Financial Risk Propensity, Business Cycles and Perceived Risk Exposure
| Date | 01 February 2018 |
| Author | Raffaele Miniaci,Alessandro Bucciol |
| Published date | 01 February 2018 |
| DOI | http://doi.org/10.1111/obes.12193 |
160
©2017 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 80, 1 (2018) 0305–9049
doi: 10.1111/obes.12193
Financial Risk Propensity, Business Cycles and
Perceived Risk Exposure*
Alessandro Bucciol† and Raffaele Miniaci‡
†Department of Economics, University of Verona, Via Cantarane 24 I-37129, Verona, Italy
(e-mail: alessandro.bucciol@univr.it)
‡Department of Economics and Management, University of Brescia, Via San Faustino 74/b
I-25122, Brescia, Italy (e-mail: raffaele.miniaci@unibs.it)
Abstract
Weanalyse whether individual financial risk propensity changes overtime with background
financial conditions, as well as personal and subjective portfolio risk exposure. We elicit
risk propensity from six different self-assessed facets collected in a long panel data set from
the DNB Household Survey, annually covering the period 1995–2015. Risk propensity is
generally higher during periods of economic growth and lower during periods of recession,
but is untrended when elicited, using questions referring to safe investments. Our risk
propensity measure is also higher following positive stock market returns or subjectively
large risk exposure in own past investments.
I. Introduction
In the aftermath of the Great Recession, there is evidence that people changed their be-
haviour and lifestyle: for instance, in the UK, theyspent less money on holidays, shared cars
and used public transportation more frequently, and they ate cheaper, lower quality food
(Crossley, Low and O’Dea, 2013). The 2007–09 financial crisis affected expectations, risk
perceptions and attitudes (e.g. Hoffman and Post, 2015). These changes in habits and risk
attitudes can have potentially relevant impacts on the asset price formation process (e.g.
Campbell and Cochrane, 1999) and the effectiveness of economic policies (De Paoli and
Zabczyk, 2013). In this work, we study empirically the link between the macroeconomic
environment, personal experience with portfolio risks and returns and household financial
risk propensity, i.e. the willingness to bear risk, in a time interval that includes recessions
as well as expansion periods.
JEL Classification numbers: D81, G11, D14.
*We thank Scott Findley, Dimitris Georgarakos, Arvid Hoffmann,Alessandro Sembenelli and the seminar par-
ticipants at the University of Copenhagen, the University of Alicante, the University ‘Ca’ Foscari’of Venice, the
2012 CeRP conference at Collegio Carlo Alberto in Turin, the 2014 Netspar International Pension Workshop in
Venice, the 2014 NIBS workshop at the University of Nottingham, the 2015 ICEEE Congress in Salerno and the
2015 IAREP-SABE Conference in Sibiu. In this paper, we use data from the DNB Household Survey. The usual
disclaimers apply.
Risk propensity, cycle and risk exposure 161
There is growing micro-based evidence that market factors havean impact on investors’
risk attitudes in household finance. Malmendier and Nagel (2011) find that cohorts who
have experienced low stock market returns in the past report lower willingness to bear
financial risk, are less likely to participate in the stock market and, if they participate,
invest a lower fraction of their wealth in stocks. This suggests that the recent shocks
to financial market returns might have a long-lasting impact on investors’ behaviour and
persistently reduce future stock market participation. Bucciol and Miniaci (2015) document
that portfolio risk varies together with the business cycle and Guiso, Sapienza and Zingales
(2013) also find evidence of a time-varying pattern for subjectiveand hypothetical measures
of risk aversion.
There is evidence that experiencing consecutive investment losses reduces the subse-
quent willingness to take risk (for a review see Barberis, 2013). A possible psychological
explanation is that negative events promote negative emotions such as fear which, in turn,
may lead individuals to be more risk averse (Lerner and Keltner, 2001). Hoffman and Post
(2015) find, for clients of a brokerage firm, that return expectations and risk tolerance
are positively correlated with past returns. Other works find that investors have difficulty
learning from their experience (Barber and Odean, 2001) and, if they learn, the process is
slow (Gervais and Odean, 2001).
To frame our work in the context of standard consumption-investment modelling, con-
sider a household that at time tmaximizes the expected present discounted value of utility
flows, with the discount factor and Uinstantaneous utility,
Vt=Et[
∞
j=t
j−tU(Cj)] (1)
under the usual intertemporal budget constraint and the no-Ponzi scheme condition. The
coefficient of relative risk aversion with respect to the beginning-of-period assets Atis
Rt=−At×(V
t/V
t) (e.g. Swanson, 2012).The (unobserved) value of Rtvaries over time and
across households because of heterogeneity in the curvature of U, consumption, holding
of financial assets, housing (e.g. Zanetti, 2014), habits (e.g. Boldrin, Christiano and Fisher,
1997), or expectations. We posit that changes in risk propensity, that is, the households’
willingness to take financial risk, as elicited by survey questions, mirror changes in Rt.
1
The goal of our analysis is to study empirically the relation between risk propensity
and common and personal experience, using specific facets of risk propensity as well as
a composite indicator. Considering common and personal experience jointly may provide
indirect evidence about the appropriateness of an asset-pricing model with internal rather
than external habit formation (e.g. Chen and Ludvigson, 2009). Our hypothesis is that risk
propensity evolves in response to market conditions (common experience) and, once we
control for it, it also varies after observing the past performance of the personal portfolio.
In this regard, we distinguish between objective and subjective experience of financial
1There is no consensus on the denomination of the variables derived from survey questions similar to those used
in this paper. Some consider them as a measure of preferences, i.e. ‘risk aversion’(e.g. Guiso et al., 2013), other
as a measure of actual household choices, i.e. ‘risk-taking’ (e.g. Malmendier and Nagel, 2011). We instead take an
agnostic position and we adopt the more neutral term ‘risk propensity’,commonly used in behavioural sciences (e.g.
Josef et al., 2016), recognizing that it is difficult to disentangle preferences, actual choices, endowments, habits and
expectations from survey questions without further data and behavioural assumptions.
©2017 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial