01_(286)_Victor1558_Flickr_CreativeCommons

Real Data Suggest Gender Biases in Investing

in Investing by

We took a look at the Personal Capital user base to see how whether the oft-cited fact that women are less risk tolerant than men holds true for our users.  Our data set reveals the following:

  • On our 1-5 scale of risk tolerance, women are on average 7% more risk-averse than men.
  • The effect of gender on risk tolerance is greater than that of any other variable.

Why is this significant?  An investor’s risk tolerance is pivotal in to determining an investing strategy and ultimately, a portfolio’s returns.  The fact that women are more risk averse than men means they may be systematically leaving returns on the table.  We calculate that this may translate to 30 basis points (bps) per year.  Over a long investing horizon, that difference could mean a woman’s portfolio is over 10% smaller than a man’s.

In the following post, I first discuss our data analysis and then compare hypothetical portfolios to illustrate the opportunity loss of a lower-risk portfolio.  I then discuss measures that can be taken to empower women – and men – to make better investing decisions.

Female Personal Capital Users Are More Risk-Averse

To see how risk tolerance compares among female and male Personal Capital users, we examine the data of users who have been for whom we have both demographic information and data on risk tolerance.  Demographic data includes: gender, age, retirement annual income range, net worth range, tax rate range, and number of years employed.[1]

Risk tolerance is a user-defined variable on 1 to 5 scale from low to high risk.  Users are prompted to select from among one of the following statements to describe their attitude toward risk:

  • Highest Safety (1) – “Market volatility makes me very uncomfortable. Safety is a much higher priority than growth for me, and I do not expect growth meaningfully above inflation.
  • Conservative (2) – “I am able to accept some volatility, but have difficulty stomaching meaningful fluctuations in account values. I expect long term growth somewhat above inflation, but am willing to sacrifice up to half of my potential long term return in exchange for less volatility.
  • Moderate (3) – “I am comfortable with moderate volatility consistent with a diversified portfolio which includes a significant allocation to stocks. I prefer to sacrifice some long term return in order to reduce risk.
  • Aggressive (4) – “My primary objective is to achieve growth, and I am comfortable with typical stock market volatility. Still, I am willing to trade a small amount of growth potential to reduce risk.
  • Highest Growth (5) – “I am willing to take a high degree of risk in pursuit of higher returns, and am very comfortable with the volatility of a 100% stock portfolio.”

We begin with a simple analysis to examine difference in risk tolerance between men and women.  The difference between the means of men and women is 0.36, or 7.2% given our 1-5 scale.  The chart below shows the results, where the bar describes mean risk tolerance with its standard error (T=5.37 in our two-sided t-test, meaning that it is statistically significant).

1_140130_pc-report-charts

We then looked at gender to control for the other demographic factors that we record that may be related to risk tolerance (for results of the multivariate analysis, see Table below).  As it pertains to gender, the result was as follows:

  • Gender has statistically significant relationship with risk tolerance.
  • Not only is this relationship is significant in the presence of any other control variable, but the magnitude of the gender effect is larger than of any other variable.
Gender_Multivariate

Implication: Lower-Risk Portfolios May Miss Return Opportunities

To illustrate how having “7% lower risk tolerance” might impact returns, we can examine the hypothetical portfolios of two individuals, Mary and Brian.

Mary and Brian are both 30 and expect to retire at 68.  They both earn $100,000, and have already saved $100,000 in their retirement accounts.  Furthermore, they have a healthy savings rate of $10,000 per year.  Given their identical situations and objectives, they can both afford to have growth-oriented portfolios.

Left on his own, Brian developed a long-term growth allocation with mostly stocks but including other asset classes to achieve moderately lower volatility.  The expected return on his portfolio is 8.8% (based on historical returns), with a standard deviation of 13.6%.  The chart below illustrates where Brian’s portfolio (the blue circle with a star) falls on the efficient frontier.[2]

2a_140130_pc-report-charts

Now, let’s turn to Mary’s portfolio.  Let’s say Mary also had an efficient portfolio with the same objectives as Brian.  However, she has a lower risk tolerance and has invested in an asset mix that reflects that preference.

As we demonstrate in the chart below, the reduction in the risk level of Mary’s portfolio moves her portfolio to the left on the efficient frontier (her portfolio is represented by the blue star).  A 7% lower risk level means a portfolio with expected returns of 8.5% with a standard deviation of 12.5%.  While her portfolio is still efficient, its expected return is 30bps less than Brian’s.

2b_140130_pc-report-charts

30bps may not seem dramatic, but if we project Mary and Brian’s portfolios into retirement, this difference equates to nearly $50,000, or 8% difference in their median expected portfolio value at retirement.[3] If they were 20 years old – i.e. with a longer time horizon to retirement – this figure would be over 10%.  The chart below illustrates this “opportunity loss.”

3_140130_pc-report-charts

Conclusion

In our study, we’ve shown that female Personal Capital users have a lower risk tolerance than male Personal Capital users by 7%.  We estimate that for two efficient portfolios, returns could be 30bps lower for the less risky portfolio.  Over a long time horizon, this differential can translate into a 10%+ difference in portfolio values.

