Institut de Recerca en Economia Aplicada Regional i Pública Document de Treball 2018/10 1/29 pág.

Research Institute of Applied Economics Working Paper 2018/10 1/29 pág.

“Exposure to risk increases the excess of zero accident claims

frequency in automobile insurance”

Guillen M, Nielsen JP, Ayuso M & Pérez-Marín A

4

WEBSITE: www.ub.edu/irea/ • CONTACT: irea@ub.edu

The Research Institute of Applied Economics (IREA) in Barcelona was founded in 2005, as a research
institute in applied economics. Three consolidated research groups make up the institute: AQR, RISK
and GiM, and a large number of members are involved in the Institute. IREA focuses on four priority
lines of investigation: (i) the quantitative study of regional and urban economic activity and analysis of
regional and local economic policies, (ii) study of public economic activity in markets, particularly in the
fields of empirical evaluation of privatization, the regulation and competition in the markets of public
services using state of industrial economy, (iii) risk analysis in finance and insurance, and (iv) the
development of micro and macro econometrics applied for the analysis of economic activity, particularly
for quantitative evaluation of public policies.

IREA Working Papers often represent preliminary work and are circulated to encourage discussion.
Citation of such a paper should account for its provisional character. For that reason, IREA Working
Papers may not be reproduced or distributed without the written consent of the author. A revised version
may be available directly from the author.

Any opinions expressed here are those of the author(s) and not those of IREA. Research published in
this series may include views on policy, but the institute itself takes no institutional policy positions.
 Abstract

Most automobile insurance databases contain a large number of policy

holders with zero claims. This high frequency of zeros may reflect the fact
that some insureds make little use of their vehicle, or that they do not wish to
make a claim for small accidents in order to avoid an increase in their
premium, but it might also be because of good driving. We analyse
information on exposure to risk and driving habits using telematics data from
a Pay-as-you-Drive sample of insureds. We include distance travelled per
year as part of an offset in a zero- inflated Poisson model to predict the
excess of zeros. We show the existence of a learning effect for large values
of distance travelled, so that longer driving should result in higher premium,
but there should be a discount for drivers that accumulate longer distances
over time due to the increased proportion of zero claims. We confirm that
speed limit violations and driving in urban areas increase the expected
number of accident claims. We discuss how telematics information can be
used to design better insurance and to improve traffic safety.

JEL classification: C35, C55, G22.

Keywords: Telematics, pay-as-you-drive, mileage.

Montserrat Guillen: Department of Econometrics, Riskcenter-IREA, University of Barcelona,

Avinguda Diagonal 690, 08034 Barcelona, Spain. E-mail: mguillen@ub.edu

Ana M. Pérez-Marín: Department of Econometrics, Riskcenter-IREA, University of Barcelona,

Avinguda Diagonal 690, 08034 Barcelona, Spain.

Mercedes Ayuso: Department of Econometrics, Riskcenter-IREA, University of Barcelona,

Avinguda Diagonal 690, 08034 Barcelona, Spain.

Jens Perch Nielsen: Cass Business School, City, University of London, 106 Bunhill Row,
London EC1Y 8TZ, United Kingdom.

Acknowledgements

The study was supported by ICREA Academia, the Spanish Ministry of Economy and Competitiveness, and the ERDF
under grants ECO2015-66314-R and ECO2016-76203-C2-2-P.
1. INTRODUCTION AND MOTIVATION

According to the World Health Organization (2017), road traffic injuries are responsible

for more than 1.2 million deaths every year. Indeed, they are the leading cause of

mortality among those aged between 15 and 29, at a cost to governments of

approximately 3% of GDP. This situation is exacerbated if we contemplate that from

the beginning of 2013 until the end of 2015, there was a 16% increase in the number of

vehicles on the world’s roads.(1)

Automobile insurance is compulsory in almost all countries and, recently, many

insurance companies have begun to collect telematics data about drivers’ exposure to

traffic (i.e. distance driven and vehicle location) and their driving behaviour (excess

speed and aggressiveness). This information can improve the insurance ratemaking

process(2-10) and also allows conclusions to be drawn about how to make driving safer.

New automobile insurance products (known by the acronyms PAYD, pay-as-you-drive,

or PHYD, pay-how-you-drive) necessitate the introduction of a GPS device in the

insured vehicle to record and store relevant information about variables that change over

time, including, for example, the number of kilometres driven per day by the insured,

the percentage of kilometres driven above the speed limit, and the percentage of

kilometres driven at night, among others. This development represents a remarkable

advance, given that, previously, automobile insurance companies could only use

variables related to certain fixed characteristics of the insured (for example age, gender,

or number of years since the driver’s license was issued) and the vehicle (age of the

automobile, engine power, etc.).

3
Most automobile insurance databases contain many policy holders with zero claims.

This high frequency of ‘zeros’ may be due to the presence of insureds that have no wish

to claim for small accidents in order to avoid a premium increase or, alternatively, it

might be due to the relative lack of use they make of their vehicles. If the vehicle is

parked in a garage, it is not exposed to the risk of accident. Here, we analyse distance

driven as a measure of exposure to risk and examine its role in the probability of an

insured having zero claims. We show how to differentiate those drivers that almost

never use their vehicles (and so have little exposure to the risk of an accident) from

those that are good drivers, i.e. those who, despite recording high mileages, are not

involved in any accidents. In what follows we refer to accidents as opposed to claims,

even though we are aware that some accidents are not reported to the insurance

company. Indeed, a detailed discussion of the difference between the number of

accidents and the number of claims has previously been reported.(11)

We discover a positive relationship between the distance driven and the number of

excess zeros observed in the number of claims. We argue that this is due to a learning

effect, where good drivers are more frequent than expected among those that drive long

distances. The overall effect of the driving distance variable is positive, however, even

if it is true that longer driving should obviously result in higher premium, there is a

discount due to the increased proportion of zeros in the frequencies, due to a learning

effect. The overall effect is still an increase in the premium, however not as much as we

would expect without the learning effect.

