support of
MC (t) or any other upper bound for ρ.
For example, if W and C are independent, W ∼ Exponential(2), C ∼ Exponential(1) and
the premium rate is c = 2.4 (for a safety loading of 20% using the expected value premium
principle), then the adjustment coefficient is
22 actuar: An R Package for Actuarial Science

R> adjCoef(mgf.claim = mgfexp(x), mgf.wait = mgfexp(x, 2),

+ premium.rate = 2.4, upper = 1)

[1] 0.1667

The function also supports models with proportional or excess-of-loss reinsurance (Centeno
2002). Under the first type of treaty, an insurer pays a proportion α of every loss and the
rest is paid by the reinsurer. Then, for fixed α the adjustment coefficient is the solution of

h(t) = E[etαC−tc(α)W ] = 1. (33)

Under an excess-of-loss treaty, the primary insurer pays each claim up to a limit L. Again,
for fixed L, the adjustment coefficient is the solution of

h(t) = E[et min(C,L)−tc(L)W ] = 1. (34)

For models with reinsurance, adjCoef returns an object of class "adjCoef" inheriting from
the "function" class. One can then use the object to compute the adjustment coefficient
for any retention rate α or retention limit L. The package also defines a method of plot for
these objects.
For example, using the same assumptions as above with proportional reinsurance and a 30%
safety loading for the reinsurer, the adjustment coefficient as a function of α ∈ [0, 1] is (see
Figure 7 for the graph):

R> mgfx <- function(x, y) mgfexp(x * y)

R> p <- function(x) 2.6 * x - 0.2
R> rho <- adjCoef(mgfx, mgfexp(x, 2), premium = p, upper = 1,
+ reins = "prop", from = 0, to = 1)
R> rho(c(0.75, 0.8, 0.9, 1))

[1] 0.1905 0.1862 0.1765 0.1667

R> plot(rho)

3.6. Probability of ruin

In this subsection, we always assume that interarrival times and claim amounts are indepen-
dent.
The main difficulty with the calculation of the infinite time probability of ruin lies in the lack
of explicit formulas except for the most simple models. If interarrival times are Exponential(λ)
distributed (Poisson claim number process) and claim amounts are Exponential(β) distributed,
then
λ −(β−λ/c)u
ψ(u) = e . (35)
cβ
If the frequency assumption of this model is defensible, the severity assumption can hardly
be used beyond illustration purposes.
 Journal of Statistical Software 23

Adjustment Coefficient
Proportional reinsurance

0.20
0.15
0.10
R(x)

0.05
0.00

0.0 0.2 0.4 0.6 0.8 1.0

Figure 7: Adjustment coefficient as a function of the retention rate

Fortunately, phase-type distributions have come to the rescue since the early 1990s. Asmussen
and Rolski (1991) first show that in the classical Cramér–Lundberg model where interarrival
times are Exponential(λ) distributed, if claim amounts are Phase-type(π, T ) distributed, then
ψ(u) = 1 − F (u), where F is Phase-type(π + , Q) with
λ
π + = − πT −1
c (36)
Q = T + tπ + ,
with t = −T e as in Section 2.1.
In the more general Sparre Andersen model where interarrival times can have any Phase-
type(ν, S) distribution, Asmussen and Rolski (1991) also show that using the same claim
severity assumption as above, one still has ψ(u) = 1 − F (u) where F is Phase-type(π + , Q),
but with parameters
e0 (Q − T )
π+ = (37)
ce0 t
and Q solution of
Q = Ψ(Q)
(38)
= T − tπ (I n ⊗ ν)(Q ⊕ S)−1 (I n ⊗ s) .
 

