Abstract
This paper investigates behavioural finance principles and in particular the managerial bias, overconfidence. During the dotcom bubble, 1998-2001, it has become clear overconfidence has played a significant role in the burst of the bubble. Overconfident and, even, rational agents were riding the bubble and invested highly in technology stock. MANAGEMENT CONTROL MEASURES TO BE TAKEN.
‘
1. Introduction
“Irrational lenders come and go — mostly they go!” – John Stumpf, chairman and CEO of Wells Fargo
Primarily, prior theories on finance solely assumed that all agents act rationally and have rational expectations. On the other hand, behavioural finance deals with the psychological side of finance, in contradiction to the Efficient Market Hypothesis (EMH) and Modern Portfolio Theory, this concept investigates the cognitive factors that impact the decision making of individuals, groups and organizations, and has been taken into account the past few decades (Ricciardi & Simon, 2000).The Efficient Market Hypothesis states that every market is efficient, all investors act according to all available information and all investors make rational decisions (Fama, 1970). A condition describing that all prices reflect the intrinsic value. According to this theory, all stock prices look like ‘random walks’, indicating that every time there is a change there has been new information available on the stock (Shiller, From efficient markets theory to behavioural finance, 2003). Nevertheless, in daily life, speculation and bubbles do exist due to information asymmetry and irrational decision making by individuals. Fromlet (2001) describes bubbles as a temporarily inexplicable observation to describe abnormalities (Fromlet, 2001). An example of a speculative bubble is the dotcom bubble, also referred to as the internet bubble or the technology bubble, which took place during 1998-2001 when internet companies saw their own capital growing with spectacular rates and finally collapsing in 2000. Cognitive factors on investor decision making are regarded as one of the causes of the dotcom bubble, in particular managerial biases as overconfidence.
This research project will investigate the knowledge professional investment companies have on irrational decision-making in the field of management control, especially when expecting a financial bubble. The focus of this research lays on management control and the measures needed to be taken in order to avoid irrational decision making. This will be done by the knowledge we have now from previous events and, in particular, the dotcom bubble. The paper is divided into three research questions to give an answer to the research question.
1. What are the principles of behavioural finance and how do they relate to managerial decision making?
2. What are the lessons learned from the dotcom bubble regarding overconfidence?
3. What are possible management control measures to avoid irrational decision-making by CEO/CFO when companies are in the midst of a stock price ‘bubble’?
The paper will combine behavioural finance to management control and internal control by CEOs and CFOs. Up to this point, this has not been researched before and will therefore add to already existing literature. It can also be used as a reference for management accountants to act upon.
First, the principles of behavioural finance will be explained, together with the relationship to investor decision making. Second, an overview of the dotcom bubble will be given to describe the effects of irrational decision making. Third, possible management control measures will be given to avoid irrational decision making by investors. At last, a short conclusion will be given to summarize the research.
2. What are the principles of behavioural finance and how do they relate to managerial decision making?
The past few centuries it has become more clear that finance cannot only be explained by abstract theories as the Efficient Market Hypothesis, but that psychological factors have to be taken into account to explain the behaviour of individuals concerning finance. The psychologists Tversky and Kahneman introduced several concepts that belong to the field of behavioural finance, such as the term availability heuristic stated as ‘a judgmental heuristic in which a person evaluates the frequency of classes or the probability of events by availability, i.e. by the ease with which relevant instances come to mind’ (Kahneman & Tversky, Availability: A heuristic for judging frequency and probability, 1973). Heuristics are often described as mental shortcuts in the decision process (Thoma, White, Panigrahi, Strowger, & Anderson, 2015). For instance, the daily tasks of professional investors, such as doing risk assessments and pricing assets, may be prone to heuristics as well. But does it mean that this results in irrational decision making by investors, following their intuition? In the research of Thoma et al. it is questioned whether the behaviour of traders, bankers and non-financial experts is based on ‘good thinking’, analytical thinking, or ‘gut feeling’, the more intuitive one. There is no difference between the three reference groups on good thinking and all are acting sufficiently rational. Kahneman states that ‘it is self-evident that people are neither fully rational nor completely selfish, and that their tastes are anything but stable’ (Kahneman D. , 2011). A framework often used in the decision making and judgement theory, in which a distinction is made between intuition and rationality, is the dual systems framework. In this framework humans show two psychological systems used to make decisions; system 1, the fast and effortless decision process which is seen to be as the intuitive judgement; and system 2 is the slow and effortful, rules based decision process which is more seen to be as ‘rational (Thoma, White, Panigrahi, Strowger, & Anderson, 2015).