Our research is innovative because (1) we introduce telematics covariates while dealing

with the excess of zeros and (2) we discuss the implications for new insurance products

and traffic safety that are obtained on the basis of distance driven.

4
Various studies have explored the potential of telematics when applied to risks of road

accidents, beginning in 1968 with a preliminary analysis by Vickrey.(12) More recently,

several papers have examined the impact of new technologies on road safety and how

driving habits can be measured,(13-20) while others have focused specifically on mileage

and new risk factors that might be included in the ratemaking process (see (2) for an

extended review). Recently, it has been proven that including standard telematics

variables significantly improves risk assessment of insureds, therefore insurers should

be able to tailor their products to the customers’ risk profile. (21) The objective for the

insurance industry is to penalize high risk drivers with higher premiums by taking into

consideration factors related to dangerous driving, including, for example, exceeding

the speed limits or not respecting safety distances. We show that having information

about the annual distance driven by the insured improves the ratemaking process

considerably not only because it is a measure of exposure to risk, but because of the

crucial role it plays in the analysis of the absence of claims, i.e. the probability of not

claiming or, in other words, the probability of zero claims. See the following papers on

the relevance of including distance driven as a traffic risk factor.(22,23)

In terms of methodology, Poisson regression models have traditionally been used to

predict the number of automobile claims in insurance. The Poisson regression model is

a special case of the generalized linear model class and serve as a benchmark

model.(24,25) However, various corrections have to be made when assuming that the

probability of zero is larger than the probability under the Poisson assumption – a so-

called excess of zeros. Various papers suggest that this excess is caused by

asymmetrical information with an insured preferring not to declare a claim so as to

avoid certain deductibles or the application of a bonus-malus system.(26,27) In this paper,

we wish to differentiate those drivers that have no claims because they rarely use their

5
vehicles during the year (in the extreme case, making no use of the vehicle at all) from

those that have no claims despite being frequent drivers. To do this, we propose using a

zero-inflated Poisson (ZIP) model corrected by distance (kilometres driven per year by

the driver). While various studies have used ZIP models(28-30) and applied them to the

context of automobile insurance(31-33), none of these contributions has analysed the role

of exposure to risk in terms of distance driven.

From the empirical point of view, we draw on a real automobile claims database for a

sample of insureds. This includes individual details about annual mileage travelled and

other aspects of driving behaviour, which enable us to study the effects of various

indicators on the probability of making a claim. We highlight the implications of this for

the design of new insurance ratemaking processes.

The rest of the paper is structured as follows. In section 2, we present the methodology

used when including distance as an offset variable in the ZIP model. The database and

some descriptive results are presented in section 3 and our main results obtained with

the models specified are analysed in section 4. Finally, a discussion and the main

conclusions drawn from this research are presented in section 5.

2. METHODOLOGY

A Poisson regression with an offset variable is the logical way to include an exposure to

risk variable in our model. Here, therefore, we opt to use a Poisson model with offset

and, by way of alternative, a two-step procedure aimed at introducing telematics data,

which serves as a correction to the classical model.

6
Zero-inflated Poisson (ZIP) regression is a model for count data with an excess of zeros.

It assumes that with probability p the only possible observation is 0, and with

probability 1–p, a Poisson (λ) random variable is observed. For example, in a different

context, the same model can be used in quality control. Thus, when a manufacturing

system is properly aligned, defects are nearly impossible, and the p is large. But when

the machine is misaligned, defects may occur according to a Poisson (λ) distribution.

This same principle is also plausible in motor insurance when modelling the number of

accidents per year. Some drivers hardly use their vehicle or use it very rarely, so for

them the probability of not being involved in an accident should be large.

Both the probability of no accidents and the mean number of defects λ in the imperfect

state (when people use their cars) may depend on covariates that are defined for each

individual. Here, we have not included subscript i to refer to the i-th observation in a

sample of size n, to make notation easier. Sometimes p and λ are unrelated; but on other

occasions p is a simple function of λ, such as p = l/(1 + λT) for an unknown constant T.

In either case, ZIP regression models are easy to fit. Maximum likelihood estimates

(MLE) are approximately normal in large samples, and confidence intervals can be

constructed by inverting likelihood ratio tests or using the approximate normality of the

MLE. The estimation can be performed with standard statistical software, such as R or

SAS, but the interpretation of the results of a ZIP regression model is not

straightforward. For example, the 1992 article by Lambert(30) reports that in an

experiment involving soldering defects on printed wiring boards, two sets of conditions

resulted in roughly the same mean number of defects; however, the perfect state was

more likely under one set of conditions and the mean number of defects in the imperfect

state was smaller under the other set. In other words, ZIP regression can show not only

7
which conditions give the lower mean number of defects but also why the means are

lower.