In the above, s = −Se, I n is the n × n identity matrix, ⊗ denotes the usual Kronecker
product between two matrices and ⊕ is the Kronecker sum defined as
Am×m ⊕ B n×n = A ⊗ I n + B ⊗ I m . (39)
24 actuar: An R Package for Actuarial Science

Function ruin of actuar returns a function object of class "ruin" to compute the probability
of ruin for any initial surplus u. In all cases except the exponential/exponential model where
(35) is used, the output object calls function pphtype to compute the ruin probabilities.
Some thought went into the interface of ruin. Obviously, all models can be specified using
phase-type distributions, but the authors wanted users to have easy access to the most com-
mon models involving exponential and Erlang distributions. Hence, one first states the claim
amount and interarrival times models with any combination of "exponential", "Erlang" and
"phase-type". Then, one passes the parameters of each model using lists with components
named after the corresponding parameters of dexp, dgamma and dphtype. If a component
"weights" is found in a list, the model is a mixture of exponential or Erlang (mixtures of
phase-type are not supported). Every component of the parameter lists is recycled as needed.
The following examples should make the matter clearer. (All examples use c = 1, the default
value in ruin.) First, for the exponential/exponential model, one has

R> psi <- ruin(claims = "e", par.claims = list(rate = 5),

+ wait = "e", par.wait = list(rate = 3))
R> psi

function (u, survival = FALSE, lower.tail = !survival)

{
res <- 0.6 * exp(-(2) * u)
if (lower.tail)
res
else 0.5 - res + 0.5
}
<environment: 0x19af2150>
attr(,"class")
[1] "ruin" "function"

R> psi(0:10)

[1] 6.000e-01 8.120e-02 1.099e-02 1.487e-03 2.013e-04 2.724e-05

[7] 3.687e-06 4.989e-07 6.752e-08 9.138e-09 1.237e-09

Second, for a model with mixture of two exponentials claim amounts and exponential inter-
arrival times, the simplest call to ruin is

R> ruin(claims = "e", par.claims = list(rate = c(3, 7),

+ weights = 0.5), wait = "e", par.wait = list(rate = 3))

function (u, survival = FALSE, lower.tail = !survival)

pphtype(u, c(0.5, 0.214285714285714), c(-1.5, 3.5, 0.642857142857143,
-5.5), lower.tail = !lower.tail)
<environment: 0x1dfba1c>
attr(,"class")
[1] "ruin" "function"
 Journal of Statistical Software 25

Finally, one will obtain a function to compute ruin probabilities in a model with phase-type
claim amounts and mixture of exponentials interarrival times with

R> prob <- c(0.5614, 0.4386)

R> rates <- matrix(c(-8.64, 0.101, 1.997, -1.095), 2, 2)
R> ruin(claims = "p", par.claims = list(prob = prob, rates = rates),
+ wait = "e", par.wait = list(rate = c(5, 1), weights = c(0.4, 0.6)))

function (u, survival = FALSE, lower.tail = !survival)

pphtype(u, c(0.146595513877824, 0.761505562273639), c(-7.66616600130962,
0.246715940794557, 7.05568145018378, -0.338063471100003),
lower.tail = !lower.tail)
<environment: 0x19ba1208>
attr(,"class")
[1] "ruin" "function"

To ease plotting of the probability of ruin function, the package provides a method of plot
for objects returned by ruin that is a simple wrapper for curve (see Figure 8):

R> psi <- ruin(claims = "p", par.claims = list(prob = prob, rates = rates),
+ wait = "e", par.wait = list(rate = c(5, 1), weights = c(0.4, 0.6)))
R> plot(psi, from = 0, to = 50)

4. Simulation of compound hierarchical models

Function simul simulates portfolios of data following compound models of the form (16)
where both the frequency and the severity components can have a hierarchical structure.
The main characteristic of hierarchical models is to have the probability law at some level
in the classification structure be conditional on the outcome in previous levels. For example,
consider the following compound hierarchical model:

Sijt = Cijt1 + · · · + CijtNijt , (40)

for i = 1, . . . , I, j = 1, . . . , Ji , t = 1, . . . , nij and with

Nijt |Λij , Φi ∼ Poisson(wijt Λij ) Cijtu |Θij , Ψi ∼ Lognormal(Θij , 1)

Λij |Φi ∼ Gamma(Φi , 1) Θij |Ψi ∼ N (Ψi , 1) (41)
Φi ∼ Exponential(2) Ψi ∼ N (2, 0.1).