Several mechanisms have been developed for the decision making process in prior literature. A mechanism that can be used for decision making under uncertainty is the expected utility theory, described by Tversky and Kahneman (1979) as a normative model that describes economic behaviour based upon the axioms of the expected utility theory assuming all axioms are being met (Kahneman & Tversky, 1979). In their research paper Prospect Theory: an Analysis of Decision under Risk they introduce a new concept, called prospect theory, after criticising the expected utility theory for being inadequate. The prospect theory instead gives more insight on economic behaviour than only on profit maximization and is being proposed as a model of choice basing decisions on gains and losses with a certain reference point (Kahneman & Tversky, 1979). These decisions can be affected by the certainty of the decision problem but also depend on how a decision is framed. Described as the conception of all aspects of the decision problems the individual makes when making a decision and is determined by the personal norms and values of the individual but also by the way a question is framed (Kahneman & Tversky, 1981). A concept adequate to prospect theory is loss aversion and refers to the tendency to act differently to gains and losses while expected utility is held constant. People tend to be risk averse for gains and risk loving for losses and therefore prefer to avoid losses than actually making gains. This can be explained by a S-shaped figure showing a convex curve for losses and a concave curve for gains.
Managerial biases affect the decision making process of managers. Especially in the investment sector it is self-explanatory that irrational decisions can result in unexpected worse results. For instance, the disposition effect, which refers to the tendency to hold on longer to projects with negative returns than to projects with positive returns. Generally, people would rather avoid situations in which they have to make a difficult decision (Shiller, From efficient markets theory to behavioural finance, 2003). An example concerning investor behaviour is that when investors face a negative outcome they may not want to short the investment in the hope it will become positive again. However, selling it off at a lower price than the initial investments still means that the agent has made a loss (Shiller, From efficient markets theory to behavioural finance, 2003).
Besides the disposition effect, financial cognitive dissonance and herding belong in the list of cognitive investor behaviour. According to Festinger (1957) cognitive dissonance is the inconsistency in behaviour and beliefs that lead to irrational decision making (Festinger, 1957). When making a decision, one is indifferent between two choices but receives an unpleasant feeling with either one of the choices. Applying this to management decisions, this may result in choosing a project basing decisions on intuition and not on the fundamental values. Another pitfall of cognitive factors on decision making is that people tend to make less thoughtful decisions by just acting upon decisions of other people, which is called herding. For instance, investing in the same stock as other investors but not relying the decisions on the actual fundamental underpinnings.
2.1 Overconfidence
Overconfidence among managers has to be considered as a significant cause of irrational decision making following prior literature. According to Plous (1993) there is not a topic that is ‘more prevalent and potentially catastrophic than overconfidence’ (Plous, 1993). Overconfident managers systematically overestimate their own skills and assume that they have a higher ability in situations than they actually have. For instance, overconfident executives attribute corporate successes to their own personal qualities. This chapter will focus on the effect of overconfidence on decision making of managers, focusing on CEOs and CFOs.
Optimism is said to be highly accompanied with overconfidence, and is referred to the tendency to feel above average. Optimism has been developed during the evolution of human beings and has been significantly valuable in past successes, it has been regarded as one of the characteristics of life to accomplishment (Eshragi & Taffler, 2012). Several cognitive biases apply on overconfidence as a result of optimism. For instance, human beings tend to overestimate their role in successes, the self-serving attribution bias, and tend to attribute successes to their own skills and in the case of losses to just bad luck (van den Steen, 2002).When they hear the outcome of an event or a question, they tend to overestimate the predictability of the outcome, called the hindsight bias. Furthermore, they tend to overestimate their value in a partnership or a joint effort when having succeeded, called the self-centric attribution bias (van den Steen, 2002). And the illusion of control, that is, people think their attributions affect uncontrollable events, while it actually cannot (van den Steen, 2002). For example choosing the numbers of a lottery ticket are subject to illusion of control.