Notice that formally we introduce an extended model of zero claims in insurance using

distance driven as the exposure to risk variable. However, while this simple model

extension primarily improves understanding of zero claims, it may have another

important effect. When factors other than just mileage are included in the model, then

essentially the extension suggested here also serves as a bias correction. If the effect of

risk exposure through distance is not log-linear, for example, then our extended model

adjusts for that. With the data provided herein, the adjustment via our extended model

improved considerably when mileage was included, and only marginally when further

variables were included. Finally, therefore, we opted only to include mileage in the

extension of the model, thus facilitating a straightforward interpretation. In this way, the

excess zeros in our extended model are simply interpreted as a function of miles driven.

In the zero part of the model, we have only a Bernoulli variable that distinguishes

between the zero event (no claim) versus the non-zero event (at least one claim), so the

expectation for this binary response random variable is exactly the probability of excess

zero claims, which should be limited to the [0,1] interval. For this reason, we have no

offset in this part and the parameter of the log-distance is not necessarily equal to one.

Below we first introduce the simple Poisson model with and without exposure as it has

traditionally been presented. Exposure, in our study, is equivalent to miles driven per

year.

8
2.1. The Poisson model

Let us assume that given xi, the dependent variable Yi follows a Poisson distribution

with parameter i , which is a function of the linear combination of parameters and
regressors, 𝛽0 + 𝛽1 𝑥𝑖1 + ⋯ + 𝛽𝑘 𝑥𝑖𝑘 . Indeed,

𝐸 (𝑌𝑖 |𝑥𝑖 ) = i = exp(𝛽0 + 𝛽1 𝑥𝑖1 + ⋯ + 𝛽𝑘 𝑥𝑖𝑘 ). (1)

The unknown parameters to be estimated are (𝛽0 , … , 𝛽𝑘 ).

2.2. The Poisson model with exposure

When exposure to risk is introduced, then an offset is included in the model. Let us call

𝑇𝑖 the exposure factor for policy holder i (i=1,…,n), in our case Ti=ln(𝐷𝑖 ), where 𝐷𝑖

indicates distance travelled. Then the model can incorporate this factor as follows:

𝐸 (𝑌𝑖 |𝑥𝑖 , 𝑇𝑖 ) = 𝐷𝑖 exp(𝛽0 + 𝛽1 𝑥𝑖1 + ⋯ + 𝛽𝑘 𝑥𝑖𝑘 ) = exp(𝑇𝑖 )𝜆𝑖 . (2)

Under this model, the probability of zero using the Poisson distribution is calculated as

follows, P(𝑌𝑖 = 0) = exp(−𝐷𝑖 𝜆𝑖 ), so it depends on the distance and, since 𝜆𝑖 is always

positive by definition, then the probability of zero claims declines naturally as distance

driven increases.

We are now ready to extend the traditional Poisson regression models above to include

excess zeros via ZIP models. This extension is also introduced with and without

exposure.

9
2.3. The zero-inflated Poisson model

In the ZIP model, the probability of zero is specified as follows:

P(𝑌𝑖 = 0) = 𝑝𝑖 + (1 − 𝑝𝑖 )P(𝑌𝑖 ∗ = 0), (3)

where 𝑝𝑖 is the probability of the perfect, zero defect state and (1-𝑝𝑖 ) is the probability

of the complementary state. The new 𝑌 ∗ variable follows a Poisson distribution with

parameter exp(𝛽0 + 𝛽1 𝑥𝑖1 + ⋯ + 𝛽𝑘 𝑥𝑖𝑘 ) and captures the claims distribution that is not

contaminated by the excess of zeros. Note that 𝑝𝑖 may depend on some covariates.

Under this model, the probability of suffering k accidents, when k is bigger than or

equal to one is:

P(𝑌𝑖 = k) = (1 − 𝑝𝑖 )P(𝑌𝑖 ∗ = k), k>0.

2.4. The zero-inflated Poisson model with exposure

Here we assume that 𝑝𝑖 is the probability of an excess of zeros for the i-th observation

and it is specified as a logistic regression model such that

exp(𝛼0 + 𝛼1 ln(𝐷𝑖 ))
𝑝𝑖 =
1 + exp(𝛼0 + 𝛼1 ln(𝐷𝑖 ))
(4)

10
The Poisson model for 𝑌 ∗ is specified as follows, with an exposure, 𝐸 (𝑌𝑖 ∗ |𝑥𝑖 , 𝑇𝑖 ) =

𝐷𝑖 exp(𝛽0 + 𝛽1 𝑥𝑖1 + ⋯ + 𝛽𝑘 𝑥𝑖𝑘 ) = 𝐷𝑖 𝜆𝑖 = exp(ln(𝐷𝑖 ))𝜆𝑖 = exp(𝑇𝑖 )𝜆𝑖 . Then,

exp(𝛼0 + 𝛼1 ln(𝐷𝑖 )) 1
P(𝑌𝑖 = 0) = + exp(−𝐷𝑖 𝜆𝑖 )
1 + exp(𝛼0 + 𝛼1 ln(𝐷𝑖 )) 1 + exp(𝛼0 + 𝛼1 ln(𝐷𝑖 ))

1
𝑃(𝑌𝑖 = k) = (exp(−𝐷𝑖 𝜆𝑖 ))𝐷𝑖 𝑘 𝜆𝑖 𝑘 /𝑘!
1 + exp(𝛼0 + 𝛼1 ln(𝐷𝑖 ))

Using the definition of the expectation of a discrete random variable, the expectation of

the Poisson part is

1 (5)
𝐸 (𝑌𝑖 |𝑥𝑖 , 𝑇𝑖 ) = (1 − 𝑝𝑖 )𝐸 (𝑌 ∗ 𝑖 |𝑥𝑖 , 𝑇𝑖 ) = 𝐷 𝜆 = 𝐷 ∗ 𝑖 𝜆𝑖
1 + exp(𝛼0 + 𝛼1 ln(𝐷𝑖 )) 𝑖 𝑖

𝐷𝑖
where 𝐷∗ 𝑖 = 1+exp(𝛼 is a transformation of the original exposure 𝐷𝑖 . So, when
0 +𝛼1 ln(𝐷𝑖 ))

we include zero-inflation there is a transformation of the exposure in the Poisson model.

