Papers Flashcards
results of field experiment on savings and commitments*
Ashraf et al 2005, out of 710 treated people, 202 take up SEED (28%), average saving balances increased 81% for clients assigned to treatment, more ‘hyperbolic’ individuals seem more willing to take up SEED, marketing treatment had no effect
explain setting of experiment on exercise and health*
Royer et al 2013, 1000 employees at Fortune 500 company, 2 years, treatments: one month financial incentive to attend company onsite exercise facility (10USD per visit up to 3 each week), half of the incentive group randomly selected and offered to put money at stake for pledge they would continue to use over next 2 months
results of experiment on exercise and health*
Royer et al 2013, pure incentive only worked temporarily and went back to normal levels of attendance, 12% take up of commitment contract people who took up commitment increased gym attendance permanently compared to pretreatment levels, hyperbolic and sophisticated (if put money have extra incentive to go)
explain setting of field experiment on work effort *
Kaur et al 2010, Indian IT firm small city 124 workers, work was digitalising questionnaires the more you do the more money you get, since effort at work is immediate and benefits delayed, effort at work is an investment good, workers offered contract that penalises them for not achieving a (self-chosen) threshold over the same contract, with no advantages
what is the main hypothesis of the field experiment on work effort *
Kaur et al 2010, worker’s effort increases as the payday gets closer (test of hyperbolic, if hyperbolic future person bears little weight) workers choose a contract which penalises them for not achieving a (self-chosen) threshold over the same contract without a penalty (are people sophisticated?, do they accept crazy contract?)
how does hyperbolic deal with getting work done before a deadline
would leave until quite last minute as the future bears little weight so would put off until deadline
explain setting of field experiment on addiction *
Gine et al 2010, 2000 smokers Philippines, offer urine test for smoking in 6 months, offered some opportunity to voluntarily sign commitment contract to stop smoking, smoker signing contract pledged own money that they would pass urine test in 6 months
results of field experiment on addiction *
gine et al 2010, 11% take up of 781 offered, average deposit 57 pesos, at 6 months, increase of 4-5% in chance of making urine test, at 12 months, similar increase at surprise test
what are visceral influences
incorporate “visceral” influences such as hunger, sexual desire, physical pain, cravings
what was the extension of Saunders (1993)*
Hirshleifer and Shumway (2003) extend to 26 countries, use weather of city where stock market located, negative relationship between cloud cover (de-trended from seasonal averages) days with completely covered skies have daily stock returns .11% lower than days with sunny skies
what is the experiment for time varying risk aversion
Guiso et al 2013, want to see if investors were emotionally affected by a stock market crash leading to increase in risk aversion, sample of students with a five minute clip of horror film Hostel, subjects who watched horror movie shows higher level of risk aversion (certainty equivalent 27% lower than those who didn’t watch movie)
what is the discount rate and what is the discount factor
factor (1/1+ρ) rate ρ
what is the field experiment of effect of mood (weather)*
Saunders (1993), days with higher cloud cover in New York are associated with lower aggregate US stock returns
explain methodology of paper looking at trading differences between men and women
Barber and Odean (2001), investment accounts of 78,000 hh 1991-96, net worth, investment experience, demographics, difference tests, cross sectional regressions
what were results from trading differences between men and women
Barber and Odean (2001), men trade 45% more than women, trading always hurts performance but men losses 54% higher, effects more pronounced when comparing single men and women, men chose riskier decisions
what were robustness and limitations of the differences between men and women
Barber and Odean (2001), Robustness: no diff in picking high-performance assets (no ability diff), control for demographics in regression, men only risk averse? (as opposed to overconfident), Limitations: correlation (not necessarily causal), is it really over-confidence or something else about gender, Problem: gender cannot be randomly assigned
what is the setting of experiment on betting behaviour in lotteries
Clotfelter and Cook (1993), look at betting behaviour in lotteries, “Maryland daily numbers lottery” -> P(Win)=0.001, P(Win|RecentWin) is still 0.001, look at 52 winning numbers in 1988, in 52/52 cases betting volume for previous winner decreases 3 days after win, relative to before
what is the setting of paper on pension contributions
