A possibilistic and probabilistic approach to precautionary saving

Summary
This paper proposes two mixed models to study a consumers optimal saving in the presence of two
types of risk: income risk and background risk. In the first model the income risk is represented by a fuzzy
number and the background risk by a random variable. In the second model the income risk is represented by a
random variable and the background risk by a fuzzy number. For each of these models, three notions of
precautionary saving are defined as indicators of the extra saving induced by the income and the background
risk on the consumers optimal choice. As a conclusion, we can characterize conditions allowing for extra
saving relative to the optimal saving under certainty, even when a certain component of risk is modeled with
fuzzy numbers.

Key words: Optimal saving, background risk, income risk, possibility theory.
JEL: CO2, D31, D81
Nowadays it is widely expressed a concern about the huge saving rate of Chinese households, which
has far-reaching international implications in terms of Chinas current account surpluses and the Western
external deficits. Many analysts and policy makers are trying to trace back the reasons for such behavior and
predict how long it will last. In this regard, many of the potential explanations for this phenomenon point to
economic uncertainty as a determinant of precautionary saving (Marcos Chamon, Kai Liu and Eswar Prasad
(2013)). In contrast to the Chinese example, US households are well known for their low saving rate, which
calls for some precise explanations as well.
Therefore, it is very important to understand the characteristics of such uncertainty in order to model
the households optimal behavior in terms of consumption and saving. Both the kind of uncertainty and the
features of the consumers preferences must be analyzed, in order to predict micro or macroeconomic saving
patterns. In this context, the notion of precautionary saving has long appeared in models of economic decision
under uncertainty. It measures the way adding a source of risk modifies the optimal saving. When the
consumers utility function is unidimensional and the risk situation is described by a random variable, extra
saving will appear in response to uncertainty when the third order derivative of the utility function is positive.
On the other hand, several authors (Neil A. Doherty and Harris Schlesinger (1983), Christian Gollier
and John W. Pratt (1996), John W. Pratt (1998)) studied economic decision processes governed by two types of
risk: primary risk (income risk) and background risk (loss of employment, divorce, illness, etc.). In our paper,
as in Mario Menegatti (2009), the presence of background risk will be associated to a nonfinancial variable and
will be uninsurable, nevertheless having an influence on the optimal solution for economic decisions (see e.g.
Louis Eeckhoudt, Christian Gollier and Harris Schlesinger (2005)).
The way the interaction of both types of risk affects optimal saving was studied by Christophe
Courbage and Beatrice Rey (2007) and Mario Menegatti (2009). These models assume that the consumers

