Financial Advice and Stock Market Participation

Transcription

1 Financial Advice and Stock Market Participation Dimitris Georgarakos Roman Inderst University of Frankfurt and CFS University of Frankfurt (IMFS) and Imperial College London International Pension Workshop Amsterdam, January 26-28, 2011.

2 Motivation Professional financial advice is pervasive In the US: More than three-quarters of investors who hold equities outside employer-sponsored plans in 2005 own equities purchased through advisers. Equity Ownership in America (2005) In Europe around 80% declare that they expect financial institutions to give them advice. 1

3 Motivation Complexity of new financial products / Increased household responsibility to save for retirement. Recent financial crisis has shattered confidence in financial institutions / Long-term repercussions on households willingness to participate in financial markets? US: We must act now to restore confidence in the integrity of our financial system (Financial Regulatory Reform, US Department of Treasury, June 2009). EU: series of projects to enhance customers confidence in the retail financial sector (coordinated by Directorate General for Consumer Protection - SANCO). 2

4 This Paper: Aim Use a model and empirical analysis to introduce financial advice in households decision to hold risky financial assets. Shed light on the role of households perceptions about: - own financial capability - legal rights as consumers of financial services - trust in financial advice Add to the growing literature on household finance (Guiso, Haliassos, and Jappelli, 2001; Campbell, 2006; Tufano, 2009) 3

5 This Paper: Main Findings (preview) Trust in financial advice matters for stock market participation only when households own financial capability is low. For households with high financial capability it is their perception of legal rights as consumers of financial services that affects stockholding. Households perceptions of legal protection differ in an economically significant way even though, after controlling for a range of socioeconomic variables, their objective circumstances are seemingly comparable. 4

6 Model Decision process (two parts): - Participate in risky assets (e.g. through opening a retirement account)? - If so, which assets to choose (e.g. which retirement plan)? (for simplicity, choice between two plans: a and b) No participation: U 0 5

7 Model Participation -> choice of particular plan/ strategy - Make an informed choice: capable investor / rely on advisor - Investor/Advisor not able to beat the market - Instead, finding the right fit (i.e. assets that give the highest benefits, given individual preferences/ needs; Inderst, Ottaviani, AER, 2009). -> Abstracting from risk aversion: U h >U l (no fit ) 6

8 Model - In addition, perceived risk to be defrauded (Guiso, Sapienza, Zingales, JF, 2008). by those who ultimately undertake the security transactions, such as brokers and dealers / by those who manage the respective accounts, such as investment funds / or ultimately by companies owners and managers - Risk of fraud depends on the legal environment / perceived risk thus depends on the households perception of the legal environment Ex-ante each equally likely to be best fit : π 0 = Pr(a = fit ) = 1/2 With posterior π (based on own information or advisor s recommendation) choose a when π π* 7

9 Self-reliant Investor: finding the right fit Investor s posterior on fit : π I Investor s financial capability: ρ I Optimal threshold: π* = 1/2 Ex ante utility of investing in risky assets incorporates the risk of being defrauded and the risk of not making the most suitable choice. No informed choice:, where Fully informed choice: 8

10 Investor relying on Advice: finding the right fit Advisor s posterior on fit : π A Can generate information with precision ρ A May receive information commission z 0 on either product. Puts weight on customer utility γ. Weight 1- γ on commission. If commission on a, then recommends a when: Cut-off rule: Customer s beliefs = trusting advice: Lower w 9

11 Comparative Results Corollary 1 (Hypothesis 1): Trust in advice only affects stockownership when the household's own capability is low. Instead, when financial capability is high, only perceived legal protection affects the participation decision. 10

12 Comparative Results: complementarily Corollary 2 (Hypothesis 2): Less capable households: more likely to invest in risky assets when both perceive consumer protection to be sufficiently high and sufficiently trust advice 11

13 Comparative Results: complementarily Corollary 3 (Hypothesis 3). More capable households or households that do not trust advice: invest in risky assets when they both perceive consumer protection to be sufficiently high and are sufficiently capable. 12