The random variables Φi , Λij , Ψi and Θij are generally seen as risk parameters in the actuarial
literature. The wijt s are known weights, or volumes. Using as convention to number the data
level 0, the above is a two-level hierarchical model.
Goulet and Pouliot (2008) describe in detail the model specification method used in simul.
For the sake of completeness, we briefly outline this method here.
A hierarchical model is completely specified by the number of nodes at each level (I, J1 , . . . , JI
and n11 , . . . , nIJ , above) and by the probability laws at each level. The number of nodes is
26 actuar: An R Package for Actuarial Science

Probability of Ruin

0.8
0.6
ψ(u)

0.4
0.2
0.0

0 10 20 30 40 50

Figure 8: Graphic of the probability of ruin as a function of the initial surplus u

passed to simul by means of a named list where each element is a vector of the number of
nodes at a given level. Vectors are recycled when the number of nodes is the same throughout
a level. Probability models are expressed in a semi-symbolic fashion using an object of mode
"expression". Each element of the object must be named — with names matching those
of the number of nodes list — and should be a complete call to an existing random number
generation function, with the number of variates omitted. Hierarchical models are achieved
by replacing one or more parameters of a distribution at a given level by any combination of
the names of the levels above. If no mixing is to take place at a level, the model for this level
can be NULL.
Function simul also supports usage of weights in models. These usually modify the frequency
parameters to take into account the“size”of an entity. Weights are used in simulation wherever
the name weights appears in a model.
Hence, function simul has four main arguments: 1) nodes for the number of nodes list;
2) model.freq for the frequency model; 3) model.sev for the severity model; 4) weights for
the vector of weights in lexicographic order, that is all weights of entity 1, then all weights of
entity 2, and so on.
For example, assuming that I = 2, J1 = 4, J2 = 3, n11 = · · · = n14 = 4 and n21 = n22 = n23 =
5 in model (41) above, and that weights are simply simulated from a uniform distribution on
(0.5, 2.5), then simulation of a data set with simul is achieved with:

R> nodes <- list(cohort = 2, entity = c(4, 3), year = c(4,

+ 4, 4, 4, 5, 5, 5))
 Journal of Statistical Software 27

R> mf <- expression(cohort = rexp(2), entity = rgamma(cohort,

+ 1), year = rpois(weights * entity))
R> ms <- expression(cohort = rnorm(2, sqrt(0.1)), entity = rnorm(cohort,
+ 1), year = rlnorm(entity, 1))
R> wijt <- runif(31, 0.5, 2.5)
R> pf <- simul(nodes = nodes, model.freq = mf, model.sev = ms,
+ weights = wijt)

The function returns the variates in a two-dimension list of class "portfolio" containing
all the individual claim amounts for each entity. Such an object can be seen as a three-
dimensional array with a third dimension of potentially varying length. The function also
returns a matrix of integers giving the classification indexes of each entity in the portfolio
(subscripts i and j in the notation above). Displaying the complete content of the object
returned by simul can be impractical. For this reason, the print method for this class only
prints the simulation model and the number of claims in each node:

R> pf

Portfolio of claim amounts

Frequency model
cohort ~ rexp(2)
entity ~ rgamma(cohort, 1)
year ~ rpois(weights * entity)
Severity model
cohort ~ rnorm(2, sqrt(0.1))
entity ~ rnorm(cohort, 1)
year ~ rlnorm(entity, 1)

Number of claims per node:

cohort entity year.1 year.2 year.3 year.4 year.5

[1,] 1 1 2 2 1 0 NA
[2,] 1 2 0 0 0 0 NA
[3,] 1 3 0 3 0 2 NA
[4,] 1 4 0 1 1 1 NA
[5,] 2 1 0 1 1 1 2
[6,] 2 2 0 0 0 0 0
[7,] 2 3 3 4 2 2 0

The package defines methods for four generic functions to easily access key quantities of the
simulated portfolio.