activity takes place during two periods and both types of risk act during the second period. The presence or the
absence of one of the two types of risk leads to several possible uncertainty situations, which enables the
definition of several notions of precautionary saving (e.g. in Menegatti (2009) two of such notions are studied).
All the optimal saving models in the literature are based on probability theory. That is, both the
primary and the background risk are modeled as random variables. However, there are risk situations for which
probabilistic models are not appropriate (e.g. for small databases). Lotfi A. Zadeh (1978)s possibility theory
offers another way to model some risk situations. Here risk is modeled with possibility distributions
(particularly, with fuzzy numbers) and the well known probabilistic indicators (e.g. expected value, variance,
covariance) are replaced by the corresponding possibilistic indicators.
Due to the complexity of economic and financial phenomena, one can have mixed situations in which
some risk parameters should be probabilistically modeled with random variables, and other risk parameters
should be possibilistically modeled with fuzzy numbers. Then we can consider the following four possible
situations: (1) a random variable captures the primary risk and also a random variable captures the background
risk; (2) a fuzzy number captures the primary risk and a fuzzy number captures the background risk; (3) a
fuzzy number captures the primary risk and a random variable captures the background risk; (4) a random
variable captures the primary risk and a fuzzy number captures the background risk.
The situation (1) was treated in the abovementioned probabilistic models. The purpose of this paper is
studying the precautionary saving motive in situations (3) and (4). In the case of situation (3), the risk situation
is described by a mixed vector (A, X); and in the case of situation (4) the risk situation is described by a mixed
vector (Y, B); where A, B are fuzzy numbers and X, Y are random variables. Let us denote the mixed models
described by situation (3) as type I models; we will also denote the mixed models described by situation (4) as
type II models. For each of those two types of models (type I and type II) we will define three notions of
precautionary saving and investigate the necessary and sufficient conditions for extra saving to arise, after
adding primary risk, background risk or both of them. The main results of the paper establish those necessary
and sufficient conditions, expressed in terms of third-order partial derivatives of the bidimensional utility
function and in terms of the probabilistic and possibilistic variances associated with the mixed vector.
Our three notions of precautionary saving will be defined below in detail. However, we can anticipate
that the first notion refers to the extra saving arising when a small income risk is introduced, relative to the
optimal saving under certainty. The second notion refers to the extra saving arising when a small background
risk is introduced, relative to the optimal saving under certainty. And finally, the third notion refers to the extra
saving arising when both a small income risk and a small background risk are introduced, relative to the
optimal saving under certainty. Each of the three notions requires its own necessary and sufficient conditions
for each kind of precautionary saving to be positive.
We can offer illustrative examples of type I and type II models in real life. For instance, suppose a
man owns a house near the beach surrounded by a wheat field he needs to harvest every year. If the rainfall
during the year is very scarce, the annual income from the wheat harvest will be lower. But, simultaneously,
the renting value of a room in his house could be higher.
Assume now that only the risk related to the harvest is insurable. However, the background risk is
related to the potential lodgers pleasure from sunbathing, which could be described by a fuzzy number:
weather near the beach is pleasant enough to sunbathe. Both risks would be negatively correlated, since
under a dry and sunny weather the owner would collect little revenue from the harvest, but probably he could

charge a higher rent to his lodger. Therefore, there may or may not be precautionary saving in response to
those two risks. That would be an example we could characterize as a mixed model of type II.
Imagine now that, instead, the main income source for the owner is the apartment rent. The size of the
rent could be now insurable for him unlike the harvest revenue. In that case, the income risk would be
characterized by a fuzzy number with a value dependent on the lodgers imprecise perception of the weather.
The situation could be then assimilated to a mixed model of type I. The objective of this paper is exploring the
conditions for precautionary saving in situations like these two described cases.
We will describe now briefly the structure of the paper. Section 1 contains a brief historical
perspective on the contributions dealing with the unidimensional precautionary saving problem, and also with
precautionary saving in probabilistic models with background risk. Section 2 recalls the indicators of fuzzy
numbers and the mixed expected utility of Irina Georgescu and Jani Kinnunen (2011). Section 3 presents the
mathematical framework in which the optimal saving models with background risk are embedded. Section 4
proposes mixed models of optimal saving of type I, with an income risk modeled with a fuzzy number and a
background risk modeled with a random variable. Section 5 deals with mixed models of optimal saving of type
II, with an income risk modeled with a random variable and a background risk modeled with a fuzzy number.
Three notions of precautionary saving will be defined for both mixed models of type I and II.
The main results of the paper establish the necessary and sufficient conditions for the positivity of
each notion of precautionary saving, in terms of the consumers preferences and the features of risks. More
specifically, the existence of each notion of precautionary saving depends on: the signs of third-order partial
derivatives of the utility function and the possibilistic and probabilistic variances associated with the mixed
vectors.
1. Literature Review
Saving under uncertainty is a topic introduced in economic research by Hayne E. Leland (1968) and
Agmar Sandmo (1970). These papers investigate the way in which the presence of risk modifies the optimal
amount of saving. The variation experienced by the optimal saving after adding risk elements is measured in
Leland (1968) and Sandmo (1970) by the concept of precautionary saving. A central result of those papers
asserts that a positive third-order derivative of a consumers utility function is a necessary and sufficient
condition for an increase in saving as the effect of risk. Miles S. Kimball (1990) defines the notion of
prudence as the sensitivity of optimal saving to the magnitude of risk, and introduces the index of absolute
prudence as a measure of this sensitivity. He proves that the theory of the absolute prudence index is
isomorphic to the Arrow-Pratt index theory on risk aversion (see e.g. Eeckhoudt, Gollier and Schlesinger
(2005)). Furthermore, Louis Eeckhoudt and Harris Schlesinger (2008) show how higher order risk changes
affect the demand for saving.
All the papers above study the optimal saving decision in the presence of a unidimensional risk. In
particular, the paper by Christopher D. Carroll and Miles S. Kimball (2008) contains a survey of the literature
on univariate precautionary saving. On the other hand, there are also economic and financial situations that
require several risk parameters. Therefore, some economic models consider, apart from the primary (income)
risk, a second kind of risk (exogenous and unhedgeable) usually called background risk (see e.g. Doherty
and Schlesinger (1983), Gollier and Pratt (1996), Pratt (1988)).
Then it was natural to raise the issue of the way in which the existence of several risk parameters
affects the optimal amount of saving. Courbage and Rey (2007), Louis Eeckhoudt, Beatrice Rey and Harris

