Age And Other Factors Explaining Risky Asset Investments

Behavioral finance opposes traditional finance theories in the sense that individuals are not capable of carrying out the dynamic optimization problems that traditional finance talks about. However, one of the most fundamental concepts for investments and financial decision making for both schools of thoughts is the concept of risk. Risk appetite is a determinant which has been widely reported in prior literature related to investors’ appetite to hold risky assets, whereas it has been argued to affect not only portfolio composition, but the overall individual decision to hold risky assets. At first, as this study explores differences among generations and more specifically dissimilarities among young investors, it is important to identify the role that the age pertains within the literature of risky asset investing. Furthermore, researchers have long been aware of the dissimilarities in financial risk tolerance among individuals of various ages. However, apart from changes in risk appetite that have occurred due to different stages of lifecycle, literature also reports that individuals may change their financial risk attitudes in response to changes in general economic conditions. The combination of the above has been characterized by academics as “the age effect” and in fact refers to three different phenomena: aging effect, cohort effect, and period effect. Yao, Sharpe and Wang (2011) argue that what is attributed to an “age effect” might be due to: 1) depreciation of human capital and decreases of investment horizons and as people age (aging effect), 2) macroeconomic shocks and life experiences that influence different generations during their formative years and do not change with age (generation/cohort effect) or 3) socioeconomic environments and financial busts that have an impact on individuals of all ages over time (period effect). Palmer (1978) argues that the separation of the three effects is crucial when a research deals with aging process and differences among generations as the ignorance of this particular dissolution leads to erroneous conclusions. However, most of the studies that focus on financial risk attitudes and portfolio-choice have failed to separate these effects.

Generally, very few studies examine the impact of cohort effects on risky asset selection, whereas the majority of the research focuses mostly on aging effects assuming no generational effects. One of the earliest studies, Morin and Suarez (1983), investigated the effect of age on risky asset holdings using 1970 Canadian Survey of Consumer Finances data. Including age as a categorial variable in their multivariate analysis, they concluded that risk aversion increases with age. However, Palmer (1978) argues that in analyses which single cross-sectional data are used, it is erroneous to attribute results solely to aging process as in fact they might be partly or entirely the outcome of cohort differences. Likewise, Bertaut and Starr-McClue (2001) assume no cohort effects in their analysis of portfolio allocation patterns in the U.S. However, their probit regression model controlled for period and aging effects, including age and years of survey as independent variables, along with a set of numerous variables such as education, wealth, marital status, income and self-reported willingness to take financial risk. Combining repeated cross-sectional data from 1989, 1992, 1995 and 1998 Federal Reserve’s Survey of Consumer Finances, they also find that the age effects are significant in the decision to hold risky assets in the sense that older individuals are less likely to own risky assets compared to younger investors. However, the cohort effect was still entangled with the aging effect.

Apart from cross-sectional studies, experimental evidence and studies based on individual questionnaire responses also corroborate the argumentation above, as findings indicate that older investors tend to hold smaller fractions of equities and other higher-risk investments compared to younger ones. Additionally, longitudinal studies like Yao and Curl (2011), also find age effect consistent with previous studies.

On the other hand, some studies discover age to be positively related to risk tolerance and risky asset holdings. In particular, King and Leape (1987) find that the likelihood of equity ownership increases with age. However, evidence could not be distinguished between age and cohort effects as their estimations were solely based on a single cross-sectional dataset. Wang and Hanna (1997) calculated the ratio of risky assets to total wealth in order to investigate the effect of age on risk tolerance. Using the 1983-1989 Survey of Consumer Finance panel dataset, they found that age has a positive effect on risk tolerance. However, panel data cannot draw conclusions about cohort effects as they track the behavior of a the same group of people over a specific timespan. Moreover, Amerkis and Zeldes in 2004 studied the relationship between age and the fraction of wealth held in the stock market. Similarly, found no evidence that this fraction decreases with age.