1. By default, the method of aggregate returns the values of aggregate claim amounts
Sijt in a regular matrix (subscripts i and j in the rows, subscript t in the columns). The
method has a by argument to get statistics for other groupings and a FUN argument to
get statistics other than the sum:
28 actuar: An R Package for Actuarial Science

R> aggregate(pf)

cohort entity year.1 year.2 year.3 year.4 year.5

[1,] 1 1 31.37 7.521 11.383 0.000 NA
[2,] 1 2 0.00 0.000 0.000 0.000 NA
[3,] 1 3 0.00 72.706 0.000 23.981 NA
[4,] 1 4 0.00 98.130 50.622 55.705 NA
[5,] 2 1 0.00 11.793 2.253 2.397 10.48
[6,] 2 2 0.00 0.000 0.000 0.000 0.00
[7,] 2 3 44.81 88.737 57.593 14.589 0.00

R> aggregate(pf, by = c("cohort", "year"), FUN = mean)

cohort year.1 year.2 year.3 year.4 year.5

[1,] 1 15.69 29.73 31.00 26.562 NA
[2,] 2 14.94 20.11 19.95 5.662 5.238

2. The method of frequency returns the number of claims Nijt . It is a wrapper for
aggregate with the default sum statistic replaced by length. Hence, arguments by and
FUN remain available:

R> frequency(pf)

cohort entity year.1 year.2 year.3 year.4 year.5

[1,] 1 1 2 2 1 0 NA
[2,] 1 2 0 0 0 0 NA
[3,] 1 3 0 3 0 2 NA
[4,] 1 4 0 1 1 1 NA
[5,] 2 1 0 1 1 1 2
[6,] 2 2 0 0 0 0 0
[7,] 2 3 3 4 2 2 0

R> frequency(pf, by = "cohort")

cohort freq
[1,] 1 17
[2,] 2 16

3. The method of severity (a generic function introduced by the package) returns the
individual claim amounts Cijtu in a matrix similar to those above, but with a number
of columns equal to the maximum number of observations per entity,
nij
X
max Nijt .
i,j
t=1

Thus, the original period of observation (subscript t) and the identifier of the severity
within the period (subscript u) are lost and each variate now constitutes a “period” of
observation. For this reason, the method provides an argument splitcol in case one
would like to extract separately the individual claim amounts of one or more periods:
 Journal of Statistical Software 29

R> severity(pf)

$main
cohort entity claim.1 claim.2 claim.3 claim.4 claim.5 claim.6
[1,] 1 1 7.974 23.401 3.153 4.368 11.383 NA
[2,] 1 2 NA NA NA NA NA NA
[3,] 1 3 3.817 41.979 26.910 4.903 19.078 NA
[4,] 1 4 98.130 50.622 55.705 NA NA NA
[5,] 2 1 11.793 2.253 2.397 9.472 1.004 NA
[6,] 2 2 NA NA NA NA NA NA
[7,] 2 3 14.322 11.522 18.966 33.108 15.532 14.99
claim.7 claim.8 claim.9 claim.10 claim.11
[1,] NA NA NA NA NA
[2,] NA NA NA NA NA
[3,] NA NA NA NA NA
[4,] NA NA NA NA NA
[5,] NA NA NA NA NA
[6,] NA NA NA NA NA
[7,] 25.11 40.15 17.44 4.426 10.16

$split
NULL

R> severity(pf, splitcol = 1)

$main
cohort entity claim.1 claim.2 claim.3 claim.4 claim.5 claim.6
[1,] 1 1 3.153 4.368 11.383 NA NA NA
[2,] 1 2 NA NA NA NA NA NA
[3,] 1 3 3.817 41.979 26.910 4.903 19.078 NA
[4,] 1 4 98.130 50.622 55.705 NA NA NA
[5,] 2 1 11.793 2.253 2.397 9.472 1.004 NA
[6,] 2 2 NA NA NA NA NA NA
[7,] 2 3 33.108 15.532 14.990 25.107 40.150 17.44
claim.7 claim.8
[1,] NA NA
[2,] NA NA
[3,] NA NA
[4,] NA NA
[5,] NA NA
[6,] NA NA
[7,] 4.426 10.16

