Accounting Theory

Behavioral Accounting

vs.

Behavioral Finance

St. Gallen, Januar 2009

---

1

INTRODUCTION AND OBJECTIVES

An economic theory which is not incorporating human behavior is not imaginable. For

reasons of simplification economic models traditionally use the concept of a rational acting

market participant. In order to face the inadequateness of this abstraction behavioral economic

science reject the assumption of the

homo economicus

and adds various findings from

supporting disciplines as psychology, sociology, and organizational theory. While the

exploration of human behavior in finance theory has a long tradition, research in the area of

psychological effects in accounting started not earlier than the mid of last century. The main

intention of modern financial reporting is the supply of useful information for actual and

potential investors within their decision-making process. As information processing of agents

on the market for equity is part of finance theory, this is the meeting point of the two

disciplines. The intention of this paper is to identify overlapping contents of behavioral

research in finance and accounting. For clarification selected studies from Behavioral Finance

Research (BFR) and Behavioral Accounting Research (BAR) literature will be presented and

comparatively analyzed. In addition varying fields of research of both schools which are not

related with each other were outlined.

---

2

BEHAVIORAL FINANCE RESEARCH (BFR)

Classic economic theory as developed by economics like ADAM SMITH, DAVID

RICARDO and JOHN STUART MILL is based on the idea of the

homo economicus

.1 The

homo economicus seeks to receive the highest possible well-being for himself given available

information about opportunities and other constraints, on his ability to achieve his

predetermined goals.2 Homo economicus is seen as "rational" in the sense that well-being as

defined by the utility function is optimized given perceived opportunities (with the greatest

extent and least possible costs).3 Furthermore, the individual is able to process all relevant

information in a appropriate manner.

Taking this into finance theory, the market for equity should be predictable and driven by

public information and the consequent rational reaction of its agents. But since organized

capital markets exist prices seem to be randomly settled and unpredictable. In 1900 LOUIS

BACHELIER stated that stock prices reflect reactions to information coming to the market in

random fashion, so they are no more predictable than "the walking pattern of a drunken

person".4 This is known as the "random walk" in finance theory. Also JOHN MAYNARD

KEYNES recognized in 1936 that the price on the market reflect the expectations of the future

trend; regardless if the expectations are based on wrong or correct assumptions. So the

question was raised if the market behaves irrational. The nobel-prize-winning answer was: no!

PAUL SAMUELSON (1965) argued that the rationality of the investors is the reason for the

fluctuation of the prices on the capital market. Prices on the capital market are the result of the

cumulated amount of discounted dividends. Random fluctuation of prices are due to random

variances of earnings and their expected trends. Neoclassical theory was established and

influenced EUGENE FAMA in 1970 to his Efficient Market Theory (EMT). The bottom line

of this theory is that there is perfect information in the stock market. All data are fully and

immediately reflected in a stock price. This means that whatever information is available

about a stock to one investor is available to all investors. Since everyone has the same

information about a stock, the price of a stock should reflect the knowledge and expectations

of all investors. No group has access to information that would enable it to earn superior risk-

adjusted returns (Arbitrage).

1 The term economic man was introduced in JOHN KELLS INGRAMs "A Hiostory of Political Economy"

in 1888; VILFREDO PARETO used the latin term homo economicus first in 1906.

2 Cf. THALER (2000), p. 134.

3 Cf. FRANZ (2004), p. 83.

4 BACHELIER (1964), p. 256.

---

3

Based on MARKOWITZ`s portfolio theory a rational investor calculates the asset price with

the Capital Asset Pricing Model (CAPM), which means that the price is the result of the linear

dependence of the expected return and the risk.5 Therefore irrational behavior could occur

either in processing information regarding the future performance or during the risk

assessment. These are the points where Behavioral Finance joins the game. Empirical findings

describing biases in behavior of investors regarding their risk assessment led

KAHNEMAN/TVERSKY (1979) to their Prospect Theory, which ennobled Behavioral

Finance to a systematic and accepted science. Moreover anomalies regarding the information

processing of market participants were identified and challenged the EMT. The following

event studies show some examples for irrational incorporation of information into stock

prices:

Example 1

"The Sorry Tale of the Cure for Cancer"

HUBERMAN/REGEV (2001) detected the informational imperfection of markets by showing

that even a "non-event" had a major market impact. On Sunday 3 May 1998, the New York

Times carried an article on a new cancer drug being researched by EntreMed. EntreMed`s

stock price rose from $12 on previous Friday up to $85 at the beginning of the week. The

stock subsequently fell back during trading on Monday 4 May to close a very respectable $52.

