Being Warren Buffett: a classroom simulation of the stock market
March 24, 2009 Nicholas J. Horton Department of Mathematics and Statistics Smith College, Northampton, MA nhorton@smith.edu http://www.math.smith.edu/~nhorton
Acknowledgements and references
Activity developed by Robert Stine and Dean Foster (Wharton School, University of Pennsylvania) Published paper: “Being Warren Buffett: A classroom simulation of risk and wealth when investing in the stock market”, The American Statistician (2006), 60:53-60. More information, the handout form and copy of the TAS paper can be found at: http://www-stat.wharton.upenn.edu/~stine A copy of these notes plus the R code to run the simulation and results from 5000 simulations can be found at: http://www.math.smith.edu/~nhorton/buffett
Horton – Causeweb - 2009
Being Warren Buffett
2
Overview
The concepts of expected value and variance are challenging for students A hands-on simulation can help to fix these ideas, in the context of the stock market Allows students to experience variance first-hand Can be implemented using dice (and calculators) in a classroom setting Computer generation of results complements the analytic and hand simulations
Horton – Causeweb - 2009
Being Warren Buffett
3
Objectives
Understanding discrete random variables to model stock market returns Calculate and interpret expectations for return from a given investment strategy Calculate and interpret standard deviations of returns from a given investment strategy Compare the risk and return for these strategies Spark thinking about diversification and rebalancing of investments
Horton – Causeweb - 2009
Being Warren Buffett
4
Background information
Imagine that you have $1000 to invest in the stock market, for 20 years Three investment possibilities are presented to students in groups of 2 or 3:
Investment
Green
Expected annual return
8.3%
SD(annual return)
20%
Red White
71% 0.8%
132% 4%
Question: Which of the three investments seems the most attractive to the members of your group?
Horton – Causeweb - 2009
Being Warren Buffett
5
Dice outcomes
The investments rise or fall based on the outcomes of a 6-sided die: Outcome 1
2
Green 0.8
0.9
Red 0.05
0.2
White 0.95
1
3 4 5
6
1.1 1.1 1.2
1.4
1 3 3
3
1 1 1
1.1
Horton – Causeweb - 2009
Being Warren Buffett
6
Example:
Suppose on the first roll your team gets the following outcomes (Green 2) (Red 5) (White 5), then on the second roll, you get (Green 4) (Red 2) (White 6)
Round Start Return 1 Value 1 Return 2 Value 2 Green $1000 0.9 $ 900 1.1 $ 990 Red $1000 3 $3000 0.2 $ 600 White $1000 1 $1000 1.1 $1100
Horton – Causeweb - 2009
Being Warren Buffett
7
Repeat the process for 20 years
1 student to roll the dice (green, red and white) 1 student to determine the return and calculate the new value on the results handout 1 student to supervise and catch errant dice
At the end of class, each team enters their results on the classroom computer Find out who are the “Warren Buffett’s” of the class
Horton – Causeweb - 2009
Being Warren Buffett
8
Group results form
Horton – Causeweb - 2009
Being Warren Buffett
9
Usually, red doesn’t do as well as green
Horton – Causeweb - 2009
Being Warren Buffett
10
But occasionally it wins big!
Horton – Causeweb - 2009
Being Warren Buffett
11
Expected returns for 20 years
Use property that the expectation of a product is the product of the expectation GREEN: $1000*(1.083)^20= $ 4,927 RED: $1000*(1.710)^20= $45,700,632 WHITE: $1000*(1.008)^20= $ 1,173
We’d always want to pick RED, no?
Horton – Causeweb - 2009
Being Warren Buffett
12
Observed returns (using simulation)
Used R to simulate 5000 20-year histories, available as “res.csv” Observed Q1, median, Q3
GREEN: $2,058 $3,621 $6,269 RED: $ 0 $ 16 $1,993 WHITE: $1,011 $1,141 $1,321
Percentage ending with less than initial investment ($1000)
GREEN: 5.9% RED: 72.7% WHITE: 25.0%
Horton – Causeweb - 2009
Being Warren Buffett
13
Another strategy (“pink”)
Consider a strategy where you balance investments between RED (dangerous) and WHITE (boring) each year Call this “PINK” Smaller average returns, but far less variable Can be calculated using existing rolls (average returns), using space on the results form
Horton – Causeweb - 2009
Being Warren Buffett
14
How to implement PINK
Pink $1000 2 $2000 0.525 $1050
3 $3000 0.05 $150
1 $1000 1 $1000
Horton – Causeweb - 2009
Being Warren Buffett
15
Connections to reality and thoughts on “pink”
GREEN performs like the US stock market (adjusted for inflation) WHITE represents the (inflation adjusted) performance of US Treasury Bills Quote from authors: “We made up RED. We don’t know of any investment that performs like RED. If you know of one, please tell us so we can make PINK!”
Horton – Causeweb - 2009
Being Warren Buffett
16
Boxplots of results (needs rescaling)
$80 billion!
Horton – Causeweb - 2009
Being Warren Buffett
17
Boxplots of results (where returns <=$50,000)
Horton – Causeweb - 2009
Being Warren Buffett
18
Teaching materials and checklist
Copies of handout describing the simulation (one per student) Copies of results sheet (one per group) Set of three die (though one will work in a pinch, one set per group) Remind students to bring calculators (or run this in a lab rather than lecture) Time requirements: between 50 and 80 minutes (depending in part on whether you calculate expected values, motivate the simulation parameters in terms of historical inflation and stock returns and whether “pink” is introduced)
Horton – Causeweb - 2009
Being Warren Buffett
19
Extensions and assessment
The activity was developed for use in both an MBA and PhD program The paper introduces concepts of “volatility drag” and “volatility adjusted return” as more advanced topics (potentially applicable as a project at the end of a undergraduate probability class), as well as connections to calculus Verifying the expected value and standard deviation of one of the investment strategies is a straightforward homework assignment (other assessments possible) Students without formal exposure to expectations of discrete random variables can still fully participate in the simulation
Being Warren Buffett 20
Horton – Causeweb - 2009
Conclusions
Hands-on activity is popular with students Helps to reinforce important but often confused concepts in the context of a real world application Small group work helps to address questions as they arise Students turn in results to allow review of results (in addition to immediate display of summary and graphical statistics)
Horton – Causeweb - 2009
Being Warren Buffett
21
Being Warren Buffett: a classroom simulation of the stock market
March 24, 2009 Nicholas J. Horton Department of Mathematics and Statistics Smith College, Northampton, MA nhorton@smith.edu http://www.math.smith.edu/~nhorton