Executives are prone to such biases as they are being in charge of corporate control and have the authority of the company. The managerial power they have, sometimes leads to distortions in the objectives between the CEO and the shareholders. Executives in general are more likely to be overconfident. Often CEOs are more known for their charisma than their actual managerial power.
Malmendier and Tate (2007) show in their study that overconfident CEOs generally feel ‘above average’ and also systematically overestimate their future returns, which can result in serious company distortions. Furthermore, it is found that overconfidence of managers led to an increase in mergers and acquisitions. They not necessarily make more mergers and acquisitions but it is 65% more likely that they get into new mergers or acquisitions compared to non-overconfident managers, a decision generally made by the CEO (Malmendier & Tate, 2007).
Itzhak et al. distinguish between optimism and miscalibration as a feature of overconfidence. In their eyes, optimism is when an overconfident manager overestimates the company’s assets. Miscalibration is a measure often used for overconfidence and is represented by the narrowness of market return intervals. In their study they observed 7000 financial executives on stock results and they show that most financial executives are miscalibrated, only 38% of the market returns are in between the confidence intervals of 80%. Furthermore, they find that overconfident CFOs use lower discount rates, expect lower IRRs for their own future cash flows, tend to spend more on capital expenditures, pay out less dividends and are more likely to engage in acquisitions consisted with the study of Malmendier and Tate (Itzhak, Graham, & Harvey, 2007). Differences in gender seem to have an effect on engaging in acquisitions and issuing debt. Huang and Kisgen state that male CFOs engage more often in acquisitions and are more likely to issue debt. Returns from acquisition are somewhat 2% lower than female executives (bron).
Malmendier and Zheng (2012) study the effect of overconfidence on financial and non-financial decisions of CEOs and CFOs. They find that overconfident CEOs and CFOs generally are more likely to issue debt in external capital. Though in bad financial time, CFOs are more likely to issue debt and CEOs instead prefer to issue equity (Malmendier & Zheng, 2012). Furthermore, overconfident CEOs tend to pay out less or no dividend, because they prefer not to finance with external equity (Deshmukh, Goel, & Howe, 2013).
On the other side, confidence should not necessarily be a bad characteristic. Hirschleifer (2012) shows that confident managers are more likely to contribute to company innovations due to their lower risk aversion (Hirschleifer, Low, & Hong Teoh, 2011). A CEO as Steve Jobs is known for his overconfidence traits, ignoring feedback and criticism from his colleagues, but made him a great innovator. According to Goel and Thakor (2008) high risk aversion leads to underinvestment (Goel & Thakor, 2008). Also the energy these manager have, may have a positive influence on company prospects.
3. What have we learned from the dotcom bubble regarding overconfidence?
During the end 90s of the previous century, technology was developing in a fast pace and internet companies began to emerge, resulting in fast increasing valuations and capital volumes in the capital market (Valliere & Peterson, 2004). It all started with the invention of the internet, which had a large impact on businesses because whole business strategies had to be changed due to technological improvements. It asked for acting strategically to keep up with the competitors and not to face bankruptcy. The evolution of the internet can be applied to the theory of Joseph Schumpeter, the well-known Austrian economist, stated that ‘innovation is the force behind economic growth and the cause of cyclical fluctuations’. A shortage in internet companies stock caused the first run on stocks of these companies. For instance, amazon.com showed a initial public offering (IPO) of $18 in 1997, rose to $106 in December 1999 and dropped to even below $15 in December 2001. The market of internet companies stock was significantly overvalued. Increasing stock prices was subject to many countries during the same time. In the United States, the Dow Jones average tripled from 1994 to 1999 (from 3600 to 11700). Alan Greenspan, former chair of the Federal Reserve Board, spoke about ‘irrational exuberance’ in the market, in a speech at the 5th of December in 1996, which immediately led to a drop in all international stock markets. He meant that trading had changed to momentum trading, trading in quick moving stock in the same direction with high volumes, and that prices may have somewhat been overvalued (Shiller, From efficient markets theory to behavioural finance, 2003). In the book irrational exuberance (2000), Robert Shiller investigates the speculative asset market from the 80s (?) until the dotcom bubble. He mentions the role of overconfidence during the bubble. The high level of overconfidence goes hand in hand with the high level of trade (Shiller, 2000). Rational investors would have had a better knowledge about their own trading ability. And in this sense, inexperienced traders, less confident traders, would rather not trade, because they know there are way more experienced traders in the market. The somewhat ‘rational investors’ may actually have had the intention to ‘beat the market’ by riding the bubble and, this way, gaining higher returns. Therefore, if the market was completely rational, markets with such a high level of speculation would not exist. And in turn, products would not be overvalued and economic bubbles would not arise.
