finance discussion writing
Alok Kumar Department of Finance Miami Business School
akumar@miami.edu; 305‐284‐1882
FIN 686 Psychology of Financial Markets and Financial Decision Making
LECTURE NOTES 8
Behavioral Corporate Finance: Managerial Biases
2
In this lecture, we will transition from the MT&C framework, which examines the implications of irrational investors for firms, and take the alternative perspective. That is, we will examine the implications if managers act irrationally.
Behavioral Corporate Finance The assumptions underlying neoclassical corporate finance
(mgr.’s and investors have rational beliefs and expectations) may not always hold…
For instance, investors’ behaviors may not be fully rational Firms/CEO’s could recognize this and respond accordingly This is the Market Timing & Catering framework
Or, managers may not always behave rationally This could lead to suboptimal policies within firms
3
Behavioral Corporate Finance What are the effects on firms if managers are not fully
rational? This is known as the Managerial Biases approach
This approach assumes that managers are humans and are subject to biases
Ultimately, biases affect the decision making of CEO’s and can impact firm policies, such as investment and financing policies
But, this approach also depends on: Investor rationality but limited governance mechanisms Or investor irrationality (not paying attention)
4
BCF: Normative Implications But first, why should be care if managers have biases?
Implications of the Managerial Biases approach… The primary source of irrationality is on the manager side This suggests that managers should be pushed
(obligated) to respond to market price signals In turn, this suggests that governance measures are
important
5
BCF: Managerial Biases
6
We will now look deeper into the Managerial Biases literature to gain insights into the dynamics between CEO’s, firms, and investors.
BCF: Managerial Biases
7
The Managerial Biases is a growing literature which has produced from valuable insights into the dynamics between firms, markets, and investors
There is a key condition underpinning this framework 1. For less‐than‐fully‐rational managers to have an impact,
corporate governance must be limited in its ability to constrain them into making rational decisions
Limited Governance
8
Limited governance is analogous to the requirement of limits to arbitrage when we examined the impacts of investor irrationality
But, is assuming limited governance reasonable? We will investigate this over the next few slides
Limited Governance
9
An important component of corporate finance is the role of corporate governance on CEO’s decisions
There are important questions that are still open… For instance, how strong should governance mechanisms
be? What are the best/most efficient mechanisms? Are existing governance tools effective? How can we precisely measure the effectiveness of
various governance mechanisms?
Limited Governance
10
Various literatures have developed around the preceding questions
We will review a few of the highlights in order to establish a foundation which will then aid us in how we think about the managerial biases
Specifically, we need to determine if: There are limits to governance, If the effectiveness of governance varies across
firms/time periods, If it matters for firms (operating performance) and
shareholders (valuations/stock returns)
Limited Governance
11
There are several governance tools/mechanisms used to monitor CEO’s actions with the aim of protecting shareholders Stock market – declining share prices Takeovers – firms or activists shareholders police CEO’s Boards of Directors CEO compensation contracts – align CEO and shareholder
interests
But, are these effective?
Limited Governance
12
While these governance tools may assist in policing CEO’s, they are not always completely effective Takeover battles and proxy fights are blunt tools Antitakeover provisions (staggered Boards, poison
pills, etc..) Boards of Directors may be pawns of management or
have their own biases Incentivized compensation contracts may not be effective
if an irrational manager incorrectly thinks that he is already acting in a way that best for shareholders (i.e., maximizing value)
Limited Governance
13
Systematic studies of the impacts of governance mechanism have also been conducted…
A frequently used method of measuring the influence of corporate governance is the Governance Index developed by Gompers, Ishii, & Metrick (2003) Combine large set of governance provisions into an index
which proxies for the strength of shareholder rights Then, study the relationship between the index of
shareholder rights and corporate performance
Limited Governance
14
Investor Responsibility Research Center publishes 24 distinct corporate governance provisions for about 1,500 firms since 1990 22 firm‐level provisions 2 state‐level laws (actually 6 but 4 are analogous to firm‐
level provisions)
Limited Governance
15
Limited Governance
16
Limited Governance
17
G‐Index: For each firm: Add one point for every provision that reduces
shareholder rights Firms in the highest decile have highest mgt. power
(Dictatorship Portfolio) Firms in lowest decile have strongest shareholder power
(Democracy Portfolio)
Limited Governance
18
Limited Governance
19
Limited Governance
20
Does corporate governance matter? Check if G‐Index is related to share returns… Democracy Portfolio outperforms Dictatorship Portfolio
by 8.5% per year So shareholder wealth is impacted by corporate
governance!