Let us study the transformation. If 𝛼1 > 1, when 𝐷𝑖 is large then 𝐷 ∗ 𝑖 tends to zero, but

when 𝛼1 < 1 then 𝐷∗ 𝑖 increases when 𝐷𝑖 increases. On the other hand, when 𝐷𝑖 tends

to zero, 𝐷 ∗ 𝑖 tends to zero.

If we examine the logistic regression part (4), we observe that 𝑝𝑖 can be understood

again as a transformation of the exposure into the [0,1] interval, which tends to zero

when 𝐷𝑖 tends to zero if 𝛼1 is positive. Moreover, the derivative of (5) with respect to

𝐷𝑖 shows how much the expected claims would change as a function of 𝐷𝑖 and indicates

that if 𝛼1 is significantly different from zero, then the derivative is not 𝐷𝑖 times 𝜆𝑖 .

Since insurance premiums are based on expected number of claims, this is an important

11
result as it potentially shows that insurance prices should not necessarily be linearly

proportional to distance driven.

3. DATA

We use information on the risk exposure and number of claims for 25,014 insureds with

car insurance coverage throughout 2011, that is, individuals exposed to the risk for a

full year. Note that in our case these data concern drivers up to a maximum age of 37,

given that the insurance product was sold primarily to young drivers. Our aim is to

discriminate between good and bad drivers in this portfolio segment and to identify the

influence of driving short distances.(16) Claim frequencies are presented in Table I, with

an expected value of 0.23 claims per person. Table I has information on the frequency

of all reported claims. The sum of reported claims that were not at fault is 3,108, while

the sum of claims at fault is 2,652. Overall 5,760 claims were reported. Descriptive

statistics for the risk exposure indicator (kilometres per year) are presented in Table II,

where we analyse drivers with and without claims separately. The rest of the indicators,

both those derived from traditional ratemaking factors and those obtained from

telematic devices, are presented in Table III, where we also present the definitions of

these variables and their main descriptive statistics.

Table I. Frequency of claims per driver (n=25,014)

in the Spanish insurance data set (all claims and by fault)
Absolute frequency per driver
Number of claims All claims Claims at fault Claims not at fault
0 20,608 22,837 22,432
1 3,310 1,750 2,111
2 889 385 424
3 165 37 40
4 34 4 6

12
 5 7 1 1
6 1 0 0
One insured driver had 6 claims, 2 were at fault and 4 where not at fault.

Table II. Descriptive statistics for the risk exposure indicator

(total kilometres travelled per year in 000s)

All Sample Drivers with no claims Drivers with claims

n = 25,014 n = 20,608 (82.4%) n = 4,406 (17.6%)
Mean 7.16 6.99 7.96
1st Quartile 4.14 4.00 4.87
Median 6.46 6.28 7.22
3rd Quartile 9.40 9.22 10.30
Standard Deviation 4.19 4.14 4.35

The results presented in Table II in relation to the annual distance travelled by the

insured drivers reveal differences between those with no claims and those with claims.

If we focus on the 25% of drivers that travelled the smallest distance over the year (1st

quartile), we observe that the insureds that claim at least one accident drove more

kilometres per year than those with no claims – the respective quartile values being 4.87

vs. 4.00. A similar pattern of behaviour is observed for the second (median) and third

quartiles with those making claims driving larger distances than those with no claims.

This result was as expected and is a clear indication of a relationship between claims

and distance driven.

The Mann Whitney test is a nonparametric test of the null hypothesis that it is equally

likely that a randomly selected value from one sample is less than or greater than a

randomly selected value from a second sample. The Mann-Whitney test shows that the

differences in the mean for the exposure risk regressor (Table II), as well as for the

other classical and telematic regressors (Table III) are statistically significant in the

cases of drivers with no claims and drivers with claims, with the exception of vehicle

age (p-value=0.331) and the percentage of kilometres driven over the speed limit

squared (p-value=0.9293). Note that the normality hypothesis of these variables is

13
 rejected when using the Kolmogorov-Smirnov test. The Kolmogorov-Smirnov test is a

nonparametric test of the equality of continuous, one-dimensional probability

distributions that can be used to compare the statistical distribution of two samples.

From a univariate point of view, drivers that made a claim for at least one accident are,

on average, younger than those that made no claim and have held their driving licence

for fewer years. A similar conclusion can be drawn in the case of ownership of a

powerful vehicle, where those insureds making at least one claim present a higher value

than those making no claims. Unexpectedly, in the case of cars parked overnight in a

garage, the percentage value is higher among those who made at least one claim than it

is among those who made no claim. We would expect such cars to be safer, but it

appears that this variable may be closely related to car type, with powerful, more

expensive cars being kept in garages. As for the new driving behaviour indicators

derived from telematics, driving at night and driving in urban areas present larger mean

values in the claims group than in the no claims group.

