Risk aversion and group dynamics in the management of student managed investment fund

Risk aversion and group dynamics in the management of student managed investment fund

Peter A. Ammermann

ABSTRACT

The Student Managed Investment Fund (SMIF) at California State University, Long Beach, has been in operation for eight years, a time span that has covered a wide spectrum of market conditions. The records of the decisions of the SMIF managers over this time period allow for an examination of both the attitudes towards risk and return of the students and some of the factors that contribute to the formation of their expectations about the risk and return of investment alternatives. Using information about the numbers of stocks analyzed, the numbers of stocks actually approved for purchase, and the numbers of students in the class in each of the eight years, three main hypotheses are examined. First, that the students are return-seeking, in which case higher anticipated returns should lead to higher levels of stock-market activity. Second, that the students are risk-averse, in which case greater anticipated volatility would lead to reduced levels of stock-market activity. Third, that, due to the group dynamics of the class, the larger the class, the greater the apparent degree of risk aversion reflected in the students’ decisions. Correlation analysis is used to examine these three hypotheses, and evidence is found in support of all three hypotheses, although the evidence in support of the third is not significant. Interestingly, the evidence also suggests that the students’ risk and return expectations are informed by the stock market activity over the two years prior to the students’ entrance into the SMIF program, roughly the length of their academic careers prior to commencing SMIF.

1. INTRODUCTION

The Student Managed Investment Fund (SMIF) at California State University, Long Beach, began operations in 1996, and the 2002-2003 academic year marked the eighth year of operations for the fund. Over these eight years, which have covered a wide variety of market conditions, the student managers of the fund have achieved a successful record of portfolio management. The documented results of the program’s history allow for an examination not only of the fund’s performance but also of a number of issues of broader significance to the theory, practice, and teaching of investments.

A key axiom of finance is that investors are risk-averse. However, the evident degree of risk aversion that investors display varies over time, depending not only on market conditions but also on the setting in which the investment decision is made. In the case of CSULB’s SMIF program, the student portfolio managers demonstrate risk aversion characteristics that appear to differ from those that are typically demonstrated by students managing simulated portfolios. This behavior is evident both in the type of securities analyzed for possible inclusion in the SMIF portfolio and in the number of stocks actually selected for inclusion in the portfolio. It is anticipated that such behavior could be a consequence of both the market environment in the which the students are making their decisions and the dynamics of the groups within which the SMIF portfolio managers are making their investment decisions.

The remainder of the paper is organized as follows. The initial sections of this paper describe the history and institutional background of the SMIF program and provide discussions of how SMIF’s institutional structure would be anticipated to affect the students’ investment performance and demonstrated levels of risk aversion. These sections are followed by our empirical analyses and conclusions.

2. BACKGROUND OF THE SMIF SYSTEM PROGRAM

The SMIF program is an honors program consisting of students who are well-motivated high achievers with high GPA’s and a strong career interest in the field of investments. Admission to the academic year program is by application with a brief essay addressing the applicant’s career interests. Finalists for the 15 to 20 combined undergraduate and graduate positions are selected based upon the strength of the essay and performance during an interview by representatives of the Board of Advisors.

The class is by and large self-directed by the student participants, who develop the syllabus and the course schedule. The two faculty facilitators form the teams; leadership of the weekly classes is under the direction of a participant CEO. While both faculty members are always present in the classroom students are charged with the responsibility of directing class activities. However, both facilitators exert “suggested subtle directives,” and their omni presence, combined with the consistent use of penetrating questions, serves to bring understated but finely tuned focus to the course. This approach is utilized to build self-confidence among the participants and has been found to be very empowering to the students.

3. INVESTMENT PERFORMANCE AND SMIF

The key goal of the SMIF program is to provide a unique educational experience while minimizing the significance of investment performance. However, it has become quite evident that students are driven by the sense of importance of managing a portfolio that performs well. Theoretically, the SMIF portfolio could outperform professional money managers. If markets were truly efficient, SMIF’s performance, which is free from salaries and overhead costs, should exceed that of professionally managed portfolios. Various measures of the collective performance of money managers indicate that 60% to 90% underperform their market benchmark. However, the $50,000 size of the SMIF portfolio may mitigate this advantage. Further, the SMIF portfolio starts each new academic year with $50,000 in cash, and at the end of the spring term each year the portfolio is liquidated. This approach is by design to allow participants to experience the idiosyncrasies of portfolio construction as well as portfolio management.