Is G‐Index related to firm operating performance? Yes! Firms with weak shareholder rights are less
profitable and have lower sales growth than firms in the same industry
1 point increase in G‐Index associated with 11.5 pct.point decline in Tobin’s Q
Limited Governance
21
More recent evidence also suggests that corporate governance can vary across firms and has important implications for shareholders
Harford, Mansi, & Maxwell (2007): CEO’s of firms with weak shareholder rights increase CAPEX and acquisitions and ultimately have lower profitability and lower valuations
Sokolyk (2011) combination of staggered Board and a Poison Pill reduces the probability of a takeover (i.e., limits the effectiveness of mergers/activist investing)
Limited Governance
22
Bebchuk, Cohen, & Ferrell (2008) develop an Entrenchment Index based on: Staggered boards, Limits to shareholder bylaw amendments, Poison pills, Golden parachutes, Supermajority requirements for merges or charter
amendments
All the above limit the extent to which a majority of shareholders can impose their will on management
Limited Governance
23
E‐Index scores range from 0 – 6 with 6 being the highest entrenchment
Increases in the E‐Index are associated with significant reductions in firm valuations and large negative abnormal returns from 1990 – 2003
Limited Governance
24
Strategy of buying firms with low E‐Index scores and, simultaneously, selling short firms with high E‐Index scores yields substantial abnormal returns
About 13% abnormal return per year.
Limited Governance
25
Overall, the evidence suggests that governance mechanisms are important Higher governance leads to higher shareholder wealth
But, the effectiveness of governance mechanisms vary across firms can be influenced by CEO’s actions
These ultimately suggest that the impact of CEO biases can manifest in firms and could be economically large
BCF: Managerial Biases
26
We will now look begin examining the role of biases on CEO’s decisions, firm outcomes, and market dynamics
To start with, we will set the stage with an example…
Scott McNealy was the CEO of Sun Microsystems. He was an extreme risk‐taker and was often portrayed as smart and optimistic but also brash and combative
A Typical CEO?
Investors were very optimistic about his firm. For example, at the peak of the Internet bubble, the price‐to‐earnings ratio of Sun was over 100!
The extreme market valuations could have contributed to his psychological biases.
27
A Typical CEO? His decisions were often biased. For example: He was slow to cut costs during times of recession because he predicted that the recession would be short even though all the data pointed in the opposite direction.
Even when industry leaders such as Cisco laid off employees, McNealy did not take any corrective action (perhaps due to his excessive optimism).
Sun held large amounts of cash and he used them to increase R&D expenditures and acquire Cobalt (perhaps due to his overconfidence).
28
A Typical CEO? His decisions were often biased. For example: Sun decided not to use Intel chips even though economically that would have been a better decision. He felt that Sun’s chip design group wound be able to close the gap (perhaps due to illusion of control).
During the most successful years, the earnings of Sun grew by almost 50% per year. In later years, the earnings were revised downward but the 50% growth rate acted as an anchor (anchor and adjustment phenomenon).
McNealy often paid more attention to issues that were more salient, even when they were less relevant for firm profitability.
29
A Typical CEO? While McNealy’s behaviour may appear unique and non‐
typical, other top corporate managers are likely to exhibit these types of behavioural biases too.
In general, corporate managers may not act in the best interests of their shareholders, not necessarily due to agency issues, but due to their own personal psychological biases.
