Previous research has shown that one’s emotional state is linked to one’s risk-taking behavior. The author’s research is unique in that the results are not derived by experiments—the common method in behavioral finance. Instead, the author uses two datasets to derive her conclusions. Some previous research has investigated the effects of lotteries on stock turnover. The author instead focuses on risk taking and uses monthly data instead of the less frequently occurring lottery jackpots.

Behavioral finance has shown that people do not always behave rationally and are not simply economic agents trying to maximize their utility. Risk taking, for example, can change over time and with one’s mood and can even be influenced by external factors, such as politics or the weather. Risk aversion is mainly driven by fear rather than by economic wealth. But not many studies in this area have results based on actual data; the author’s research is unique in that it does not include any experimental settings and is based on pure data analysis.

The author applies the difference-in-differences approach to derive her conclusions, with two datasets from two different time periods. The first dataset is used to draw conclusions for the second dataset. The author supports the use of this approach with the fact that the opening of a casino can be considered an exogenous event and the investment behavior of gamblers and nongamblers is similar before casino openings.

This research is particularly interesting for investment professionals advising individual clients. Because individual clients might not have a financial background and are not always bound by regulations, their investment decisions might be influenced by their emotional states. Financial professionals with a better understanding of how these emotional states can affect their clients’ preferences will be better equipped to cope with them.

Gambling became more popular in the United States in the late 1980s and in the 1990s because of changes in public opinion and state budget deficits. The author, therefore, selects the 1991–96 time period, in which more than 100 casinos opened, for her study. She uses survey data for October and November 2012 from Amazon Mechanical Turk, an online forum for survey data. The author considers this survey dataset, consisting of 1,750 investors, to be appropriate because it reflects the US population well.

She then evaluates the gambling behavior of the respondents and concludes that gamblers tend to be significantly older and have higher incomes than the population average.

The second dataset consists of investors’ brokerage data between 1991 and 1996, resulting in a total of 40,498 investors. This dataset also reveals that gamblers tend to be older, male, and retired and have higher incomes. One of the weaknesses of this dataset is that only the ZIP code data for 1996 are collected and no relocations between 1991 and 1996 are taken into account. The casino opening data and brokerage data are then compared, and only the opening of the first casino within 50 miles in a particular area is considered an event.

The author applies the portfolio beta, stock volatility, and portfolio volatility as measurement criteria and concludes that the opening of casinos results in more financial risk taking for gamblers who are older, male, and married and have higher incomes. Furthermore, she concludes that the casino effect is not restricted to a 50-mile radius and that the effect persists when performing the same analysis for a 100-mile radius. She notes that her analysis has shown increased risk taking for stock investments and that this increase is persistent through time. The author does not assess the impact of a casino opening on an investor’s entire portfolio volatility, but previous research has shown that household portfolio allocations do not change much over time.

This article is unique in terms of both the research method—purely data driven—and the research topic—how nonfinancial decisions can influence financial decisions. The author provides a good introduction to the development of gambling in the United States before describing her data-selection method. Because the difference-in-differences methodology is not widely known, it would have been helpful to introduce this approach with an example. Nevertheless, the article as a whole is very detailed, and the author has performed a deep analysis on the topic, making it a very interesting article for financial professionals.