4. SMIF INVESTMENT GUIDELINES

In addition to the cash-to-cash paradigm the portfolio must adhere to diversification guidelines including a limitation of no more than five percent of the portfolio invested in any one issue and no more than fifteen percent invested in any single industry. Fixed income instruments at or above investment grade must comprise 25% to 50% of the portfolio with the balance in equities. Stop-loss orders are utilized at ten percent of purchase price, or a trailing ten percent stop loss may be utilized at the discretion of the SMIF participants. Prior to the purchase of a security, which requires the approval of a two-thirds majority, a long-term price target must be established, prorated to the length of time remaining to the liquidation of the portfolio at the end of the spring term. Again, this brings emphasis to the portfolio construction aspect of the program while also recognizing the significance of the management of assets once they have been placed in the portfolio. The SMIF program is overseen by a three member Board of Advisors consisting of a retired CEO from a major savings and loan firm, a former branch manager and current broker in a major full service brokerage firm, and a senior tenured professor.

5. SMIF VERSUS CLASSROOM SIMULATIONS

Most instructors of investment courses employ classroom simulations of portfolio construction and management. These simulations tend to focus on performance either by design or by the driven desires of the students. With the will to excel by being recognized for outstanding performance, simulations often lead to excessive risk-taking. The “real money” dimension of the SMIF program has resulted in a more conservative posture with a focus on the preservation of capital. Adverse incentives in simulations, such as recognition by peers for achieving outstanding performance, or recognition for mismanagement in the attainment of the worst performing portfolio, often serve as motivating factors. This inherent shortcoming is frequently overlooked by the instructor who is often more interested in having students quickly construct portfolios to expedite coverage of text material. On the other hand the SMIF program has not institutionalized risk aversion, but students in the program are sensitive to the career enhancing nature of their pursuit and thus are more conscious of the implications of a well constructed and managed portfolio. Finally, simulations are generally driven by instructor directed activities with a pre-specified timeline, while the SMIF program is process driven. Program participants recognize that investing is rather methodical in nature, and sound principals simply cannot be ignored in the quest to rush quickly to invest the proceeds of a fictitious portfolio. The primary emphasis in the SMIF endeavor is to allow students the opportunity to apply what they have previously learned in the courses that are prerequisites for admission to the SMIF program. This process orientation is perhaps the most significant differentiating aspect of SMIF.

6. SMIF AND RISK AVERSION

Group dynamics play a significant role in the level of risk aversion in the SMIF program. Teams, generally consisting of four members, must first convince the other members of their team that a proposed security warrants sufficient merit to be considered for the portfolio. Subsequently, via a team presentation, compelling evidence must be presented in a manner that is sufficiently persuasive to motivate a two-thirds majority of the class members to vote affirmatively to include the security in the portfolio. This also differentiates the SMIF approach from that of simulated classroom portfolios where students generally have sole responsibility for issues that are included in portfolios. Each academic year the SMIF program culminates with an annual report that sets forth the collective activities of the participants. The report includes the names, and generally pictures, of the participants. This brings permanence to the performance perspective and also seems to assure that participants seriously take responsibility for their investment approach with due consideration of risk. The annual reports are widely utilized by students as supplements to resumes presented to prospective employers in their job related search activities. This serves to substantiate that SMIF participants have indeed supplemented their academic pursuits with real-world portfolio construction and management accomplishments.

7. HYPOTHESES AND EMPIRICAL RESULTS

At the end of each academic year, the student portfolio managers of SMIF must develop an annual report that documents their portfolio construction and management process over the prior year. With the results now in for the 2002-2003 academic year, there are now eight years’ worth of observations available for empirically examining the risk versus return expectations and attitudes of the SMIF students. Specifically, the three hypotheses of interest are as follows.

Hypothesis One: The students are return-seeking In other words, the higher the anticipated returns for stocks, the greater the number of stocks the students will analyze and will accept for purchase.

Hypothesis Two: The students are risk-averse. The greater the anticipated risk or volatility of stocks, the fewer the number of stocks the students will analyze and will accept for purchase.

Hypothesis Three: A greater number of students in the SMIF class leads to a greater degree of risk aversion. Given the structure of the class, the need to achieve consensus in the presentation and selection of stocks should lead to behavior that appears more risk-averse the greater the number of students in the class. Or, to look at it another way, the more uncertainty there is about a given stock, the less likely it will be that the students will be able to come to a consensus to examine that stock more closely and then to vote to buy it, and so the greater the degree of uncertainty there is about the market as a whole, the smaller the number of stocks for which the class could come to a favorable consensus.