We will now investigate the literature focusing on psychological biases among firm managers
30
Bounded Rationality: Foundation of Biases
31
As with individual investors, CEO’s may also operate as boundedly‐rational agents (Simon 1955)
That is, some type of cognitive or information‐gathering cost prevents managers from making fully optimal decisions
Bounded‐rational managers then cope with the complexity by using rules of thumb to achieve an acceptable level of performance while avoiding significantly negative outcomes
Mgr. Biases: Overconfidence
32
There are good reasons to think that overconfidence may play a role among CEO’s decisions… The average individual displays at least some level of
overconfidence (82% of students place themselves in top 30% of safest drivers (Svenson 1981))
Managerial decisions tend to be complex, a setting where overconfidence is one’s abilities is likely to be pronounced
Managerial decisions also tend to be idiosyncratic, which reduces the potential for de‐biasing through learning
Mgr. Biases: Overconfidence
33
There are good reasons to think that overconfidence may play a role among CEO’s decisions… (continued) Since overconfidence naturally leads to more risk‐taking,
even if there is no overconfidence in the population of potentialmanagers, those that are overconfident are more likely to perform extremely well (and extremely poorly), placing them disproportionately in the ranks of upper (and former) management
We will now look to existing studies to gain insights into the role of managerial overconfidence on firms
Mgr. Biases: Overconfidence
34
Overconfidence may impact the firms across the life‐cycle spectrum
For instance, do you think a rational individual would devote monetary & human capital resources to starting a firm given: About 50 percent of small businesses fail in the first four
years (BLS) 77% of small firms rely on the entrepreneurs’ or their
families’ wealth for initial funding
Mgr. Biases: Overconfidence
35
Mgr. Biases: Overconfidence
36
Moreover, it is interesting individuals start business while the leading causes of small business failure are: Incompetence: 46 percent; Unbalanced experience or lack of managerial experience:
30 percent; Catch‐all category (neglect, fraud, and disaster): 13
percent; Lack of experiences in line of goods or services: 11
percent.
Mgr. Biases: Overconfidence
37
Yet, entrepreneurs are very confident that they will succeed! Three out of four millennial business owners expect to see
their businesses improve over the next year 68% of entrepreneurs think their startup is more likely to
succeed than comparable enterprises (Cooper et al. 1998) 33% of which believe their success is all but guaranteed At the initial starting point, only 5% of entrepreneurs
expect to encounter difficulty during their endeavor (Landier and Thesmar 2009)
Mgr. Biases: Overconfidence
38
Experimental evidence directly links overconfidence to starting a business (Camerer and Lovall 1999) Subjects who enter think the total profit earned by all
entrants will be negative, but their own profit will be positive
Mgr. Biases: Overconfidence
39
Overconfidence may also influence investment decisions of more mature firms
For instance, forecasts of construction or development costs tend to be significantly underestimated by managers in mature firms (Merrow et al. 1981; Statmand and Tyebjee 1985) Holds across a range of firms: hardware, drugs, chemicals,
energy
Mgr. Biases: Overconfidence
40
Ben‐David, Graham, and Harvey (2010) test if CFO’s are overconfident through survey‐based measures Ask financial executives to estimate the mean and variance
of their firm’s past stock returns CFO’s are significantly overconfident about how their firms
performed This overconfidence spills‐over to affect their firms’
investments More overconfidence is associated with higher levels
of investment
Mgr. Biases: Overconfidence
41
Overconfidence can be difficult to measure through external observation Previous studies have principally relied on surveys
One way empirical measure is to observe if CEO’s voluntarily hold in‐the‐money stock options in his/her own firm (Malmendier and Tate 2005) Since the CEO’s human capital is already so exposed to firm‐
specific risk, voluntary holding in‐the‐money options is a strong vote of confidence in the firm’s prospects on his/her abilities to ensure such prospects manifest
Mgr. Biases: Overconfidence
42
Using in‐the‐money options proxy, Malmendier and Tate show that firms’ investments are sensitive to CEO’s level of overconfidence As overconfidence increases, investment increases
Mgr. Biases: Overconfidence
43
Overconfidence is also likely to impact firms’ investments in R&D
This is because the payoff to R&D is inherently uncertain and, thus, greater self‐belief in one’s ability to succeed may lead to higher investment
Hirshleifer, Low, and Teoh (2010) use the options proxy and media characterizations of CEO’s to measure overconfidence Overconfident CEO’s invest more in R&D and this
translates into higher patent and patent citation counts