Schlesinger (2007), Mario Menegatti (2009), Diego Nocetti and William T. Smith (2011) investigate the effect
of the primary risk and the background risk on the optimal saving decision, considering the case of a consumer
with a bidimensional utility function. Specifically, in the papers by Courbage and Rey (2007) and Menegatti
(2009), they established necessary and sufficient conditions for a positive precautionary saving to appear in a
bivariate context. In the particular case of Menegatti (2009), he defined two notions of precautionary saving:
the first one explores the impact of a small income risk on saving, relative to the optimal saving under
certainty; the second one explores the impact of both sources of risk (primary and background) on saving,
relative also to the optimal saving under certainty. Moreover, the paper by Nocetti and Smith (2011) develops a
precautionary saving model with an infinite horizon. Finally, the paper by Elyes Jouini, Clotilde Napp and
Diego Nocetti (2013) develops a matrix-measure concept of prudence.
2. Preliminaries
2.1. Preliminaries on Fuzzy Numbers
Let us now introduce some preliminary concepts on fuzzy theory. They will be useful to set up the
consumers optimization problem, in which one of the sources of risk will be represented by a fuzzy number.
Let
. A fuzzy set

be a nonempty set of states. A fuzzy subset of

is normal if

is a function

for some

. The support of

is defined by

. Next assume that

,

the

level

set

of

is

is the topological closure of

The fuzzy set

A fuzzy subset

is fuzzy convex if
of

defined

by

).

is a convex subset of

for all

is called a fuzzy number if it is normal, fuzzy convex, continuous and with

bounded support. If
and

. For

are fuzzy numbers and

then the fuzzy numbers

are defined by

.
A nonnegative and monotone increasing function

is a

weighting

function if
We fix a weighting function
for
(interpreted as a utility function).

and a fuzzy number

. Let

such that
be a continuous function

The possibilistic expected utility

is defined by:

(1)

If

is the identity function then from (1) one obtains the

weighted possibilistic expected value

(Robert Fullr and Peter Majlender (2003)):

(2)
If
then
possibilistic variance (Wei-Guo Zhang and Ying-Luo Wang (2007)):

one

obtains

the

weighted

(3)

When

for

and

are the possibilistic mean value and the possibilistic variance of Christer Carlsson and Robert
Fullr (2001).
2.2. Mixed Expected Utilities
The concept of mixed expected utility was introduced by Georgescu and Kinnunen (2011) in order to
build a model of risk aversion with mixed parameters: some of them were described by fuzzy numbers and
others by random variables. This same notion has been used by Irina Georgescu (2012b) to study mixed
investment models in the presence of background risk.
In this section we will review this definition of mixed expected utility and some of its properties. For
clarity purposes, we will deal only with the bidimensional case. Then a mixed vector will have the form

( A , X ) , where
only consider the case
Let

is a fuzzy number and

a random variable. Without loss of generality, we will

(A, X) .

be a random variable w.r.t. a probability space

its expected value and by

its variance. If

function, then

is a random variable and

(probabilistic) expected utility of

w.r.t.