Benartzi (2001), many companies offer employees to invest parts of pension contributions in company stock, question: does stock price depend on past stock performance, potential case of over-inference
what is the method and results of paper on pension contributions
Benartzi (2001), S&P 500 companies with retirement program, yearly data on own-company investments for 2/3 of companies, 2.5 mn participants, $102bn assets, company that performed badly (well) over last 5 yr are far less (more) likely to invest pension (statistically significant), large effect of past returns, effect depends on long-term performance
what are the challenges to paper on pension contributions
Benartzi (2001), objection: is it insider information?, Unlikely: high-allocation companies perform worse than low-allocation ones in the following years
what is the projection bias example with snacks
Read and Van Leeuwen (1998), office workers choose healthy snack or unhealthy, snack delivered a week later (in late afternoon), Group 1 asked when plausibly hungry (late afternoon) -> 78% choose unhealthy snack, Group 2 asked when plausibly satiated (after lunch) -> 42% choose unhealthy snack, projection bias, think current state of world will continue
setting and motivation behind experiment on clothes ordering
Conlin et al. (2007), look at catalogue orders of cold weather items on cold days, Obviously: high utility of cold-weather clothes for today, Projection bias: forecast high utility into future, order ‘too much’ if agents were rational, there shouldn’t be any correlation b/w temperature fluctuations and orders/returns
how is there projection bias in the clothes ordering experiment
Conlin et al. 2007, high utility of cold weather clothes for today, projection bias: forecast high utility into future
what is the data from the experiment on clothes ordering (projection bias)
Conlin et al. 2007, data from US company selling outdoor apparel and gear, 12m items 1995-1999, date of order and shipping, address & returned yes/no, weather data matched to shipping address
what are some further results / limitations of the experiment on clothes ordering (projection bias)
Conlin et al. 2007, similar estimates with HH fixed effects (->selection bias, don’t get this 40/41), no effect on items unrelated to weather, no such effect for snowfall and rainfall, structural model: estimates of projection bias α≈0.3-0.4 people factor in utility of item today by degree of 30-40% (not necessarily take model as true)
what are limitations of experiment on clothes ordering (projection bias)
cannot distinguish between mispredicting utility or weather, weak support for role of γR, structural estimates suggest different α for each single ite, alpha should be the same for every item
what is the motivation behind the experiment replicating Samuelson’s observations from coin flip and what paper
Benartzi and Thaler (1999) vary presentation format: words vs possible outcomes, problem with MLA (myopic loss aversion): people don’t integrate problems, do people make better decisions if integration done for them?
what is the setting of the replication of Samuelson’s coin flip
Benartzi and Thaler (1999), subjects asked if they would accept: Samuelson’s bet, Samuelson’s bet 100 times, Samuelson’s bet 100 times (outcome distribution), 3 subject groups: UGs, random people, MBAs
what were results of replication of coin flip
Benartzi and Thaler (1999), people prefer 1 over 2… apart from MBAs, all groups prefer 3 over 2, 1) samuelson’s bet, 2) samuelson’s bet 100 times, 3) samuelson’s bet 100 times (outcome distribution)
what is the repeated gambles experiment
Benartzi and Thaler (1999), three equal gambles with equal exp. value (=6) and distribution: 1)90% +0.10$; 10% -0.50$; played 150 times 2)50% +0.25$; 50% -0.15$; played 120 times 3) 10% +0.75$; 90% -0.01$; played 90 times
what is the results of the repeated gambles experiment
as the possible amount to lose decreases –> % subjects accepting gamble increases, BUT: far more subjects choose gamble 1) if only shown distribution, Prospect theory ->subjects systematically overestimate the prob. of losing money in positive expected value gambles, seeing distribution can make gamble more attractive. 1)90% +0.10$; 10% -0.50$; played 150 times 2)50% +0.25$; 50% -0.15$; played 120 times 3) 10% +0.75$; 90% -0.01$; played 90 times
background of pension plan experiment (repeated investments)
Benartzi and Thaler (1999), pension plans: workers contribute % of monthly income to retirement account, investment choice by worker, mots popular choices: low risk assets small returns
setting of pension plan experiment (repeated investments)
Benartzi and Thaler (1999), survey recent employees at US university and ask whether they prefer Fund A (high risk/return) or B (low risk/return), ReturnA-ReturnB set to historic equity premium, show distribution of 1) 1-year returns, 2) 30-year returns, 3) retirement income after 30-yr of paying in