In one of the very few studies that separated cohort from aging effect, Jianakoplos and Bernasek (2006) employed the 1989, 1995, and 2001 SCF data to investigate differences in financial risk-taking generated by different age groups and birth cohortd. Their results support that risk taking decreases with age, whereas they also reveal the existence of cohort effects underling that young generations are less willing to take financial risk compared to their older generations.

Furthermore, a small but influential body of research examines the influence of macroeconomic shocks and socioeconomic enviroments on the willingness of individuals to hold risky assets. Nevertheless, very few of these studies incorporate the parameter of aging and cohort effects into their analysis. Guiso, Sapienza and Zingales (2008) highlight that apart from the evaluation of the risk return trade-off given the available data and the latest experiences, involvement in risky activities requires also trust that the analyzed data is reliable and the overall system is fair. Furthermore, they argue that episodes related to vast macroeconomic shocks, like the collapse of Enron and Lehman Brothers, may not only alter the distribution of expected pay-offs but also the fundamental trust in the system that distributes those payoffs. After they examined 1,943 Dutch households as a part of the annual Dutch National Bank Household Survey, lack of trust found to be a crucial determinant in individuals’ risk appetite, in the sense that trusting individuals are significantly more likely to buy stocks and risky assets.

In their study Malmendier and Nagel in 2011, examined whether macroeconomic shocks shape risk attitude and affect peoples’ willingness to participate in stock markets, emphasizing at cohorts that were impacted by the Great Depression or the inflationary shocks of the 1970s. By analyzing cross-sectional data from 1960-2007 Survey of Consumer Finances waves, they found that individuals who experienced low stock market returns are less likely to participate in the stock market and more likely to keep a lower portion of their wealth in stocks compared to other groups of investors who have experienced higher returns. However, even though they suggested that more recent experiences have a stronger influence on young adults, their analysis did not emphasize on the estimation of particular cohort effects.

Likewise, Schuman and Corning in 2012 argued that national and world events that people experienced in early life tend to shape future attitudes and actions whereas such events are more likely to be remembered through the entire life. Added to that Nagel in 2012 highlighted that repercussions of macroeconomic shocks tend to affect young investors in a greater magnitude as they hold shorter life history, financial experience and knowledge. This is also known in literature as recency bias in which investors recall and weight excessively recent experiences or events as part of their decision-making process.

However, apart from the degree that external experiences formulate personal risk appetite, macroeconomic shocks also affect the overall financial situations. In more detail, Hanna and Chen in 1997 reported that the risks that individuals are capable to afford, depend also on their financial conditions. Therefore, it is natural to consider that fluctuations in wealth affect the proportions of risky assets that individuals hold in the sense that, good economic times rise individuals’ risk tolerance in the sense that they boost available resources, whereas poor economic climates increase risk aversion. On the other hand, Brunnermeier and Nagel (2008) investigated the effect of wealth fluctuations on household portfolio allocation and found no positive effect of liquid wealth changes on risky asset allocation. The authors explained their findings by enertia and implied that households are slow rebalancing their portfolios.

Apart from risk appetite and its interaction with age external experiences, literature also highlights a vast spectrum of other demographic, socioeconomic and attitudinal determinants that influence individuals’ choice on whether to hold risky investments. Particularly, studies that have assessed the impact of factors such as education, wealth, marital status and gender have reported important correlations related to risky asset possession.

To begin with, wealth is one of the most well-established determinants in the existing literature related to stock market participation. Empirically, most findings support a positive relationship between wealth and risky asset holdings. Riley and Chow (1992) explored the effect of different demographic factors on risk attitudes using a sample of 17,000 American households. The main advantage of this study is that they employed panel data, fact which enables them to track households’ wealth and stock holdings across the time. Their estimations were based on a Relative Risk Aversion Index (RRAI) that was defined as 1-(investment in risky assets/total wealth) and was incorporated as a dependent variable in the multivariate model. Furthermore, their results indicated that increase in income and wealth decrease the risk aversion of households. On the contrary, in cases of less wealthy individuals, some studies have also reported decreased possessions of risky assets as levels of wealth increased. In particular, Siegel and Hoban (1982) stated that once a certain level of financial security is reached, individuals can tolerate additional financial risk. On the other hand, in lower bounds of the wealth distribution, individuals are less inclined to tolerate risk, therefore possess larger amount of risky assets.