$split
cohort entity claim.1 claim.2 claim.3
[1,] 1 1 7.974 23.40 NA
[2,] 1 2 NA NA NA
30 actuar: An R Package for Actuarial Science

[3,] 1 3 NA NA NA
[4,] 1 4 NA NA NA
[5,] 2 1 NA NA NA
[6,] 2 2 NA NA NA
[7,] 2 3 14.322 11.52 18.97

4. The method of weights extracts the weights matrix from a simulated data set:

R> weights(pf)

cohort entity year.1 year.2 year.3 year.4 year.5

[1,] 1 1 0.8361 2.115 1.2699 1.1555 NA
[2,] 1 2 1.7042 1.709 0.7493 1.0892 NA
[3,] 1 3 1.6552 1.762 1.5240 1.5100 NA
[4,] 1 4 1.5681 1.614 2.2358 2.1594 NA
[5,] 2 1 0.7229 1.907 2.2950 1.0595 0.9564
[6,] 2 2 0.5307 0.758 0.6868 0.9738 2.0823
[7,] 2 3 1.6995 2.320 1.6208 2.0114 1.2583

In addition, all methods have a classification and a prefix argument. When the first is
FALSE, the classification index columns are omitted from the result. The second argument
overrides the default column name prefix; see the simul.summaries help page for details.
Function simul was used to simulate the data in Forgues, Goulet, and Lu (2006).

5. Credibility theory
Credibility models are actuarial tools to distribute premiums fairly among a heterogeneous
group of policyholders (henceforth called entities). More generally, they can be seen as pre-
diction methods applicable in any setting where repeated measures are made for subjects with
different risk levels.
The credibility theory facilities of actuar consist of the matrix hachemeister containing the
famous data set of Hachemeister (1975) and the function cm to fit credibility models.

5.1. Hachemeister data set

The data set of Hachemeister (1975) consists of private passenger bodily injury insurance
average claim amounts, and the corresponding number of claims, for five U.S. states over
12 quarters between July 1970 and June 1973. The data set is included in the package in
the form of a matrix with 5 rows and 25 columns. The first column contains a state index
(state), columns 2–13 contain the claim averages (ratio.1, . . . , ratio.12) and columns
14–25 contain the claim numbers (weight.1, . . . , weight.12).

5.2. Hierarchical credibility model

The linear model fitting function of R is named lm. Since credibility models are very close
in many respects to linear models, and since the credibility model fitting function of actuar
borrows much of its interface from lm, we named the credibility function cm.
 Journal of Statistical Software 31

Function cm acts as a unified interface for all credibility models supported by the package. Cur-
rently, these are the unidimensional models of Bühlmann (1969) and Bühlmann and Straub
(1970), the hierarchical model of Jewell (1975) (of which the first two are special cases) and
the regression model of Hachemeister (1975). The modular design of cm makes it easy to add
new models if desired.
This subsection concentrates on usage of cm for hierarchical models.
There are some variations in the formulas of the hierarchical model in the literature. We
compute the credibility premiums as given in Bühlmann and Jewell (1987) or Bühlmann
and Gisler (2005). We support three types of estimators of the between variance structure
parameters: the unbiased estimators of Bühlmann and Gisler (2005) (the default), the slightly
different version of Ohlsson (2005) and the iterative pseudo-estimators as found in Goovaerts
and Hoogstad (1987) or Goulet (1998). See Belhadj, Goulet, and Ouellet (2008) for further
discussion on this topic.
The credibility modeling function assumes that data is available in the format most practical
applications would use, namely a rectangular array (matrix or data frame) with entity obser-
vations in the rows and with one or more classification index columns (numeric or character).
One will recognize the output format of simul and its summary methods.
Then, function cm works much the same as lm. It takes in argument: a formula of the
form ~ terms describing the hierarchical interactions in a data set; the data set containing
the variables referenced in the formula; the names of the columns where the ratios and the
weights are to be found in the data set. The latter should contain at least two nodes in
each level and more than one period of experience for at least one entity. Missing values are
represented by NAs. There can be entities with no experience (complete lines of NAs).
In order to give an easily reproducible example, we group states 1 and 3 of the Hachemeister
data set into one cohort and states 2, 4 and 5 into another. This shows that data does not
have to be sorted by level. The fitted model using the iterative estimators is:
R> X <- cbind(cohort = c(1, 2, 1, 2, 2), hachemeister)
R> fit <- cm(~cohort + cohort:state, data = X, ratios = ratio.1:ratio.12,
+ weights = weight.1:weight.12, method = "iterative")
R> fit