Table III. Explanatory variables* included in the models and descriptive statistics

Drivers with no
All sample Drivers with claims
claims
Description
Std. Std. Std.
Mean Mean Mean
Dev. Dev. Dev.

Traditional ratemaking factors

Age Age of the insured driver (in years) 27.57 3.09 27.65 3.09 27.18 3.10
Age2 Age squared of the insured driver
Male (%) Sex of the insured driver (1 if male, 0 female) 48.91 - 48.61 - 50.32 -
Age Driving Licence Experience of the insured driver 7.17 3.05 7.27 3.07 6.73 2.94
Vehicle age Age of the insured vehicle 8.75 4.17 8.76 4.19 8.69 4.11
Power Power of the insured vehicle 97.22 27.77 96.98 27.83 98.36 27.46
1 if the vehicle is parked in a garage over night, 0
Parking (%) 77.38 - 77.21 - 78.17 -
otherwise

New telematic ratemaking factors

Percentage of kilometres travelled at night during
Km per year at night (%) 6.91 6.35 6.85 6.32 7.16 6.49
the year
Km per year at night (%)2 Percentage of kilometres travelled at night squared
Percentage of kilometres during the year above the
Km per year over speed limit (%) 6.33 6.83 6.28 6.87 6.60 6.59
speed limits
2 Percentage of kilometres during the year above the
Km per year over speed limit (%)
speed limits squared
Percentage of kilometres travelled in urban areas
Urban km per year (%) 25.87 14.36 25.51 14.31 27.56 14.47
during the year
* In addition to risk exposure (km per year in 000s)

14
4. RESULTS

Tables IV and V present the zero-inflated Poisson models including exposure to risk

(kilometres driven per year) as the offset variable in the models as discussed in section

2. Figure 1 gives an overview of the estimated models.

Figure 1: Summary of the estimated models

Traditional software programs facilitate the maximum likelihood estimation of these

models, their results being obtained using SAS, PROC GENMOD. To compare the

models, we use the Akaike Information Criterion (AIC), calculated as twice the number

of parameters in the model minus twice the value of the log-likelihood in the maximum.

The best model is the one that presents the smallest AIC value. 1

1
The AIC penalizes the number of parameters less strongly than the Bayesian information criterion
(BIC), which is calculated on the basis of the logarithm of the number of observations as opposed to
multiplying the number of parameters by two, as with the AIC.

15
Table IV highlights a clear improvement in the results when considering all the model

regressors (the lowest AIC value being obtained for the first specification). These

results seem to validate the conclusions drawn in previous studies, (2–4) in which the

relevance of the new indicators related to distance travelled and driving habits is

highlighted, but where they are used in conjunction with the classical regressors.

Individual significance is observed for a large number of parameters, including those of

the logit model in its zero-inflation part. On first inspection, the positive sign of the

parameter associated with the log-distance in the logistics part might seem surprising

and it could be interpreted erroneously. This value (0.404) in the first column does not

mean that the greater the distance driven, the greater is the probability of the insured

having zero claims. Rather it means that the greater the distance driven, the greater is

the proportion of excess zero claims, indicating a deviation from the Poisson

distribution that can be captured by the ZIP model.

Table IV. Zero-inflated Poisson model with offsets (Log of km per year in 000s). All
types of claim.
All variables Non-telematics Telematics

Coefficient (p-value) Coefficient (p-value) Coefficient (p-value)

Poisson part
Intercept -2.148 0.045 -0.829 0.440 -3.461 <.001
Age -0.094 0.232 -0.123 0.121
Age2 0.002 0.221 0.002 0.131
Male -0.068 0.029 -0.011 0.719
Age Driving Licence -0.059 <.001 -0.067 <.001
Vehicle Age 0.014 <.001 0.017 <.001
Power 0.003 <.001 0.001 0.017
Parking 0.029 0.420 0.032 0.381

Log of km per year 1.000 -- 1.000 -- 1.000 --

(thousands) - offset
Km per year at night (%) -0.004 0.312 -0.001 0.771
Km 0.0001 0.467 0.000 0.931
per year at night (%)2
Km per year over speed 0.019 0.001 0.018 0.001
limit (%)
Km per year over speed -0.001 0.001 -0.001 0.003
limit (%)2
Urban km per year (%) 0.026 <.001 0.027 <.001

Zero-inflation part
Intercept (Logit) -0.847 <.001 -1.639 <.001 -0.795 <.001

16
Log of km per year 0.404 <.001 0.824 <.001 0.406 <.001
(thousands) (Logit)
AIC 28,877.112 29,427.423 29,005.172
BIC 28,999.019 29,508.694 29,070.189

In the case of the classical variables, all the parameters for gender, driving experience,

vehicle age and the power of the vehicle are statistically significant. Thus, we find an

increasing expectation in the number of claims for women drivers as opposed to men,

inexperienced drivers as opposed to experienced, and owners of old and powerful

vehicles as opposed to owners of newer and less powerful cars. As for the new telematic

regressors, two – the percentage of kilometres per year driven over the speed limit and

the percentage of urban kilometres driven per year – are significant in explaining the

expected number of claims. Thus, the number of claims increases as these two

regressors increase. No significance is observed in the case of night driving. Finally, if

we compare the results of the second and third specifications (columns 2 and 3,

respectively), the best results are obtained for the model that only includes variables

related to driving habits (telematics), as indicated by its lower AIC value.