These three hypotheses are examined using correlation analysis of the number of stocks analyzed, the number of stocks voted for purchase, and the percent of stocks voted for purchase out of all the stocks analyzed, with the number of students in the class and with two separate measures for the return and risk of stocks. In studying return-seeking and risk-averse behavior, a key issue that is always involved is how to measure the expected return and the risk that is seen in the eyes of the participants. Because the recent history of the stock market is likely to inform the perceptions that the students have coming into the class, two measures of historical return and historical volatility are used–the average monthly return and the monthly volatility over the year prior to the start of the SMIF class’s investment activities, and the average monthly return and the monthly volatility over the two years prior to the start of the SMIF class’s investment activities.

The results of the correlation analysis are shown in Table 1. The evidence in the table provides support for all three of the hypotheses. For Hypothesis Three, however, although the estimated correlations do have the correct sign (a negative correlation with the number of students in the class, suggesting that a greater number of students in a SMIF class is associated with fewer stocks presented, with fewer stocks voted for purchase, and with a smaller percentage of stocks voted for purchase), none of these correlations is significant. The strongest of these three correlations is the correlation between number of students and number of stocks presented, for which the estimated correlation is -0.613, with a significance level of 0.108. It is important to note, though, that the small number of observations (there is only eight years’ worth of data, so N = 8) makes it more difficult to find significant results. Nonetheless, the results obtained do suggest that the team / consensus structure does contribute to more risk-averse behavior among the class than might otherwise be the case.

With regard to the first hypothesis (Hypothesis One: return-seeking behavior), a positive correlation would be expected between the measures of interest in the stock market (numbers of stocks analyzed and presented, numbers of stocks voted for purchase, and percentage of stocks voted for purchase) and the measures of anticipated returns (specifically, the average monthly returns over the year prior to the commencement of the SMIF class and the average monthly returns over the two years prior), and these are what are found. While the correlations between the return measures and the numbers of stocks analyzed are statistically insignificant, they are, nonetheless, positive. Moreover, both the measures of return (prior one year’s and prior two year’s) are significantly positively correlation with both the percentages of stocks purchased (significant at the 0.05 level) and the numbers of stocks purchased (significant at the 0.01 level).

Finally, with regard to the second hypothesis (Hypothesis Two: risk-averse behavior), the evidence is somewhat weaker, but once again it is supportive of the hypothesis, with negative correlations found between the measures of interest in the stock market and the measures of anticipated volatility or risk. In this case, though, the only significant results are those between the average monthly volatility over the prior two years and both the numbers of stocks analyzed and the numbers of stocks purchased.

This latter result is interesting–the average monthly volatility over the prior two years has a stronger association than the average monthly volatility over the prior year with the measures of the students’ interest in the stock market. This suggests that the students’ decisions and expectations are being more strongly affected, either implicitly or explicitly, by the more extended record of the stock market, especially in terms of assessing the risk of the market and are not being too heavily influenced by shorter-term activity in the stock market. Notably, this longer, two-prior-year horizon would also roughly coincide with the period of time that the average SMIF student spends at CSULB prior to enrolling in the SMIF program.

8. CONCLUSION

While the typical classroom approach to teaching the subject of investments appropriately provides an opportunity to allow students to pursue experimentation via simulations the limitations are vast and perhaps in some circumstances even counter productive. Outcomes are very fleeting, lasting only for the duration of a few weeks, and the sole focus is on the course itself. On the other hand the SMIF program subjects students to a “greater authority” to include not only their classroom peers but also the recognition of participating in an honors program that is favorably viewed throughout the college, the university and the community at large. Outcomes of their academic year experiences are memorialized in the annual report. And, finally, the entire SMIF experience brings a career focus that is otherwise unavailable to students by providing an opportunity to engage in real world experience that is not generally afforded by academia. These real-world aspects of the program do appear to be successful in encouraging behavior on the part of the students that is both long-term-oriented but responsibly risk-averse, an outcome that is often not the case for simulated portfolio classroom portfolios.

Table 1: Results of Correlation Analysis

Correlations

NUANALYZ NUPURCH

NUANALYZ Pearson Correlation 1 .694

Sig. (2-tailed) . .056

NUPURCH Pearson Correlation .694 1

Sig. (2-tailed) .056 .

