Learning to Wait: A Laboratory Investigation

Abstract

Human subjects decide when to sink a fixed cost "C" to seize an irreversible investment opportunity whose value "V" is governed by Brownian motion. The optimal policy is to invest when "V" first crosses a threshold "V"* = (1 + "w"*)"C", where the wait option premium "w"* depends on drift, volatility, and expiration hazard parameters. Subjects in the Low "w"* treatment on average invest at values quite close to optimum. Subjects in the two Medium and the High "w"* treatments invested at values below optimum, but with the predicted ordering, and values approached the optimum by the last block of 20 periods. Copyright Copyright © 2009 The Review of Economic Studies Limited.

Holding on for too long? An experimental study on inertia in entrepreneurs’ and non-entrepreneurs’ disinvestment choices

“…As mentioned in the introduction, the experimental investigation of the real options approach is still at an early stage: Closest to the spirit of our investigation is a recent paper by Oprea et al (2009) that analyses whether individual behavior in an investment setting might, by learning, approximate the optimal exercise frontier for available options. Further, Rauchs and Willinger (1996), focusing on how increased expected information affects subjects' choices, provide evidence for an irreversibility effect, while Sirmans and Yavas (2005) try to elicit, in a very simple setting, subjective valuations for an option by asking the participants to submit a bid for it.…”

That's How We Roll: An Experiment on Rollover Risk

“…The first one corresponds to continuous-time experiments, the second to "timing experiments", with a special emphasis on the experimental bank-runs literature. Continuous time experiments started years ago, with Friedman and Cheung (2009) and Morgan and Brunnermeier (2010) (whose working papers appeared around 2003/04), but it has not been until recently that this experimental technique has taken off with Oprea et al (2009) and Anderson et al (2010) looking into strategic investment decisions, Oprea et al (2011) studying the evolutionary equilibrium of the hawk and dove game, Friedman and Oprea (2012) experimenting with the effects of response delay in a repeated prisoners dilemma game, and Rabanal (2012) looking at mortgage default timing. While none of these papers directly address any of the questions of our paper, they are a good reference for the methodological design of our experiment.…”

“…Following Anderson et al (2009), each tick will be 1/5 of a second (i.e., 200 milliseconds). Each of the 60 rounds has a random end which is governed by a Poisson process, and has an expected length of 150 ticks (30 seconds), at which point the long-term investment matures and the firm ceases to exist.…”

“…In HX, the value of the long-term investment (y t ) follows a geometric Brownian motion. Given the discrete nature of computer internal clocks, we will need to discretize this Brownian motion, and to do so we will use the procedure described in Anderson et al (2009).…”

“…Standard errors in parentheses * p < 0.10, * * p < 0.05, * * * p < 0.01 censored beyond the latest observed stopping value (in total 38 observations). Then, following Anderson et al (2009), we eliminate all observations in those sessions where the minimum value of the firm was not low enough. In our case this means all observations for rounds where the minimum value was (strictly) greater than $0.9 (in total 3020 observations).…”