Call:
cm(formula = ~cohort + cohort:state, data = X, ratios = ratio.1:ratio.12,
weights = weight.1:weight.12, method = "iterative")

Structure Parameters Estimators

Collective premium: 1746

Between cohort variance: 88981

Within cohort/Between state variance: 10952
Within state variance: 139120026

The function returns a fitted model object of class "cm" containing the estimators of the
structure parameters. To compute the credibility premiums, one calls a method of predict
for this class:
32 actuar: An R Package for Actuarial Science

R> predict(fit)

$cohort
[1] 1949 1543

$state
[1] 2048 1524 1875 1497 1585

One can also obtain a nicely formatted view of the most important results with a call to
summary:

R> summary(fit)

Call:
cm(formula = ~cohort + cohort:state, data = X, ratios = ratio.1:ratio.12,
weights = weight.1:weight.12, method = "iterative")

Structure Parameters Estimators

Collective premium: 1746

Between cohort variance: 88981

Within cohort/Between state variance: 10952
Within state variance: 139120026

Detailed premiums

Level: cohort
cohort Indiv. mean Weight Cred. factor Cred. premium
1 1967 1.407 0.9196 1949
2 1528 1.596 0.9284 1543

Level: state
cohort state Indiv. mean Weight Cred. factor Cred. premium
1 1 2061 100155 0.8874 2048
2 2 1511 19895 0.6103 1524
1 3 1806 13735 0.5195 1875
2 4 1353 4152 0.2463 1497
2 5 1600 36110 0.7398 1585

The methods of predict and summary can both report for a subset of the levels by means of
an argument levels. For example:

R> summary(fit, levels = "cohort")

Call:
cm(formula = ~cohort + cohort:state, data = X, ratios = ratio.1:ratio.12,
 Journal of Statistical Software 33

weights = weight.1:weight.12, method = "iterative")

Structure Parameters Estimators

Collective premium: 1746

Between cohort variance: 88981

Within cohort variance: 10952

Detailed premiums

Level: cohort
cohort Indiv. mean Weight Cred. factor Cred. premium
1 1967 1.407 0.9196 1949
2 1528 1.596 0.9284 1543

R> predict(fit, levels = "cohort")

$cohort
[1] 1949 1543

The results above differ from those of Goovaerts and Hoogstad (1987) for the same example
because the formulas for the credibility premiums are different.

5.3. Bühlmann and Bühlmann–Straub models

As mentioned above, the Bühlmann and Bühlmann–Straub models are simply one-level hier-
archical models. In this case, the Bühlmann–Gisler and Ohlsson estimators of the between
variance parameters are both identical to the usual Bühlmann and Straub (1970) estimator,
and the iterative estimator is better known as the Bichsel–Straub estimator.

