Marcello Bofondi: Comments ppt 195.0 KB

The Luxemburg Wealth Study: Enhancing
Comparative Research on Household Finance
Comments on:
Cognitive Abilities and Portfolio Choice
by D. Christelis, T. Jappelli, and M. Padula
and
Financial Litteracy and Stock Market Participation
by M. van Rooij, A. Lusardi and R. Alessie
Marcello Bofondi
Bank of Italy
The main question
Can cognitive abilities / financial literacy help to explain the
so called “stock-holding” puzzle?
Why is it important to answer this question?
Cocco et al. (2005): welfare loss from non-participation in
stock markets can be substantial.
Motivation (1)
Recent financial innovation created a large set of quite
complicated, difficult to understand, financial assets.
Following Campbell (2006) one can speculate that:
“the existence of naïve household permits an equilibrium […]
in which confusing financial products generate a crosssubsidy from naïve to sophisticated households”
Motivation (2)
A “financial market participation channel” may therefore add
to other factors (such as the widening wage differential
between high-skilled and low-skilled workers) and contribute
to increase wealth inequality.
Policy implications
Given that both papers find a positive effect of cognitive
abilities and financial literacy on stock-holding the most
obvious policy response should be to increase numeracy and
literacy through education.
Unfortunately there is some evidence that formal economic
education may not help (Mundell, 2006).
Policy implications (1)
But we, as financial economists and researchers at central
banks, can work in other directions.
By enhancing competition in financial markets, proposing new
and more effective market regulation, consumer protection
policies and disclosure rules, we can help to lower all those
explicit and implicit transaction costs that prevent may
households to participate in financial markets (Campbell,
2006).
Some specific comments on
Cognitive Abilities and Portfolio Choice
by D. Christelis, T. Jappelli, and M. Padula
Cognitive abilities …
Very interesting paper with many results, not only about
cognitive abilities.
Besides the questions that measure cognitive abilities a second
unique characteristic of the SHARE is to have comparable
cross country data. Maybe some more effort could be made to
exploit this feature.
Cognitive abilities …
Explicit transaction costs (fees and commissions paid by
households to access the financial markets) vary across the
different countries covered by SHARE. Some information is
available about average mutual funds commissions (the TER
computed by Fitzrovia).
An individual with lower cognitive abilities has higher
implicit transaction costs. Therefore the impact of cognitive
abilities could be different depending on the level of explicit
transaction costs.
Cognitive abilities …
Cognitive abilities are influenced by health and education. In
order to measure the “net effect” of cognitive abilities the
authors control for these individual characteristics.
Then education is excluded from the controls in order to
measure the “gross” effect of cognition. Why not doing the
same for health?
Cognitive abilities …
The authors perform a number of interesting sample splits.
The first one is by marital status. It is not totally clear why.
The authors find that numeracy has a much stronger effect for
couples than for single and that memory is significant only for
coulples.
Very similar result are obtained when splitting he sample by
the degree of social interaction.
I would interpret the result of the two sample splitting in the
same way. After all marriage is a sort of social interaction …
Some specific comments on
Financial Litteracy and Stock Market Participation
by M. van Rooij, A. Lusardi and R. Alessie
Financial literacy …
Main contributions:
- literacy indexes
- noise in the responses
- disentangle the effect of basic and advanced financial
literacy controlling for possible endogeneity problems
Financial literacy …
The data set contains information on both self-assessed and
objective literacy.
The two measures are correlated, but one could exploit the
difference between the two, to verify if overconfidence has a
role in stock-holding.
Overconfidence may be even more harmful than nonparticipation!
Financial literacy …
The data set contains information on the different sources of
advice when making financial decisions.
This information could be exploited to verify if the role of
financial literacy is magnified or reduced by the different
sources of advice.
Some more comments on both papers …
Cognitive abilities and Financial literacy …
1- Sample split by wealth
2- What about the role of cognition and literacy on financial
liabilities?