The course of the dotcom bubble is well represented by the model of Redhead () in the appendix. After the introduction of the internet, people became more interested and enthusiastic for the new product and showed emotionally driven trade behaviour. Hereby, stock prices went up and trading behaviour changed to momentum trading, significantly investing in stock that is moving in only one direction. Due to herding behaviour of investors, prices increased more. Investors became more confident and attributed their successes in trading to their own skills, subsequently leading to even higher stock prices. Which is supported by the study of Odean and Terrance, finding that overestimating investment skills led to excessive trading (Odean & Terrance, 1999). Kyle and Wang show that excessive trading leads to more aggressively trading rational investors, which in turn leads to higher profits for confident managers (Kyle & Wang, 1997). At this point, the highest level of the bubble, investors still maintained their opinions there is no bubble and prices reflect the intrinsic values, by ignoring all information that is not supporting their view. Thereafter, prices start to decline due to higher supply and lower demand in the market. Leading to, again, herding behaviour by all selling their stock short, and subsequently to the bubble crash in 2000.
In the study of Smith (1988) it is found that bubbles tend to be caused by, in particular, inexperienced traders in comparison to experienced traders (Smith, Suchanek, & Williams, 1988). Adding to it, bubbles tend to occur more often in situations where managers had not experienced a bubble or a similar situation before, referring to learning by doing. Greenwood and Nagel state that younger managers during the dotcom bubble strongly invested in technology stock while their benchmark showed a less optimistic view (Greenwood & Nagel, 2009). Also their portfolios contained more technology stock than older managers and showed trend chasing behaviour.
Nevertheless, it must be noted that a bubble might be driven on purpose by ‘rational speculators’. Abreu and Brunnermeier show that arbitrageurs, who know that the market is overvalued, ride the bubble to maximize profits and just before the collapse of the bubble they leave the market (Abreu & Brunnermeier, 2003). However the moment of leaving the market is a difficult decision, leaving too early may not give a great return and leaving too late leaves one with big losses. Rational arbitrageurs may fail their arbitrage strategies because they are somewhat unclear about their exit strategies (Abreu & Brunnermeier, 2003). Investors drive up prices by newly obtained information to have traders aggressively invest in stocks and in turn create higher returns due to higher capital values (de Grauwe & Grimaldi, 2004).
Valliere & Peterson (2007) spoke in their article inflating the bubble about lessons we learned from the dotcom bubble. For instance they note that people thought that this new ‘internet’ industry had to be treated differently than normal business models. Due to this the normal risk assessment controls were neglected, if this was done sufficiently it may have deflated the bubble. Also, investors should have looked at the fundamental underpinnings instead of all the environmental factors as the media, a typical example of herding by investors. Following the research, ‘they should continue to use the press to highlight emerging opportunities, but not rely on it as an adequate source’ (Valliere & Peterson, 2004). Green adds that a lack of strategy contributed to the crash of the bubble. Many investment companies enlarged their organizations but a lack of focus in the market led to excessive capital, a disaster for future profits (Green, 2004).
4. What are possible management control/internal control/cost control measures to avoid irrational decision-making by CEO/CFO when companies are in the midst of a stock price ‘bubble’?
Management control was first introduced by Anthony in 1965 as ‘the process by which managers ensure that resources are obtained and used effectively and efficiently in the accomplishment of the organization’s objectives’ (Anthony, 1965). Likewise, internal control is explained by the committee of sponsoring organizations (COSO) of the Treadway Commission, as a joint initiative providing a framework to improve financial reporting, internal control and risk management, as the process of achieving objectives by the board of directors, management and other personnel inside the company in the field of efficient and effective operations, accurate financial reporting and the compliance of laws and regulations (COSO, n.a.). Even though there is a small difference between management control and internal control, both do have the same objective, keep the company ‘in control’ following the objectives of the company. The decision making of CEOs and CFOs have a large impact on the control systems inside the company as the final decision is often made by one of them, CEOs in special. This chapter will focus on CEO and CFO decision making and the measures to be taken to avoid irrational decision making like occurred during the dotcom bubble.