14 Data Eurobarometer 60.2 (End of 2003): Employment and Social Policies, Financial Services, Harmful Internet Content, and Product Safety / 15 EU Countries Key questions: I usually trust the advice given by financial institutions. My rights as a consumer are adequately protected in relation to financial services. Financial capability: - Education: college/ non-college - Perceive financial matters to be complicated - Gender Stock Ownership (direct and indirect) 13

15 Descriptive Statistics To tal Sample 14 Colle ge Graduate s Le s s than Colle ge Educatio n Total Stocks Trust in financial advice Consumer rights are protected A ge Male Couple Single D ivorced Children Self Employed Retired U nemployed H igh School College Median Income band Number of observations

16 Empirical specification Controls for: gender; marital status(couple, divorced, single, widowed); children; employment status(employed, self-employed, retired, unemployed); education(less than high school, high school, college degree), income(12-income bands); country dummies. 15

17 Baseline results Total Sample College Graduates Less than College Education Marg. Eff. z-stat Marg. Eff. z-stat Marg. Eff. z-stat Trust in financial advice ** *** Consumer rights are protected *** ***

18 Marginal Effects for Different HH Segments All Marg. Eff. z-stat Marg. Eff. z-stat Marg. Eff. z-stat Trust in financial advice All Marg. Eff. z-stat Marg. Eff. z-stat Marg. Eff. z-stat Consumer rights are protected *** *** ** All Marg. Eff. z-stat Marg. Eff. z-stat Marg. Eff. z-stat Trust in financial advice *** *** All A. College Graduates Consumer Rights Protected subsample Trust in Advice subsample Consumer Rights Not Protected subsample No Trust in Advice subsample B. Less than College Education Consumer Rights Protected Consumer Rights Not subsample Protected subsample Trust in Advice subsample No Trust in Advice subsample Marg. Eff. z-stat Marg. Eff. z-stat Marg. Eff. z-stat Consumer rights are protected **

19 Marginal Effects for Different HH Segments All Marg. Eff. z-stat Marg. Eff. z-s tat Marg. Eff. z-stat Trus t in financial advice All Marg. Eff. z-stat Marg. Eff. z-s tat Marg. Eff. z-stat Cons ume r rights are prote cte d *** *** ** All Marg. Eff. z-stat Marg. Eff. z-s tat Marg. Eff. z-stat Trus t in financial advice *** *** All A. College Graduates Consumer Rights Protected subsample Trust in Advice subsample Consumer Rights Not Protected subsample No Trust in Advice subsample B. Less than College Education Consumer Rights Protected Consumer Rights Not subsample Protected subsample Trust in Advice subsample No Trust in Advice subsample Marg. Eff. z-stat Marg. Eff. z-s tat Marg. Eff. z-stat Cons ume r rights are prote cte d **

20 HH not expecting Advice I expect financial institutions to give me advice. (~ 80%) (Those who do not expect to receive advice from financial institutions arguably should not rely on it.) Bivariate Probit. Marginal effects on the probability to invest in stocks conditional on reporting that advice is not expected Consumer Rights Protected subsample Consumer Rights Not Protected subsample Marg. Eff. z-stat Marg. Eff. z-stat College *** *** College Graduates subsample Less than College Education subsample Marg. Eff. z-stat Marg. Eff. z-stat Consumer rights are protected ***

21 Tests for Endogeneity Trust in FA: When you are about to buy [detergents/ toiletries] do you always, some times, rarely or never pay attention to the following on the packaging: [instructions on how much to use] and [pictures/ text on how to use it]. households may choose not to pay attention to the recommended amount of use because they think that the company recommends a higher dose of the product than what they consider to be sufficient. CP: do you think that it is very easy, fairly easy, difficult or very difficult to win a dispute with an insurance company. 20