5.4. Regression model of Hachemeister

The regression model of Hachemeister (1975) is a generalization of the Bühlmann–Straub
model. If data shows a systematic trend, the latter model will typically under- or over-
estimate the true premium of an entity. The idea of Hachemeister was to fit to the data
a regression model where the parameters are a credibility weighted average of an entity’s
regression parameters and the group’s parameters.
In order to use cm to fit a credibility regression model to a data set, one has to specify a vector
or matrix of regressors by means of argument xreg. For example, fitting the model

Xit = β0 + β1 (12 − t) + εt , t = 1, . . . , 12

to the original data set of Hachemeister is done with

R> fit <- cm(~state, hachemeister, xreg = 12:1, ratios = ratio.1:ratio.12,

+ weights = weight.1:weight.12)
R> fit
34 actuar: An R Package for Actuarial Science

Call:
cm(formula = ~state, data = hachemeister, ratios = ratio.1:ratio.12,
weights = weight.1:weight.12, xreg = 12:1)

Structure Parameters Estimators

Collective premium: 1885 -32.05

Between state variance: 145359 -6623.4

-6623 301.8
Within state variance: 49870187

Computing the credibility premiums requires to give the “future” values of the regressors as
in predict.lm, although with a simplified syntax for the one regressor case:

R> predict(fit, newdata = 0)

[1] 2437 1651 2073 1507 1759

6. Documentation
In addition to the help pages required by the R packaging system, the package includes
vignettes and demonstration scripts; running

R> vignette(package = "actuar")

and

R> demo(package = "actuar")

at the R prompt will give the list of each.

7. Conclusion
The paper presented the facilities of the R package actuar in the fields of loss distribution
modeling, risk theory, simulation of compound hierarchical models and credibility theory. We
feel this version of the package covers most of the basic needs in these areas. In the future we
plan to improve the functions currently available and to start adding more advanced features.
For example, future versions of the package should include support for dependence models in
risk theory and better handling of regression credibility models.
Obviously, the package left many other fields of Actuarial Science untouched. For this situa-
tion to change, we hope that experts in their field will join their efforts to ours and contribute
code to the actuar project. The project will continue to grow and to improve by and for the
community of developers and users.
Finally, if you use R or actuar for actuarial analysis, please cite the software in publications.
Use
 Journal of Statistical Software 35

R> citation()

and

R> citation("actuar")

for information on how to cite the software.

Acknowledgments
The package would not be at this stage of development without the stimulating contribution
of Sébastien Auclair, Louis-Philippe Pouliot and Tommy Ouellet.
This research benefited from financial support from the Natural Sciences and Engineering
Research Council of Canada and from the Chaire d’actuariat (Actuarial Science Chair) of
Université Laval.
Finally, the authors thank two anonymous referees for many improvements to the paper.

References

Asmussen S, Rolski T (1991). “Computational Methods in Risk Theory: A Matrix-Algorithmic

Approach.” Insurance: Mathematics and Economics, 10, 259–274.

Bates D, Maechler M (2008). Matrix: A Matrix package for R. R package version 0.999375-7,
URL http://CRAN.R-project.org/package=Matrix.

Belhadj H, Goulet V, Ouellet T (2008). “On Parameter Estimation in Hierarchical Credibility.”

In preparation.

Bühlmann H (1969). “Experience Rating and Credibility.” ASTIN Bulletin, 5, 157–165.

Bühlmann H, Gisler A (2005). A Course in Credibility Theory and its Applications. Springer-
Verlag. ISBN 3-5402575-3-5.

Bühlmann H, Jewell WS (1987). “Hierarchical Credibility Revisited.” Bulletin of the Swiss

Association of Actuaries, 87, 35–54.

Bühlmann H, Straub E (1970). “Glaubgwürdigkeit für Schadensätze.” Bulletin of the Swiss

Association of Actuaries, 70, 111–133.

Centeno MdL (2002). “Measuring the Effects of Reinsurance by the Adjustment Coefficient
in the Sparre-Anderson Model.” Insurance: Mathematics and Economics, 30, 37–49.

Daykin C, Pentikäinen T, Pesonen M (1994). Practical Risk Theory for Actuaries. Chapman
& Hall, London. ISBN 0-4124285-0-4.

Denuit M, Charpentier A (2004). Mathématiques de l’assurance non-vie, volume 1, Principes

fondamentaux de théorie du risque. Economica, Paris. ISBN 2-7178485-4-1.
36 actuar: An R Package for Actuarial Science

Forgues A, Goulet V, Lu J (2006). “Credibility for Severity Revisited.” North American

Actuarial Journal, 10(1), 49–62.