. Let

is a continuous
is the

We are going to choose a fixed weighting function

function

. We will denote by

and a bidimensional, continuous utility

be our particular mixed vector. Assume that the level sets

of

have the form

will

be

the

. For any

random

for any

variable

defined

by

Let us now define our concept of mixed expected utility, which will be subject to maximization in our
approach to optimal saving.
Definition 2.1 (Irina Georgescu, 2012a, Georgescu and Kinnunen 2011)
The mixed expected utility
vector

associated with

and the mixed

is defined by
(1)

Remark 2.2
(i) If the fuzzy number

is the constant

, then

.
(ii)

If

the

random

variable

is

the

constant

then

The following two propositions are essential to prove the main theorems to be discussed in the
following sections:

Proposition 2.3 (Georgescu, 2012a, Georgescu and Kinnunen 2011)

Let

be two bidimensional utility functions and

then
.

Proposition 2.4 (Georgescu, 2012a, Georgescu and Kinnunen 2011)

. If

If the utility function

has the form

then

3. A Probabilistic Approach to Optimal Saving

The optimal saving models presented in Courbage and Rey (2007) and Menegatti (2009) consider the
existence of two types of risk, background risk and income risk, both of them being mathematically
represented by random variables. In this section we will present the general features of these models as a
reference to start building our main models in the following sections.
These two-period models proposed by Courbage and Rey (2007) and Menegatti (2009) are
characterized by the following data:
and
the variable
for period

are consumers utility functions for period

represents the income, and

, the variables

for period

and

, resp.

is a nonfinancial variable.

have the certain values

and

, there is un uncertain income (described by the random variable

background risk (described by the random variable

We denote by

) and a

).

and

mentions the following four possible situations for the variables

. In Menegatti (2009) the author

and

(a)

(simultaneous presence of income risk and background risk)

(b)

(income risk and no background risk)

(c)

(background risk and no income risk)

(d)

(no uncertainty)

Consider now the following expected lifetime utilities corresponding to the situations (a), (c) and (d),
respectively:
(1)
(2)
(3)
where

is the level of saving. According to Menegatti (2009), the optimization problem can be

formulated as follows:

(4)
(5)
(6)
with the optimal solutions

and

.
The differences
,
are called precautionary saving and twosource
precautionary saving, respectively, in Menegatti (2009). The author finally presents some necessary and
sufficient conditions such that
and
generalizes some results previously obtained by Courbage and Rey (2007).

, which

We intend to offer in this paper an alternative setting to Menegatti (2009)s by allowing for the
possibility that either the income risk or the background risk are represented formally by a fuzzy number.
Therefore, the structure of the following sections follows a parallel line to Menegatti (2009)s, although instead
of his random vector

(Y , X) we will consider mixed vectors of the type ( A , X or (Y , B) .

4. Mixed Models of Type

The mixed models of this section are based on the hypothesis that the income risk is described by a
fuzzy number

and the background risk is described by a random variable

notation introduced in the previous section. A fuzzy number

random variable

corresponds to the variable

we have a mixed vector

We will fix a weighting function

. We will preserve the

corresponds to the variable

and the

. Thus instead of Menegatti (2009)s random vector

.

and denote

and

. In this case the situations (a)-(d) of Section 3 become

,
,
,
,
In this section we will study how the optimal saving changes as we follow the routes
,
and
. The first two are analogous to
the cases studied in Menegatti (2009) for the probabilistic models. We will define three notions of

"precautionary saving" and will establish necessary and sufficient conditions for the positivity of these
indicators.
Assume that the bidimensional utility functions

and

are strictly increasing with respect to each

component, strictly concave and three times continuously differentiable. We denote by

(resp.

) the first, the second and the third partial derivatives of

(resp.

).

