Toward a behavioral theory of real options: Noisy signals, bias, and learning

• Author: Hart E. Posen,John S. Chen,Michael J. Leiblein
• Published date: 01 April 2018
• DOI: http://doi.org/10.1002/smj.2757
• Date: 01 April 2018

RESEARCH ARTICLE

Toward a behavioral theory of real options: Noisy

signals, bias, and learning

Hart E. Posen

1

| Michael J. Leiblein

2

| John S. Chen

3

1

School of Business, University of Wisconsin-

Madison, Madison, Wisconsin

2

Fisher College of Business, The Ohio State

University, Columbus, Ohio

3

Warrington College of Business Administration,

University of Florida, Gainesville, Florida

Correspondence

Michael Leiblein, Department of Management

and Human Resources, 2100 Neil Avenue, Fisher

College of Business, Ohio State University,

Columbus, OH 43220.

Email: leiblein.1@osu.edu

Research Summary: We develop a behavioral theory of

real options that relaxes the informational and behavioral

assumptions underlying applications of financial options

theory to real assets. To do so, we augment real option

theory’s focus on uncertain future asset values (prospec-

tive uncertainty) with feedback learning theory that con-

siders uncertain current asset values (contemporaneous

uncertainty). This enables us to incorporate behavioral

bias in the feedback learning process underlying the

option execution/termination decision. The resulting com-

putational model suggests that firms that inappropriately

account for contemporaneous uncertainty and are subject

to learning biases may experience substantial downside

risk in undertaking real options. Moreover, contrary to

the standard option result, greater uncertainty may

decrease option value, making commitment to an invest-

ment path more effective than remaining flexible.

Managerial Summary: Executives recognize the need to

make uncertain investments to grow their business while

mitigating downside risk. The analogy between financial

options and real corporate investments provides an

appealing method to consider the practical challenge of

such investment decisions. Unfortunately, the “real

options”analogy seems to break down in practice. We

identify how a second form of uncertainty confounds real

options intuition, leading managers to overestimate the

value of uncertain investments. We present a behavioral

real options model that accounts for both forms of uncer-

tainty and suggest how uncertainty interacts with behav-

ioral bias in the option execution/termination decision.

Our model facilitates assessment of the conditions under

which investments in uncertain opportunities are usefully

The authors contributed equally to the study and are listed in the order determined by a random draw.

Received: 15 June 2016 Revised: 27 November 2017 Accepted: 6 December 2017 Published on: 22 February 2018

DOI: 10.1002/smj.2757

1112 Copyright © 2017 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/smj Strat Mgmt J. 2018;39:1112–1138.

considered as real options, and provides a means to eval-

uate their attractiveness.

KEYWORDS

Bayesian learning, behavioral bias, computational

model, decision-making under uncertainty, experiential

learning, real option logic, sequential decision-making,

strategic decision-making

1|INTRODUCTION

Strategy scholars conventionally think of the value of a real option as increasing with “volatility”—

uncertainty about the prospective (future) value of an asset. Increasingly, however, it has been rec-

ognized that under some conditions this association does not seem to hold in practice. We suggest

that for real, as opposed to financial, options there is a second form of volatility—uncertainty about

the contemporaneous (current) value of an asset. This article develops a behavioral real options the-

ory that embodies both types of uncertainty and, in doing so, considers the role of behavioral bias in

learning that underlies the option execution/termination decision. The model suggests important

boundary conditions to the value of flexibility associated with real options. In particular, contrary to

the standard real option result, under some conditions greater uncertainty makes a commitment to an

investment path more effective than remaining flexible. Moreover, firms that do not account for the

existence of contemporaneous uncertainty may experience substantial downside risk in undertaking

real options.

The promise and challenge of applying financial option theory to real assets (e.g., investment pro-

jects, technologies, physical assets) has been recognized by both financial (e.g., Dixit & Pindyck,

1994; McDonald & Siegel, 1986) and managerial scholars (e.g., Adner & Levinthal, 2004; Barnett,

2008; Kogut & Kulatilaka, 2004; McGrath & Nerkar, 2004; Reuer & Leiblein, 2000; Sakhartov &

Folta, 2014). One important challenge concerns the limited availability of information on current real

asset values. As Bowman and Moskowitz (2001, p. 775) note, real assets “do not have a readily

observable stock price ... [and this makes] it difficult to draw conclusions about the appropriate course

of action at the time of the option’s expiration.”The implications for practical applications of real

options are substantial. Coff and Laverty (2001, p. 73) recognize that “managers may erroneously

exercise options …or drop options that would lead to a competitive advantage.”Other evidence sug-

gests that manag ers both “overestimate the value of uncertain projects, encouraging companies to

over-invest in them”when applying real options (van Putten & MacMillan, 2004, p. 1) and underuti-

lize real options decision tools (Lander & Pinches, 1998; Teach, 2003). Van Putten and MacMillan

(2004) argue that when uncertainty about the future value of assets is very high, real options are in a

“flee zone”and should be avoided. These observations suggest that unrecognized boundary conditions

may limit the applicability of the current formulation of real options theory.

To develop a behavioral real option theory, it is useful to consider why the canonical real

options valuation framework is, to borrow the language of Thaler and Mullainathan (2008), “unbe-

havioral.”Herbert Simon (1990) suggests a behavioral theory must incorporate two dimensions of

reality analogous to the two blades of a pair of scissors. In Simon’s“scissors”analogy, one blade

POSEN ET AL.1113