ACADEMIA Letters
Does Miscalibration of Financial Risk Tolerance Describe
Portfolio Holdings?
John Grable, The University of Georgia
Eun Jin Kwak, The University of Georgia
Wookjae Heo, Purdue University
Introduction
The idea that sometimes investors mis-calibrate their willingness to take risk has been exam-
ined intermittently over the past two decades. The concept of miscalibration is closely asso-
ciated with the notion of overconfdence (Nosic and Weber, 2010) and is commonly thought
of as an over- or underestimation bias (Grable et al., 2009). A study by Hallahan et al. (2004)
illustrates how miscalibration is generally estimated. They asked a sample of investors to
complete a test designed to uncover a test-taker’s willingness to engage in fnancial risk-taking
behavior. They also asked the test-takers to estimate their risk tolerance using a stated risk-
preference measure. When they compared the two, Hallahan and associates noted that 73%
of test-takers underestimated their risk tolerance, 23% overestimated their risk tolerance, and
4% were accurate (i.e., risk-tolerance scores were calibrated). Their research showed that
miscalibration is likely a widespread occurrence.
Risk-tolerance miscalibration can have serious ramifcations for those engaged in fnancial
decision-making. Odean (1999) and Pan and Statman (2012) noted overconfdent investors
take more risk when allocating investment assets. When someone is overconfdent, the worry
is that their portfolio allocation will exceed their true willingness to take a risk, increasing
the possibility that they will sell at a loss when portfolio volatility exceeds their comfort or
preference level. By extension, those who exhibit underconfdence are more likely to invest
Academia Letters, May 2022
Corresponding Author: John Grable, grable@uga.edu
Citation: Grable, J., Kwak, E.J., Heo, W. (2022). Does Miscalibration of Financial Risk Tolerance Describe
Portfolio Holdings? Academia Letters, Article 5384.
1
©2022 by the authors — Open Access — Distributed under CC BY 4.0