Management control systems exist to help the company pursue its objectives as good as possible. During a bubble like the dotcom bubble, it was evident that some agents intuitively were riding the bubble, in the hope stock to become more profitable, irrespectively of what the fundamental value was. Behaviour that arouse from managerial biases as overconfidence. Questionable is if during this time was known if the management control system did its task or that unconsciously irrational behaviour of executives took over and let greedy investors do their work. Subsequently resulting in the burst of the bubble.
The aftermath of the dotcom bubble resulted in many committed frauds, investors received misleading information and shareholder money was being misused. The company WorldCom is a typical example of such a case. WorldCom was a company in telecommunication which inflated their assets by $11bn and finally got caught by the internal control of the company in 2002. CEO Bernie Ebbers was sentenced for fraud and received 25 years of jail. After a few more fraud cases, in 2002, the Sarbones-Oxley act was being enacted to improve the regulation of internal controls and corporate governance. All companies have to comply to these regulations since in order to prevent against fraud. It resulted in better information distribution from CEO to shareholder/investors and to less aggregate investment (Goel & Thakor, 2008).
Following prior research of the implications of overconfident executives it is evident that it can be crucial for the firm. Especially in times of a bubble, where a possible burst can ruin the firms soundness completely, irrational decision should be avoided. As management control is in charge of avoiding this, it is important that competent executives make well considered decisions. Executives are chosen by the board of directors and in the study of Goel and Thakor (2008) it is found that during the selection procedure especially the overconfident managers are chosen (Goel & Thakor, 2008). The decision of the board depends on the relation between the ability of the incumbent CEO and the CEO overconfidence. The board fires both the low ability CEO and the excessively taking risk CEO and the excessively overconfident CEO. They also show that the overconfident managers are also most likely to get fired or ask for resignation. All this, is a costly process and would preferably be avoided. To reduce the chances on overconfident CEOs (or CFOs), it is advisable to start with the selection procedure. Choi et al. find that forced CEO turnovers in a firm with good corporate governance results in significantly higher firm returns. In firms with relatively weak corporate governance, firm performance and accounting measures have improved. The board of directors should take into account that overconfidence may result in worsening firm performance. In the study of …. it is found that non confident managers tend to underinvest, not in favour of the company, and overconfident managers to overinvest. Therefore it is best when a just confident CEO is charge of corporate control.
Chen et al. study the relationship of overconfident managers to ineffective internal control, which they find to be correlating (Chen, Lai, Liu, & McVay, 2014). Due to the efficient work of the ‘good managers’ inefficient internal control is less likely to happen with them. However a company with an overconfident CEO but a strong corporate governance is more likely to maintain effective controls. Also when an non-confident manager is replaced by a confident manager it is more likely that ineffective internal control is being maintained. Besides they make the distinction between CEOs and CFOs, they find that CEOs are related to inefficient investment and CFOs to bad financial reporting. CEOs have a bad influence on the internal control of the company as a whole and the CFO more on account level. They also find that overconfident managers are less likely to improve their weaknesses.
Executive compensation as a compensation of the success of a company, definitely contributes to the confidence of a CEO of CFO. Positive feedback, as executive compensation in the form of options or bonuses was given in enormous forms. Also when the dotcom bubble bursted, some companies were providing higher expectations of their earnings to affect the compensation by options and still huge bonuses were provided although companies could not afford this. Executive compensation may also result in excessive risk taking due to increasing overconfidence, in contradiction to the effect compensation should create. Paredes states that CEO or executive overconfidence is subject to corporate governance because they are head of corporate control and in charge of all the final decisions (Paredes, 2004). Compensation should be in line with the objectives of the shareholders, but it should not result in excessive trading due to CEOs becoming highly overconfident.
‘
Appendix
Figure 1. A behavioural model of the dotcom bubble.
‘
References
Abreu, D., & Brunnermeier, M. (2003). Bubbles and Crashes. Econometrica vol. 71, 173-204.
Anthony, R. (1965). Planning and Control Systems: A Framework for Analysis. Boston: Harvard University.
Biais, B., Hilton, D., Mazurier, K., & Pouget, S. (2005). Judgmental Overconfidence, Self-Monitoring and Trading Performance in an Experimental Financial Market. The Review of Economic Studies, pp. 287-312.