Next we will use, as in Menegatti (2009), the following Taylor approximation:

v1 ( y s, x) v1 (a s, x ) v11 ( a s, x )( y a) v12 (a s, x )( x x )
(1)

Using the notion of mixed expected utility from Subsection 2.2, we introduce the following expected
lifetime utilities:
(2)
(3)

(4)
(5)
,
situation

are the analogues of

of Section 3, and

from above. By taking into account our formula (1) from Subsection 2.1:

(6)

If we differentiate, from (6) one obtains:

which, by formula (1) of Subsection 2.1, can be written as

comes from the

(7)

If we differentiate, from (3)(5) it follows that:

(8)

(9)
(10)

Proposition 4.1 The functions

and

Proof. Deriving the above expression of

are strictly concave.

it follows:

Since
and
one obtains
can prove that the other three functions are strictly concave. This ends the proof.

. Similarly one

We consider now the following optimization problems:

(11)
(12)
(13)
(14)
in

which
and

,
are optimal solutions.

By Proposition 4.1, the four optimal solutions are given by:

and

.
Taking into account (7)(10), the optimal conditions are written:
(15)

(16)
(17)

(18)
Following the line of Menegatti (2009), we will introduce three notions of "mixed precautionary
saving":

corresponds to precautionary saving from Menegatti (2009)

modification of the optimal saving when moving from
, i.e. by adding the income risk
difference

to
in the presence of the background risk

. The

expresses the modification of the optimal saving by moving from the certainty

situation
the income risk

and measures the

to the situation
and the background risk

, i. e. by adding

. Finally,

measures the modification of the

optimal saving by moving from

by adding the background risk

to

, i.e.

in the presence of the income risk

Next we intend to give necessary and sufficient conditions for the positivity of the three indicators.

Proposition 4.2 Let

be a mixed vector with

and

. The following inequalities are equivalent:

(i)
(ii)

;
.

Proof. Using the approximation formula (1) and Proposition 2.3, by applying the mixed expected
utility operator one obtains:

One

notices

that

and
. Also

, thus the

previous relation becomes:

(19)

A similar computation shows that

(20)

Taking into account that

therefore

By

is strictly concave it follows that

iff

is strictly decreasing,

. By (8) and (15):

approximating

and

with the values given by the formulas (20) and (19), from the previous relation one
obtains:
(21)
But

, thus
.

This ends the proof.

iff

iff

The property (i) of the previous proposition says that the effect of adding the income risk
presence of background risk

is the increase in the optimal saving. In particular, from Proposition 4.2 it

follows that if
income risk

in the

then
and for any background risk

for any
.

Next we study the change of the optimal saving on the route

Proposition 4.3 Let

be a mixed vector with

and

. The following are equivalent:

(i)
(ii)
.

Proof.

Taking

into

account

that

By the formulas (9) and (13) one has the equalities:

Formula (11) gives the following approximation:

thus

The following equivalences follow

iff

is

strictly

concave,

iff

iff

This ends the proof.

Condition (i) of Proposition 4.3 says that the effect of adding the income risk
risk

and the background

is the increase in the optimal saving. In particular, from Proposition 4.3 it follows that if
and

then for any mixed vector

we have

Corollary 4.4 Assume that

is a mixed vector and

. If

then

, where

and

Proof. The proof follows from Propositions 4.2 and 4.3, taking into account that
and

This ends the proof.

Finally consider now the change of the optimal saving on the route

Proposition 4.5 Let

be a mixed vector with

and

. The following are equivalent:

(i)
(ii)

;
.

Proof. Using the approximation formula (1) and applying Proposition 2.3 it follows:
(22)

By (10) and (15):

Replacing

and
with their approximate values obtained by applying (22) and (19) one obtains:

Since

is

strictly

concave,

iff

iff

This ends the proof.

Condition (i) of the previous proposition says that adding the background risk
the income risk

in the presence of

leads to an increase in the optimal saving.

Corollary 4.6 If

and

then

Proof. Propositions 4.2, 4.3 and 4.5 are applied.