Our model predicts the highest number of expected claims for younger women, with

little driving experience, driving old and powerful vehicles, driving in urban zones, and

exceeding the speed limit. Note that this result is in line with the results reported by

Mercer in 1989.(23)

Previous research(34) has shown that it may be interesting to include Age and Gender

interaction in the model. The results for all the models, which are available from the

authors, show that this interaction is not significant. In practice, Gender cannot be used

for pricing insurance in the EU, but it can certainly be used for risk evaluation and it can

help to understand male/female differences with implications on traffic safety. Our

17
conclusion for this sample is that there is no interaction between Age and Gender. There

are potentially two reasons for that (1) the sample consists of drivers aged less than 37

years, so Age may not have enough range to show a significantly different effect by

Gender or (2) as found by other authors, the influence of Gender is masked by the fact

that men on average drive significantly longer distances than women. The relationship

between distance driven and Gender was discovered by independent researchers in

(35)
different EU countries considering average daily distance in a Spanish data set , or

using average trip distance for a Belgian sample(36) or even taking both average trip

distance and total distance in another European portfolio sample (37),. They all concluded

that Gender differences in the risk of accidents are, to a large extent, attributable to the

fact that men drive longer average distances than women.

Similar results are obtained when only claims at fault are considered (Table V), with the

exception that the age of the driver is now significant while gender is not. Here, again, a

better goodness of fit is obtained for the specification that includes all variables (both

telematic and non-telematic) and the model that includes only the telematics variables

(the lowest AIC value being obtained for served column 1). As in Table IV, a lower

AIC is obtained for the specification using only telematic variables as opposed to that

using only classical variables (columns 2 and 3, respectively).

Table V. Zero-inflated Poisson model with offsets (Log of km per year in 000s). Claims
for which the policyholder was at fault
All variables Non-telematics Telematics

Coefficient (p-value) Coefficient (p-value) Coefficient (p-value)

Poisson part
Intercept -0.697 0.653 0.278 0.857 -3.892 <.001
Age -0.224 0.050 -0.224 0.049
Age2 0.004 0.039 0.004 0.045
Male 0.000 0.998 0.076 0.093
Age Driving License -0.083 <.001 -0.088 <.001

18
Vehicle Age 0.013 0.015 0.016 0.004
Power 0.001 0.163 0.001 0.351
Parking -0.035 0.497 -0.025 0.637

Log of km per year 1.000 -- 1.000 -- 1.000 --

(thousands) - offset
Km per year at night (%) 0.0052 0.386 0.010 0.083
Km per year at night -0.0001 0.685 -0.0002 0.272
(%)2
Km per year over speed 0.035 <.001 0.031 0.000
limit (%)
Km per year over speed -0.001 0.001 -0.001 0.001
limit (%)2
Urban km per year (%) 0.024 <.001 0.026 <.0001

Zero-inflation part
Intercept (Logit) -0.228 0.151 -0.765 <.001 -0.140 0.358
Log of km per year 0.442 <.001 0.743 <.001 0.441 <.001
(thousands) (Logit)
AIC 16,912.217 17,125.313 17,004.642
BIC 17,034.124 17,206.584 17,069.659

The age of at-fault drivers is inversely related to the expected number of claims, that is,

a higher number of accidents are expected among younger drivers. However, the

significance of the age squared parameter indicates a non-linear relationship between

the two variables. Inexperienced drivers (measured in terms of the number of years in

which they have been in possession of a driving licence) and drivers of old vehicles

show a higher expected number of claims than that recorded by their more experienced

counterparts and drivers of newer vehicles. In common with the result in Table IV, the

percentage of kilometres per year driven over the speed limit, and additionally here the

percentage of kilometres driven at night, have an impact on the expected number of

claims in which the driver is at fault. The percentage of kilometres driven at night is

significant at the 10% level when we only consider the telematic variables but the AIC

value for this model is lower than that obtained for the first model.

Results for the models on the not at fault claims indicate similar conclusions. We have

not discussed the not at fault cases because in insurance premium calculation only

claims at fault are of main interest. Claims at fault indicate that the driver has caused an

19
accident, while not at fault means that the accident was due to someone else. If the

accident is caused by someone else, then the insured driver should not pay a higher

insurance premium compared to someone who did not report a claim.

Comparisons with the classical Poisson model with offsets (without considering zero

inflation), both for the total sample and for claims where the policyholder is at-fault, are

not included here, but they do not enable us to see the impact of distance on the excess

of zeros. These results are available on request from the authors. The goodness of fit

results are always better in the zero-inflated models because they take into account

differences between false zeros (non-risk exposure) and true zeros (risk exposure and

zero claims).

In a similar context, it has been shown that prediction models for hurricane power

outage can be improved by a new two-step outage prediction model and the inclusion of

additional environmental variables that increase the overall accuracy(38). Our model also

improves the classical approach by introducing telematics information into the

prediction of the number of claims and this can be done in a two stage model

approach(2).

We have performed a holdout analysis, and we have tested the models against test sets

which were not used in the training process. We have chosen a 70% training sample,

versus a 30% holdout sample. In all cases we have confirmed the conclusions on the

significance of the parameter that we had in the initial analysis. The Chi squared test of

differences between observed and fitted frequencies was equal to 946.7 for the whole

sample. The hold out analysis indicates very similar values (1,041.3 with 6 degrees of

freedom in the training sample and 1,005.9 with 6 degrees of freedom in the test sample

for the model of all claims and all variables).