Gerber HU (1979). An Introduction to Mathematical Risk Theory. Huebner Foundation,

Philadelphia.

Goovaerts MJ, Hoogstad WJ (1987). Credibility Theory. Number 4 in Surveys of Actuarial

Studies. Nationale-Nederlanden N.V., Netherlands.

Goulet V (1998). “Principles and Application of Credibility Theory.” Journal of Actuarial

Practice, 6, 5–62.

Goulet V, Pouliot LP (2008). “Simulation of Compound Hierarchical Models in R.” North

American Actuarial Journal. Submitted for publication.

Hachemeister CA (1975). “Credibility for Regression Models with Application to Trend.”

In “Credibility, Theory and Applications,” Proceedings of the Berkeley Actuarial Research
Conference on Credibility, pp. 129–163. Academic Press, New York.

Hogg RV, Klugman SA (1984). Loss Distributions. John Wiley & Sons, New York. ISBN
0-4718792-9-0.

Jewell WS (1975). “The Use of Collateral Data in Credibility Theory: A Hierarchical Model.”
Giornale dell’Istituto Italiano degli Attuari, 38, 1–16.

Kaas R, Goovaerts M, Dhaene J, Denuit M (2001). Modern Actuarial Risk Theory. Kluwer
Academic Publishers, Dordrecht. ISBN 0-7923763-6-6.

Klugman SA, Panjer HH, Willmot G (1998). Loss Models: From Data to Decisions. John
Wiley & Sons, New York. ISBN 0-4712388-4-8.

Klugman SA, Panjer HH, Willmot G (2004). Loss Models: From Data to Decisions. John
Wiley & Sons, New York, second edition. ISBN 0-4712157-7-5.

Lumley T (2008). survival: Survival Analysis, Including Penalised Likelihood. R package

version 2.34, URL http://CRAN.R-project.org/package=survival.

Neuts MF (1981). Matrix-Geometric Solutions in Stochastic Models: An Algorithmic Ap-

proach. Dover Publications. ISBN 978-0-4866834-2-3.

Ohlsson E (2005). “Simplified Estimation of Structure Parameters in Hierarchical Credibility.”

Presented at the Zurich ASTIN Colloquium, URL http://www.actuaries.org/ASTIN/
Colloquia/Zurich/Ohlsson.pdf.

Panjer HH (1981). “Recursive Evaluation of a Family of Compound Distributions.” ASTIN

Bulletin, 12, 22–26.

Pinheiro J, Bates D, DebRoy S, Sarkar D, the R Development Core Team (2007). nlme:
Linear and Nonlinear Mixed Effects Models. R package version 3.1-86, URL http://CRAN.
R-project.org/package=nlme.
 Journal of Statistical Software 37

R Development Core Team (2008). R: A Language and Environment for Statistical Computing.
R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.
org/.

Venables WN, Ripley BD (2002). Modern Applied Statistics with S. Springer-Verlag, New
York, 4 edition. ISBN 0-3879545-7-0.

Wheeler B (2008). SuppDists: Supplementary Distributions. R package version 1.1-2, URL

http://CRAN.R-project.org/package=SuppDists.

Wuertz D (2007). Rmetrics: Financial Engineering and Computational Finance. R package

version 260.72, URL http://CRAN.R-project.org/package=Rmetrics.

Yan J (2007). “Enjoy the Joy of Copulas: With a Package copula.” Journal of Statistical
Software, 21(4), 1–21. URL http://www.jstatsoft.org/v21/i04/.

Affiliation:
Vincent Goulet
École d’actuariat
Pavillon Alexandre-Vachon
1045, avenue de la Médecine
Université Laval
Québec (QC) G1V 0A6, Canada
E-mail: [email protected]
URL: http://vgoulet.act.ulaval.ca/

Journal of Statistical Software http://www.jstatsoft.org/

published by the American Statistical Association http://www.amstat.org/
Volume 25, Issue 7 Submitted: 2007-11-16
March 2008 Accepted: 2008-02-14