It’s true that too much risk is a bad thing, exposing an investor to unnecessary losses.  It’s possible that men tend to adopt too high-risk strategies relative to their needs.[4] However, we believe that it’s more likely that women are underestimating the appropriate risk.  First, that’s because as studies have shown, as women’s knowledge and confidence increases, so does risk tolerance.[5] This implies that women tend to have too low risk relative to what’s appropriate.  Second, because women live longer, meaning longer investing time horizons, they generally should be taking more risk relative to men.[6]

Either way, our study underscores why figuring out the right level of risk of a portfolio is so critical.  Further,  because there is a clear possibility of a systematic gender bias, it’s imperative for individuals and financial planners to be on guard for gender risk bias in evaluating an investor’s appropriate risk levels. Making an appropriate risk assessment is the first step for both men and women toward making better investing decisions.

Find Out Your Appropriate Risk and Target Allocation with Personal Capital’s Free Financial Software

 


 

[1] Personal Capital collects demographic data during the “enrollment” process, when users sign up to become customers.  The data sample represents over 2,000 users.  For risk tolerance, which is collected for all users, n > 250,000.  The enrollment selection may introduce a bias into the data if people who choose to enroll have other factors (such as wealth and education) that we have not accounted for that impact risk tolerance.
[2] The goal in portfolio design is to minimize a portfolio’s level of risk for its level of expected return. In other words, it is to be on the “efficient frontier” – which represents the set of investment portfolios that provide the greatest return for the risk you take.  The first critical step in portfolio design is mapping out the amount of risk that is appropriate for an investor to take.  Understanding that level, you can find out where to be on the efficient frontier.
[3] We use Monte Carlo analysis to project potential outcomes of the allocations. We use the expected annual return and standard deviation and runs these inputs through a computer to simulate a large number of random scenarios. The values reported here represent the median scenario – or the midpoint of the simulations – and can be considered an expected value based on historical results. This analysis is inflation adjusted, but assumes no fees, taxes or cash flows.
[4] Further research is needed to show whether Personal Capital users overestimate or underestimate their risk.
[5] Eckel, Catherine and Grossman, Philip.  “Men, Women and Risk Aversion: Experimental Evidence.” Handbook of Experimental Economic Results, Volume I, 2008.
[6] Sherman, Hanna and Yao, Rui.  “The Effect of Gender and Marital Status on Financial Risk Tolerance,” Journal of Personal Finance, 2005.

 

The following two tabs change content below.
Catha Mullen

Catha Mullen

Catha Mullen is passionate about helping people make healthier financial decisions, which is why she joined Personal Capital. Personal Capital helps people live better financial lives by providing technology-enabled advisory services, in addition to free financial software. She's got an MBA from Stanford and AB from Princeton.

5 comments

  1. Sam Dogen

    Such a fascinating study Catha! I don’t doubt that women are more conservative than men when it comes to investing. Perhaps it’s due to interests and knowledge of investing. For example, 75%+ of the people on the trading floors of my old Wall St. jobs were men. The more you THINK you know about investing, the more aggressive you become.

    But, as well all know, it’s very hard to outsmart the markets in the long run. As a result, I’d rather go with the conservative side, especially the older I get.

    Everybody needs to look out for themselves when it comes to their money, even if they are in a relationship. One never knows!

    Sam

    Reply
  2. Michael Ruderman

    Really fascinating! I wonder if this 7% delta between men and women is consistent across age groups. It’d be really interesting if younger women and older women differed in their risk tolerances (in a substantially different way from men of the same age). That might inform some strategies to address this, as targeting women in their 20s or 60s is likely a lot easier and a lot more effective than just targeting all women.

    Reply
  3. Barbara Friedberg

    Interesting statistics. Yet, in general, those who are averse to risk (as well as educated about market history) and have a lower tolerance for portfolio volatility, may be willing to accept a lower anticipated return in exchange for less risk. An additional caveat is that future returns are estimates and are based upon history (which may or may not be predictive).

    Reply
  4. Joe Saul-Sehy

    Great piece.

    I’ve seen similar studies that have tied some of the difference in risk tolerance to the difference in life expectancy. Because women live longer, they’re more likely to be careful with funds, expecting to need more later in life.

    Reply

Trackbacks/Pingbacks

  1.  Charts of the day, female risk-aversion edition | Felix Salmon

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Disclaimer. This communication and all data are for informational purposes only and do not constitute a recommendation to buy or sell securities. You should not rely on this information as the primary basis of your investment, financial, or tax planning decisions. You should consult your legal or tax professional regarding your specific situation. Third party data is obtained from sources believed to be reliable. However, PCAC cannot guarantee that data's currency, accuracy, timeliness, completeness or fitness for any particular purpose. Certain sections of this commentary may contain forward-looking statements that are based on our reasonable expectations, estimate, projections and assumptions. Forward-looking statements are not guarantees of future performance and involve certain risks and uncertainties, which are difficult to predict. Past performance is not a guarantee of future return, nor is it necessarily indicative of future performance. Keep in mind investing involves risk. The value of your investment will fluctuate over time and you may gain or lose money.