20
In order to evaluate the variable importance, we have estimated the models using

standardized covariates, so that we can compare the coefficients. This analysis reveals

that the most important factor that determines the risk of a crash is the percentage in

urban driving, followed by the age of the driver’s license. The third factor is the

percentage of speed limit violation. The least relevant factors are the age of the vehicle,

gender of the driver, percent of night distance driven and parking in a garage.

5. CONCLUSIONS

We have shown that the part of the zero accident frequency not explained by traditional

insurance risk factors increases with the distance driven by the policy holder. This

means that when considering policy holders with the same characteristics but with

different exposures to risk in terms of distance driven per year, we can conclude that

those with a greater exposure present a larger proportion of excess zero claims than

those with less exposure. This can be understood as an indication of a learning effect, or

in terms of distance driven, that even if exposure to risk increases with distance driven,

the probability of not making a claim also increases compared to that of drivers in the

group that drive a shorter distance. This finding is evidence of the fact that good drivers

– if we identify them with those reporting no claims – are more frequent than expected

among the group of drivers that drive long distances than among those that drive shorter

distances, all other things being equal.

This conclusion has a direct impact on the future design of PAYD insurance products,

insofar as the premium paid should not be strictly proportional to the distance driven.

Moreover, the premium should take into account the learning effect analysed here. One

21
possible solution would be to make the marginal increase in the insurance price per

kilometre driven dependent on the accumulated distance. Here, we have shown that this

relationship is not linearly dependent, as we report that the zero-inflation part plays a

significant role. Taking the derivative of (5) makes this non-linearity immediately

apparent.

The probability of excess zeros increases with distance. The coefficient for the

logarithm of the number of kilometres driven per year in the logit model (which predicts

zero inflation) is positive, i.e. the probability of observing false zeros increases with

increasing distance. Moreover, we have shown that the ZIP model gives better results in

terms of goodness of fit than those obtained with the classical Poisson model (non-zero-

inflated Poisson model).

Here, therefore, we have shown both the significance of the impact of the distance

variable coefficient and the positive relationship between traffic violations involving

excess speed and urban driving with the expected number of claims. These results are in

line with reports issued by official traffic institutions where it is argued that speed limit

violations should be considered in the design of insurance premiums so that safer

driving is rewarded.(20)

Previous traffic studies published in Risk Analysis (22,23) have stressed the desirability of

including risk exposure in terms of distance driven. We have shown that indeed vehicle

telemetry, and the collection of information using GPS-based technology such as

percentages of kilometres driven at night, over the speed limit, and in urban zones,

among others can be included in the ratemaking process thus improving the results

obtained when just using classical driver variables, such as age and gender. This opens

22
the question whether pay-as-you drive should also consider a different price per mile

depending on the time of the day and the location.

Our study shows that ZIP models with mileage as their offset variable can improve the

definition of drivers’ risk profiles and provide valuable policy guidelines that might be

implemented to improve driving behaviour. Furthermore, the higher premium

associated with a higher percentage of kilometres driven in an urban area (as a

consequence of a higher expected number of claims) could discourage the use of private

vehicles in cities, as called for by various European institutions (not least to reduce

levels of pollution). Clearly, similar conclusions can be drawn in terms of traffic

violations, with an increase in the premium for drivers with a tendency to exceed the

speed limit.

REFERENCES

1. World Health Organization (2017). “10 facts on global road safety. Updated July

2017” available at http://www.who.int/features/factfiles/roadsafety/en/.

2. Ayuso, M, Guillen M., Nielsen, JP. Improving automobile insurance ratemaking

using telematics: incorporating mileage and driver behaviour data.

Transportation, 2018; accepted, in press..

3. Lemaire J, Park SC, Wang KC. The use of annual mileage as a rating variable.

ASTIN Bulletin, 2016; 46(1): 39-69.

4. Ferreira J, Minikel E. Measuring per mile risk for Pay-As-You-Drive auto

insurance. Transportation Research Record: Journal of the Transportation

Research Board, 2013; 2297(10): 97-103.

23
5. Paefgen J, Staake T, Fleisch E. Multivariate exposure modelling of accident risk:

Insights from Pay-as-you-drive insurance data. Transportation Research Part A:

Policy and Practice, 2014; 61: 27-40.

6. Paefgen J, Staake T, Thiesse F. Evaluation and aggregation of pay-as-you-drive

insurance rate factors: a classification analysis approach. Decision Support

Systems, 2013; 56: 192-201.

7. Langford J, Koppel S, McCarthy D, Srinivasan S. In defence of the ‘low-

mileage bias’. Accident Analysis and Prevention, 2008; 40(6): 1996-1999.

8. Sivak M, Luoma J, Flannagan MJ, Bingham CR, Eby DW, Shope JT. Traffic

safety in the U.S.: Re-examining major opportunities. Journal of Safety

Research, 2007; 38(3): 337-355.

9. Litman, T. Pay-As-You-Drive pricing and insurance regulatory objectives.

Journal of Insurance Regulation, National Association of Insurance

Commissioners, 2005; 23(3): 35-53.

10. Edlin AS. Per-mile premiums for auto insurance. In: Arnott R, Greenwald B,

Kanbur R, Nalebuff B (eds.) Economics for an imperfect world: essays in honor

of Joseph E. Stiglitz. Cambridge, MA: MIT Press, 2003.