From another point of view, most of the prior research has focused on transaction and information costs in order to explain why people avoid to invest in stock markets. Therefore, for wealthier individuals and households it is easier to overcome the fixed costs of investing in the stock market. However, literature reports that even between the wealthiest individuals, a significant fraction individuals do not own stocks at all. Due to that, Campbell (2006) reports that it is not the wealth or income of a household that determines the choice to interact with stock markets, but the psychological barriers that make participation distressing.

Moreover, individuals’ educational background is also widely discussed as an important driver of stock market participation and risk. In more detail, the majority of studies find education to be positive related to risk taking and engagement with stock markets. Nonetheless, there is no general consensus on how education affects individual investment decisions. Zhong and Xiao (1995) determined education according to the number of years that investors have gone to school. Using cross-sectional data from 1989 Survey of Consumer Finances, their tobit regression results prove that taking risk increases as the level of education increases. On the other hand, using the same dataset, Jianakopoulos and Bernasek (1998) found that predicted risky assets decreased with the level of education for single men, single women and married couples.

Furthermore, different researchers identify education with financial literacy. Lusardi and Mitchell in 2007, posit that higher levels of education are essential to acquire information and essential skills in order to facilitate higher risk tolerance and actively involvement to financial markets. In line with that, Cole and Shastry (2009) report that education can affect decision making and alter individual behavior in several ways as it can increase cognitive skills and financial literacy, enhance job opportunities as well as attitudes and beliefs. In addition, other studies identify an indirect impact of education on investment decision-making. Hallahan et al. (2004) do not find a significant relationship between education and risk tolerance. However, their findings reveal that education is correlated with wealth, which as discussed in the previous paragraphs, is correlated with risky asset possession. Furthermore, early empirical work finds clear differences between genders in investment decisions and strategies. In more detail, Embrey and Fox (1998) conducted a survey in order to identify differences among men and women in investment decision. Their analysis proved that females are more risk averse as 62 percent of the studied women were not willing to take any risk at all, compared to 34 percent and 60 percent of men. However, the findings of their study are subject to criticism, since the average age in the sample was 60 for women and 46 for men. Therefore, results may be more driven by age rather than gender differences. In line with the aforementioned, different studies find women to have more conservative attitudes regarding to their portfolio choices, whereas men trade more aggressively and are more likely to invest in stocks. On the other hand, recent research suggests that due to higher educational attainments, women nowadays have more knowledge about the opportunities in the stock market which make them more likely to participate in it and take financial decisions.

Added to that, the relationship between individual stock market participation and marital status of the investor has been examined by several studies. Generally, the relationship among the individual risk tolerance level and the marital status seems fairly clear. Back in 1975, Corn et al suggested that married individuals invest a lower proportion of wealth in risky assets compared to single investors as married investors are more receptive to social risks. Likewise, the results of Hanna and Yao (2005) indicated that single men tolerate higher levels of risk, followed by married men, single women and lastly married women providing consistency with the previous studies which find females to be more risk averse than males. Bertocchi, Brunetti and Torricelli (2011) examine the joint effect of marital status and gender on risky portfolio choice. The estimated a binary choice model for the decision to invest in risky assets and ran two separate sets of regressions for the male and the female subsamples.

As risky assets they defined stocks, corporate bonds and foreign assets. Their results revealed a higher tendency for married women to invest in risky assets than single ones, while a marital status gap does not apply to men.

Lastly, apart from the traditional explanations related to risky attitudes, up to date literature incorporates alternative parameters. Particularly, recent studies explained that determinants such as cognitive skills, social interaction and the influence of peers personal values or political orientation also interact with risky holding investment decisions.