In the next example we will show that there exist mixed vectors
and

and the utility functions

with
,

for which condition

does not imply

s1* ( A, X ) s1 (a, x ) 0

Example 4.7 Let

number with
uniform repartition on

be two real numbers such that

,
. It is known that

. Let

for any

and
and

be a fuzzy
the

. A simple calculation shows that

and

. Then
(23)

Assume that the utility function

has the form:

with
,

.
One notices that

;
.
Then from (23) it follows:

(24)

One notices

iff

iff
iff
From (24) and these equivalences it follows:
(25)
iff
Replacing in (25)
taking into account Proposition 4.3 one obtains:

and

and

(26)

iff

For

from (26) it follows

iff

iff

.
Then,

if

we

will

have

. On the other hand,

thus by Proposition 4.2,

In particular, the above example shows that the converse of Corollary 4.6 is not true.

5. Mixed Models of Type

The mixed models of this section assume that the income risk is represented by a random variable
and the background risk by a fuzzy number
utility functions

and

. We keep the hypotheses of Section 4 on the bidimensional

We fix a weighting function

. We denote
Similarly as in previous sections, we consider the following cases:

and

We analyze the way the optimal saving changes on the following three routes:
,

and

. For each of these three cases we

will introduce a notion of "precautionary saving" and we will prove necessary and sufficient conditions for the
positivity of these three indicators.

( a 2 ) (d 2 )
Corresponding to the cases

we introduce four expected lifetime utilities:

(1)
(2)

(3)
(4)
Deriving (1)(4) it follows
(5)

(6)

(7)
(8)

Similarly to the previous section, it is proved that

concave functions. We form the four maximization problems:

and

are strictly

(9)
(10)
(111)
(12)
in

which

and

are the optimal solutions.

By (5)(8) the optimal conditions

and

,
will be written:

(13)
(14)
(15)
(16)
We consider the following notions of precautionary saving:
. The precautionary saving
the route

and

measures the change of the optimal saving on

on the route

on the route

and

The following three propositions offer necessary and sufficient conditions for the positivity of the
three indicators.

Proposition 5.1 Let

be a mixed vector with

and

. The following are equivalent:

(i)
(ii)

Sketch of the proof. Applying the approximation formula (1) of Section 4 and Propositions 2.3, 2.4
similarly to the proof of Proposition 4.5 one reaches:
(17)

(18)

By (6) and (13) one obtains:

(19)

Replacing

in

(19)

and

with their approximate values from (18) and (19) we find the solution:
(20)
With the same type of argument as in the proof of Proposition 4.2,

iff

Proposition 5.2 Let

be a mixed vector with

and

. The following are equivalent:

(i)
(ii)
.

Proposition 5.3 Let

be a mixed vector with

and

. The following are equivalent:

(i)
(ii)

Corollary 5.4 Let

be a mixed vector with

. If

and
and

then

The proofs of Propositions 5.2 and 5.3 and Corollary 5.4 are similar to those of Propositions 4.3 and
4.5 and Corollary 4.6.
Similarly to Section 4 one proves that the converse of Corollary 5.4 is not true.

The positivity conditions of the three notions of precautionary saving

and

express the fact that on the routes

and

,
,

the optimal saving increases. The above results characterize

these conditions in terms of the third partial derivatives of

6. Conclusions
The optimal saving models of this paper combine methods of probability and possibility theory. For
mixed twoperiod models the way the optimal saving changes is studied in two cases:
(I) the income risk is a fuzzy number and the background risk is a random variable.
(II) the income risk is a random variable and the background risk is a fuzzy number.
For each of the two types of models three notions of mixed precautionary saving have been
introduced. These indicators measure the variation of the optimal saving as a result of adding income risk in
the presence of background risk, adding background risk in the presence of income risk or simultaneously
adding income risk and background risk to a certain situation. The main results of the paper characterize the
positivity of the three indicators, which indicates that by the mentioned modifications the level of optimal
saving increases.
Our results indicate that, also when the consumers environment is fuzzy, there are potentially reasons
for extra saving relative to certainty. And we have characterized the conditions for such precautionary saving.
That characterization could be useful to predict the behavior of aggregate saving in response to well defined
risks affecting a countrys population. A possible extension of our paper could be a numerical comparison of
the differences in saving predicted by our model and Menegatti (2009)s.
The mixed models of the paper follow a parallel line with the probabilistic model of Menegatti (2009),
where the background risk and the income risk are random variables. It remains to study a purely possibilistic
optimal saving model, in which both the income risk and the background risk are fuzzy numbers.