11. Boucher JP, Denuit M, Guillen M. Number of accidents or number of claims?

An approach with zero-inflated Poisson models for panel data. Journal of Risk

and Insurance, 2009; 76: 821-846.

12. Vickrey W. Auto accidents, tort law, externalities and insurance: an economist’s

critique. Law and Contemporary Problems, 1968; 33(3): 464-487.

13. Shafique MA, Hato E. Use of acceleration data for transportation mode

prediction. Transportation, 2015; 42(1): 163-188.

24
14. Xu Y, Shaw SL, Zhao Z, Yin L, Fang Z, Li Q. Understanding aggregate human

mobility patterns using passive mobile phone location data: a home based

approach. Transportation, 2015; 42(4): 625-646.

15. Ellison AB, Bliemer MCJ. Greaves SP. Evaluating changes in driver behaviour:

a risk profiling approach. Accident Analysis and Prevention, 2015; 75: 298-309.

16. Ayuso M, Guillen, M, Perez-Marin, AM. Time and distance to first accident and

driving patterns of young drivers with pay-as-you-drive insurance. Accident

Analysis and Prevention, 2014; 73: 125-131.

17. Underwood G. On-road behaviour of younger and older novices during the first

six months of driving. Accident Analysis and Prevention, 2013; 58: 235-243.

18. Jun J, Guensler R, Ogle J. Differences in observed speed patterns between crash-

involved and crash-not-involved drivers: application of in-vehicle monitoring

technology. Transportation Research Part C, 2011; 19: 569-578.

19. Elias W, Toledo T, Shiftan Y. The effect of daily-activity patterns on crash

involvement. Accident Analysis and Prevention, 2010; 42(6): 1682-1688.

20. Ayuso M, Guillen M, Alcañiz M. The impact of traffic violations on the

estimated cost of traffic accidents with victims. Accident Analysis and

Prevention, 2010; 42: 709-717.

21. Baecke, P, Bocca, L. The value of vehicle telematics data in insurance risk

selection processes, Decision Support Systems, 2017; 98, 69–79.

22. Segui-Gomez M, Lopez-Valdes FJ, Guillen-Grima F, Smyth E, Llorca J, Irala J.

Exposure to traffic and risk of hospitalization due to injuries. Risk Analysis,

2011; 31(3): 466-474.

23. Mercer GW. Traffic accidents and convictions: group totals versus rate per

kilometer driven. Risk Analysis, 1989; 9(1): 71-77.

25
24. Gourieroux C, Monfort A, Trognon A. Pseudo maximum likelihood methods:

Theory. Econometrica: Journal of the Econometric Society, 1984a; 681-700.

25. Gourieroux C, Monfort A, Trognon A. Pseudo maximum likelihood methods:

applications to Poisson models. Econometrica: Journal of the Econometric

Society, 1984b; 701-720.

26. Chiappori PA, Salanié B. Testing for asymmetric information in insurance mar-

kets. Journal of Political Economy, 2000; 108 (1): 56-78.

27. Dionne G, Vanasse C. Automobile insurance ratemaking in the presence of

asymmetrical information. Journal of Applied Econometrics, 1992; 7 (2): 149-

165.

28. Cameron AC, Trivedi PK. Regression analysis of count data. Cambridge

University Press, 2nd Ed., USA, 2013.

29. Winkelmann R. Econometric analysis of count data. Springer-Verlag, 4th Ed.,

Berlin, 2003.

30. Lambert D. Zero-inflated Poisson regression with an application to defects in

manufacturing. Technometrics, 1992; 34(1): 1-14.

31. Boucher JP, Denuit M, Guillen M. Risk classification for claim counts: a

comparative analysis of various zero-inflated mixed Poisson and hurdle models.

North American Actuarial Journal, 2007; 11(4): 110-131.

32. Lord D, Washington S, Ivan J. Poisson, Poisson-Gamma and zero-inflated

regression models of motor vehicle crashes: balancing statistical fit and theory.

Accident Analysis and Prevention, 2005; 37: 35-46.

33. Sarul, LS, Sahin, S. An application of claim frequency data using zero inflated

and hurdle models in general insurance. Journal of Business, Economics and

Finance, 2015; 4(4): 732-743.

26
34. Mercer, GW. Influences on Passenger Vehicle Casualty Accident Frequency and

Severity: Unemployment, Driver Gender, Driver Age, Drinking Driving and

Restraint Device Use. Accident Analysis and Prevention, 1987; 19: 231-236.

35. Ayuso, M, Guillen, M, Pérez-Marín, AM. Telematics and gender discrimination:

some usage-based evidence on whether men’s risk of accidents differs from

women’s. Risks, 2016; 4(2): 1-10.

36. Verbelen, R, Antonio, K, Claeskens, G. Unraveling the predictive power of

telematics data in car insurance pricing. Applied Statistics, 2018. Online

https://doi.org/10.1111/rssc.12283.

37. Wüthrich, MV. Covariate selection from telematics car driving data. European

Actuarial Journal, 2017; 7(1): 89-108.

38. McRoberts, DB, Quiring, SM, Guikema, SD. Improving hurricane power

outage prediction models through the inclusion of local environmental factors.

Risk Analysis, 2016; First published: 25 October 2016,

https://doi.org/10.1111/risa.12728.

27
Institut de Recerca en Economia Aplicada Regional i Pública Document de Treball 2014/17, pàg. 5
Research Institute of Applied Economics Working Paper 2014/17, pag. 5