Chen, S., Lai, S., Liu, C., & McVay, S. (2014). Overconfident managers and internal controls.
COSO. (n.a.). Internal control – Integrated framework. COSO.
de Grauwe, P., & Grimaldi, M. (2004). Bubbles and crashes in a behavioural finance model. CESifo Working Paper Series No. 1194.
Deshmukh, S., Goel, A., & Howe, K. (2013). CEO Overconfidence and Dividend Policy. Journal of Financial Intermediation.
Eshragi, A., & Taffler, R. (2012). Fund manager overconfidence and decision making: Evidence from mutual funds.
Fama, E. (1970). Efficient capital markets: A review of theory and empirical work. Journal of Finance, pp. 383-417.
Festinger, L. (1957). A theory of cognitive dissonance.
Fromlet, H. (2001). Behavioural Finance-Theory and Practical Application. Business Economics, pp. 63-70.
Goel, A., & Thakor, A. (2008). Overconfidence, CEO selection and corporate governance. Journal of Finance, pp. 2737-2784.
Green, J. (2004). venture capital at a new crossroads: Lessons from the bubble. Journal of management development vol. 23, pp. 972-976.
Greenwood, R., & Nagel, S. (2009). inexperienced investors and bubbles . Journal of financial economics, pp. 239-258.
Hirschleifer, D., Low, A., & Hong Teoh, S. (2011). Are overconfident CEOs better innovators? Journal of Finance, forthcoming.
Itzhak, B., Graham, J., & Harvey, C. (2007). Managerial overconfidence and corporate policies. NBER Working Paper No. 13711.
Kahneman, D. (2011). Thinking, fast and slow. Macmillan.
Kahneman, D., & Tversky, A. (1973). Availability: A heuristic for judging frequency and probability. Cognitive psychology, pp. 207-232.
Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis under risk. Econometrica, 263-292.
Kahneman, D., & Tversky, A. (1981). The framing of decisions and the psychology of choice. Science, 453-458.
Kyle, A., & Wang, A. (1997). Speculation duopoly with agreement to disagree: Can overconfidence survive the market test? Journal of Finance 52 , pp. 2073-2090.
Malmendier, U., & Tate, G. (2007). Who makes acquisitions? CEO overconfidence and the market’s reaction. Journal of Financial Economics.
Malmendier, U., & Zheng, H. (2012). Managerial duties and managerial biases. UC Berkeley.
Odean, & Terrance. (1999). Do investors trade too much? American Economic Review vol.89, pp. 1279-1298.
Paredes, T. (2004). Too Much Pay, Too Much Deference: Is CEO Overconfidence the Product of Corporate Governance? Washington U. School of Law Working Paper No. 04-08-02 .
Plous, S. (1993). The psychology of judgement and decision making. New York: McGraw-Hill.
Ricciardi, V., & Simon, H. (2000). What is behavioural finance? Business, Education & Technology Journal, Vol. 2, No. 2, pp. 1-9.
Shefrin, H., & Statman, M. (1985). The disposition to sell winners too early and ride losers too long: Theory and evidence. Journal of finance40, no. 3, pp. 777-790.
Shiller, R. (2000). Irrational Exuberance. New York: Princeton University Press.
Shiller, R. (2003). From efficient markets theory to behavioural finance. Journal of Economics Perspectives vol. 17, 83-104.
Smith, V., Suchanek, G., & Williams, A. (1988). Bubbles, crashes and endogenous expectations in experimental spot markets. Econometrica, pp. 1119-1151.
Smith, V., Suchanek, G., & Williams, A. (1988). Bubbles, crashes, and endogeneous expectations in experimental spot asset markets. Econometrics vol.56, pp. 1119-1151.
Thoma, V., White, E., Panigrahi, A., Strowger, V., & Anderson, I. (2015). Good thinking or gut feeling? Cognitive reflection and intuition in traders, bankers and financial non-experts. PLoS ONE.
Valliere, D., & Peterson, R. (2004). Inflating the bubble: examining dot-com investor behaviour. Venture capital: An international journal of entrepreneurial finance vol. 6, pp. 1-22.
van den Steen, E. (2002). Skill or luck? Biases of rational agents. MIT Sloan School of Management Working paper4255-02.
here…