References

Carlsson, Christer, and Robert Fullr. 2001. On Possibilistic Mean Value and Variance of Fuzzy
Numbers. Fuzzy Sets and Systems, 122: 315326.

Carlsson, Christer, and Robert Fullr. 2011. Possibility for Decision. Springer.

Carroll, Christopher D., and Miles S. Kimball. 2008. Precautionary Saving and Precautionary
Wealth. In The New Palgrave Dictionary of Economics, Ed. Steven N. Durlauf and Lawrence E. Blume.
Palgrave Macmillan. Second Edition.
Chamn Marcos, Kai Liu, and Eswar Prasad. (2013). "Income Uncertainty and Household Savings
in China. Journal of Development Economics 105:164177.

Courbage, Christophe, and Batrice Rey. (2007). Precautionary Saving in the Presence of Other
Risks. Economic Theory, 32: 417424.

Doherty, Neil A., and Harris Schlesinger. (1983). Optimal Insurance in Incomplete Markets.
Journal of Political Economy, 91: 10451054

Dubois, Didier, and Henry Prade. (1988). Possibility Theory . New York: Plenum Press.

Eeckhoudt, Louis, Christian Gollier, and Harris Schlesinger. (2005). Economic and Financial
Decisions under Risk. Princeton University Press.

Eeckhoudt, Louis, and Harris Schlesinger. (2008). Changes in Risk and the Demand for Saving.
Journal of Monetary Economics, 55: 1329-1336.

Eeckhoudt, Louis, Batrice Rey, and Harris Schlesinger. (2007). A Good Sign for Multivariate
Risk Taking. Management Science, 53: 114-124.

Fullr, Robert, and Peter Majlender. (2003). On Weighted Possibilistic Mean and Variance of
Fuzzy Numbers. Fuzzy Sets and Systems, 136: 363374.

Georgescu, Irina. (2012a). Possibility Theory and the Risk. Springer.

Georgescu, Irina. (2012b). Combining Probabilistic and Possibilistic Aspects of Background Risk.
The 13th IEEE International Symposium on Computational Intelligence and Informatics CINTI 2012,
Budapest, Hungary, 2022 November, pp. 225229.

Georgescu Irina, and Jani Kinnunen. (2011). Multidimensional Risk Aversion with Mixed
Parameters. 6th IEEE International Symposium on Applied Computational Intelligence and Informatics SACI
2011, May 19-21, Timisoara, Romania, pp. 6368.

Gollier, Christian, and John W. Pratt. (1996). Risk Vulnerability and the Tempering Effect of
Background Risk. Econometrica, 64: 11091124.

Jouini, Elys, Clotilde Napp, and

Journal of Economic Theory.

Diego Nocetti. Forthcoming. On Multivariate Prudence.

Kimball, Miles S. (1990). Precautionary Saving in the Small and in the Large. Econometrica, 58:
5373.

Leland, Hayne E. (1968). Saving and Uncertainty: the Precautionary Demand for Saving.
Quarterly Journal of Economics, 82: 465473.

Menegatti, Mario. (2009). Optimal Saving in the Presence of Two Risks. Journal of Economics,
96: 277288.

Nocetti, Diego, and William T. Smith. (2011). Price Uncertainty, Saving and Welfare. Journal of
Economic Dynamics and Control, 35: 1139-1149.

Pratt, John W. (1988). Aversion to a Risk in the Presence of Other Risks. Journal of Risk and
Uncertainty, 1: 395413.

Sandmo, Agmar. (1970). The Effect of Uncertainty on Saving Decision. Review of Economic
Studies, 37: 353360.

Zadeh, Lotfi A. (1978). Fuzzy Sets as a Basis for a Theory of Possibility. Fuzzy Sets and Systems,
1: 328.

Zhang, Wei-Guo, and Ying-Luo Wang. (2007). A Comparative Study of Possibilistic Variances and
Covariances of Fuzzy Numbers. Fundamenta Informaticae, 79: 257263.