22 Tests for Endogeneity Two-step procedure of Rivers and Vuong (1988), (see also Wooldridge, 2002). Statistically valid instruments. Fail to reject the null of no endogeneity at any conventional level of statistical significance. Rivers Vuong test for exogeneity H0: no endogeneity F-test F-test F-test Chi 2 Model estimated in: p-value p-value p-value p-value Total Sample Auxiliary regressions eq1: trust_advice eq2: protect_rights eq3: trust_protect *** *** *** College Graduates *** *** *** Less than College Education *** *** ***

23 Robustness Van Rooij, Lusardi, and Alessie (2007): 37%(M) and 12%(F) in the advanced literacy group. Barber and Odean (2001): male overconfidence affects genderspecific trading behavior. Think Finances: Not Think Finances: Complicated Complicated Male Respondent Female Respondent Marg. Eff. z-stat Marg. Eff. z-stat Marg. Eff. z-stat Marg. Eff. z-stat Trust in financial advice ** ** Consumer rights are protected *** ***

24 Robustness Think Finances Complicated No Yes College Education Yes No Marg. Eff. z-stat Marg. Eff. z-stat Trust in financial advice ** Consumer rights are protected *** Trust in financial advice ** Consumer rights are protected

25 Attitudes towards risk whether safety of detergent or toiletry products is not important or very important. Importance of features: It is safe. It tells me how to protect myself. It tells me what to do in case of accident, injuries or health problems. There are clear symbols or pictures of risks and dangers. 24

26 The Role of Perceptions Trust in Advice and Perceived Consumer Protection by Country FI SE UK IE DK DE NL BE LU FR AT IT ES PT GR trust_advinst consr_protect 25

27 The Role of Perceptions Generalized Trust (WVS) Generally speaking, would you say that most people can be trusted or that you need to be very careful in dealing with people? 1. Most people can be trusted 0. Can't be too careful 26

28 Regressions with regional controls I. Probit model II. Probit model III. Linear regression Marg. Eff. z-stat Marg. Eff. z-stat Coeff. Est. t-stat Trust in financial advice ** ** Consumer rights are protected *** *** Trust other people (regional average) * ** Socio-economic characteristics Regional Dummies Country Dummies yes yes yes yes yes yes Number of Observations

29 Antidirector Rights Index (ADRI) La Porta, Lopes de Silvana, Shleifer, and Vishny (JPE, 1998): The index is formed by adding 1 when: (1) the country allows shareholders to mail their proxy vote to the firm; (2) shareholders are not required to deposit their shares prior to the General Shareholders Meeting; (3) cumulative voting or proportional representation of minorities in the board of directors is allowed; (4) an oppressed minorities mechanism is in place; (5) the minimum percentage of share capital that entitles a shareholder to call for an Extraordinary Shareholders Meeting is less than or equal to 10 percent (the sample median); or (6) shareholders have preemptive rights that can only be waved by a shareholders vote. 28

30 ADRI vs. Trust in Advice and Perceived Consumer Protection by Country ADRI GR IT SEFR DE ES PT NL UK ATDK Consumer Rights Protected BE IE FI ADRI GR IT ES PT UK FR SE NL DE BE IE FI ATDK Trust Financial Advice 29

31 Conclusions Households who need to rely on advice: trust in advice becomes a key determinant for stockholding. For more educated households (or those who do not perceive financial decisions to be complicated): perception of how well their rights as consumers of financial services are protected matters. Controlling for objective differences in the legal environment and in legal enforcement households have different perceptions of their rights 30

32 Conclusions For households not relying on advice (with higher capability or who do not trust advice), high financial capability and a high perception of their legal rights jointly produce a high willingness to hold risky assets. Households with low financial capability must put trust both in advice and in legal protection to generate a large effect on their propensity to hold risky assets. 31

33 Conclusions To foster stockholding among households with low own financial capability, trust in advice is a key prerequisite. The observed economically significant differences in households perception of legal protection could suggest a further avenue to improve efficiency. 32