NUSTUD Pearson Correlation -.613 -.448

Sig. (2-tailed) .106 .266

PERPURCH Pearson Correlation .096 .776 *

Sig. (2-tailed) .821 .024

Prior1Ret Pearson Correlation .613 .875 **

Sig. (2 tailed) .106 .004

Prior1Vol Pearson Correlation -.609 -.610

Sig. (2 tailed) .109 .108

Prior2Ret Pearson Correlation .587 .874 **

Sig. (2-tailed) .126 .005

Prior2Vol Pearson Correlation -.710 * -.723 *

Sig. (2 tailed) .048 .043

NUSTUD PERPURCH

NUANALYZ Pearson Correlation -.613 .096

Sig. (2-tailed) .106 .821

NUPURCH Pearson Correlation -.448 .776 *

Sig. (2-tailed) .266 .024

NUSTUD Pearson Correlation 1 -.133

Sig. (2-tailed) . .753

PERPURCH Pearson Correlation -.133 1

Sig. (2-tailed) .753 .

Prior1Ret Pearson Correlation -.555 .722 *

Sig. (2 tailed) .153 .043

Prior1Vol Pearson Correlation .356 -.302

Sig. (2 tailed) .386 .468

Prior2Ret Pearson Correlation -.698 .735 *

Sig. (2-tailed) .054 .038

Prior2Vol Pearson Correlation .545 -.367

Sig. (2 tailed) .162 .372

Prior1Ret Prior1Vol

NUANALYZ Pearson Correlation .613 -.609

Sig. (2-tailed) .106 .109

NUPURCH Pearson Correlation .875 ** -.610

Sig. (2-tailed) .004 .108

NUSTUD Pearson Correlation -.555 .356

Sig. (2-tailed) .153 .386

PERPURCH Pearson Correlation .722 * -.302

Sig. (2-tailed) .043 .468

Prior1Ret Pearson Correlation 1 -.714 *

Sig. (2 tailed) . .047

Prior1Vol Pearson Correlation -.714 * 1

Sig. (2 tailed) .047 .

Prior2Ret Pearson Correlation .915 ** -.474

Sig. (2-tailed) .001 .235

Prior2Vol Pearson Correlation -.786 * .958 **

Sig. (2 tailed) .021 .000

Prior2Ret Prior2Vol

NUANALYZ Pearson Correlation .587 -.710 *

Sig. (2-tailed) .126 .048

NUPURCH Pearson Correlation .874 ** -.723 *

Sig. (2-tailed) .005 .043

NUSTUD Pearson Correlation -.698 .545

Sig. (2-tailed) -.054 .162

PERPURCH Pearson Correlation .735 * -.367

Sig. (2-tailed) .038 .372

Prior1Ret Pearson Correlation .915 ** -.786 *

Sig. (2 tailed) .001 .021

Prior1Vol Pearson Correlation -.474 .958 **

Sig. (2 tailed) .235 .000

Prior2Ret Pearson Correlation 1 -.638

Sig. (2-tailed) . .089

Prior2Vol Pearson Correlation .638 1

Sig. (2 tailed) .089 .

N = 8

* Correlation is significant at 0.05 level (2 tailed).

** Correlation is significant at the 0.01 level

(2-tailed).

NUANALYZ = the total number of stocks analyzed and presented by the

given year’s SMIF class

NUPURCH = the total number of stocks the given year’s SMIF class

voted to purchase

NUSTUD = the number of students in the given year’s SMIF class

PERPURCH = the percent of stocks voted for purchase out of all the

stocks analyzed and presented

Prior1Ret = the average monthly return for the S&P 500 over the year

prior to the start of the SMIF class’s investment activities

Prior1Vol = the average monthly risk or volatility for the S&P 500

over the year prior to the start of the SMIF class’s investment

activities

Prior2Ret = the average monthly return for the S&P 500 over the two

years prior to the start of the SMIF class’s investment activities

Prior2Vol = the average monthly volatility for the S&P 500 over the

two years prior to the start of the SMIF class’s investment activities

Dr. Peter A. Ammermann earned his Ph.D. at Virginia Tech in 1999. He is currently an Assistant Professor at California State University, Long Beach, and a co-administrator of CSULB’s SMIF program.

Dr. L. R. Runyon earned his D.B.A. at the University of Southern California in 1969. He is currently the Chair of the Department of Finance, Real Estate, and Law at California State University, Long Beach, and the administrator of CSULB’s SMIF program.

COPYRIGHT 2003 International Academy of Business and Economics

COPYRIGHT 2004 Gale Group