Capturing Investor Attitudes Towards Risk. A New Approach
Luc Arrondel and André Masson from the French National Research Council, interviewed by Daniele Fano, Editor of Trends in Savings and Wealth and Head of Pioneer Investments Economic Research, Paris September 20, 2006
Daniele Fano: First of all, thank you very much for accepting to have this interview.
I found your approach broad and innovative.
You have worked extensively on the original data of the French Survey of Household Wealth. Unsatisfied with the abstract questions and the disappointing outcome of measures of preferences in similar surveys, you propose new ways for inferring investor attitudes towards risk and introduce new concepts. You also propose to correlate financial choice with other areas of individual choice.
For people like us who like to do field work all this is extremely interesting.
Let me start by the concept side of your conclusions.
You end up proposing one risk parameter, which is a mix of risk aversion, prudence, temperance and loss aversion. This "score" is an average indicator of the answers given by the interviewee to a series of questions concerning his/her attitude towards risk…
André Masson: … in different domains of life, for small risks and for big risks, losses and gains and so on ….
Daniele Fano: …one parameter but very rich in dimensions… and, as we look at the time dimension, you propose four parameters, first of all time preference over the life-cycle, secondly impatience (over the short run or due to time inconsistency), and finally altruism in two dimensions towards the family and towards society.
So far for the new concepts.
You also say that when you look at individuals you are not interested in numerical measures of risk aversion, but essentially at the ordinal aspects. From my reading there is even something more than ranking in your approach, I believe you imply there are discontinuities across individual attitudes. Did I interpret this well?
André Masson: Yes but first, let us you come back to the recent developments of theory : the more theory makes progress the more risk and time parameters you have to introduce, up to now more than ten preference parameters. If experimental research shows that expected utility does not work, we have to introduce in addition loss aversion and other parameters that can explain savings behavior. For example, if you take annuities, one of the reasons why people do not choose them is loss aversion, i.e. the probability you will have invested for nothing if you die too early. It does not mean that the parameters of standard (basic) theory do not play a role, but rather that additional explanatory factors are needed if you want to fully explain the diversity of accumulation patterns and portfolio choices. But what we would like to achieve is a balance between the high number of parameters that have been proposed and the need for simplicity. This is why we went first of all for rankings, allowing thus for a more limited number of parameters to be estimated.
Luc Arrondel: Concerning time dimension, the distinction between several preferences has been dictated by the data. I would like to tell the story of an economist from INSEE and an example of a far-sighted individual ; but he shows also (short-term) impatience, for example if a book does not satisfy him he will stop reading it after just a few pages, and this may be, in this case, rational, because he has many other things to do, his time is scarce. This shows the importance of disconnecting short-term impatience from longer-term time preference.
André Masson: There, it is not the Laibson concept of hyperbolic discounting, it is the fact that time is a cost in the short-term, something different from the perception of time preference over a longer time-horizon. We found however that one parameter on an ordinal basis is enough to capture a variety of factors concerning risk attitudes. Of course it is a “rich” parameter.
Luc Arrondel: It is interesting to refer to the paper by Barsky, Juster, Kimball and Shapiro on preference parameters and behavioral heterogeneity . They draw from psychologists’ findings and look whether you have a behavioral general disposition towards risk that can also explain your portfolio choice as well as other choices. Of course there are counter-examples, such as the world record man of car dragster speed-racing, a very dangerous sport. He is obviously a risk-lover in sports (his professional life), but does he invest in risky stocks? No, it happens that in financial behavior he takes no risks because he feels that he needs to balance a risky job with a very conservative portfolio.
Quick Access