Three weeks later the stock price was still above $30. Now it could be argued that this was the

efficient market incorporating of new information. But that was exactly the problem. This

wasn`t new information. The potential breakthrough had been reported no less than five

month earlier in numerous popular newspapers including the New York Times. Enthusiastic

public attention induced a permanent rise in EntreMed`s share price, even though no

genuinely new information had been given to the market.

5 Cf. SHARPE (1964), LINTNER (1965), MOSSIN (1966).

---

4

Example 2

"Massively Confused Investors Making Conspicuously Ignorant Choices"

RASHES (2001) examines the co-movements of stocks with similar ticker symbols. Ticker

symbol confusion is found to lead to interesting price movements in stocks for which there is

no new information introduced to the market. MCI Communication (MCIC) is a large

telecommunication company in the USA. From the end of 1996 through 1997, it was engaged

in merger negotiations with Worldcom. Until the acquisition, it traded on NASDAQ with the

ticker MCIC. Massmutual Corporate Investors is a closed end fund that trades on NYSE

under the ticker MCI. The fund invests most of its assets in long-term corporate debt and

converts. In spite of the fact, that MCI and MCIC have nothing to do with each other, there is

an unusual degree of co-movement between the two stocks. Between November 1 1996 and

November 24 1997, of the 24 days during which 10,000 or more shares of MCI changed

hands, 18 are days on or immediately prior to significant announcements relating to MCIC

merger plans. The mistake done by investors is obvious.RASHES goes on to document

another five cases in which ticker confusion has caused some bizarre market movements.

Example 3

"Dot.coms and Name Changes"

COOPER et. al. (2001) studied firms that changed their names to include some mention of

dot.com. Across their sample of firms drawn from AMEX, NYSE and NASDAQ between

June 1 1998 and July 31 1999, COOPER et al. find that in the five days surrounding the

announcement of a name change to incorporate dot.com into corporate title, an abnormal

return of 53% is generated. Even in the + 60 days event window an abnormal return of 23% is

generated. But perhaps the most bizarrely finding was that firms whose core business is

unrelated to the internet, but who nevertheless decide to amend their name to include a

dot.com reference, generate a 23% abnormal return in the five days surrounding the

announcement, and even worse a 140% abnormal return in the + 60 event window.

These examples show some biases in behavior of investors during their information

processing for stock valuation. Other directions in Behavioral Finance investigate anomalies

in finance related behavior in the areas of Asset Allocation, Portfolio Construction and Risk

Management, and Corporate Finance (see figure 1).6

6 The systematization of fields in BFR follows MONTIER (2007).

---

5

Figure 1

Schools of BFR

Behavioral Finance

Stock

Asset

Portfolio

Corporate

Valuation

Allocation

Construction

Finance

JDM by Investors

JDM by Investors and

JDM by Investors

JDM by Managers

and Analysts

Analysts

and Analysts

regarding risk

management

JDM = Judgment and Decision Making

BFR in the area of

stock valuation

deals with the behavior of Investors and Analysts

regarding their information processing related to their investment decisions. As showed above

biases inconsistence with the EMT, e.g. imperfect information, are investigated. In the field of

stock valuation the assumptions of the CAPM are also doubt and experimental and empirical

evidences towards the violation of these assumptions exist. Most of the studies deal with the

discount factor for the cost of capital in the valuation process.7 As a reaction to these findings

FAMA and FRENCH developed the three-factor model (FF3) to improve the CAPM in 1993.

BFR in the field of

asset allocation

investigates the rational distribution of investments in

stocks and bonds. The equity risk premium stands in the focus of this research area.8

Furthermore, the well-know bias that initial public offerings (IPOs) or seasoned equity

offerings (SEOs) underperform the broader market, is tried to be explained by behavioral

approaches in this field.9

Biases in the modern

portfolio theory

are mostly due to anomalies in the behavior of

investors and analysts regarding their risk assessment or in technical terms their estimation of

correlation and covariance. As an seminal work DANIEL and TITMAN (1998) showed that a

characteristic model performs better in predicting expected returns than the classic covariance

7 Cf. e.g. LINTNER (1965), HORWITZ et. al (2000).

8 Cf. BARBERIS et al. (2000), CLAUS/THOMAS (2000), WELCH (1999).

9 Cf. NELSON (1999), BAKER/WURGLER (2000), SCHILL (2000).

---