BP’s Failure to Debias (Page 5 of 6)

Subsequent to April 20 some analysts adjusted their perspective, especially about the importance of corporate culture. Morgan Stanley Europe’s July 28 report states, “Investors will need more clarity on the impact of asset sales and further reassurances of a cultural change regarding safety, a process started by Tony Hayward, before BP can regain a multiple in line with its industry peers.” Similarly, the July 29 report by Oppenheimer states: “CEO Change. Although we believe Bob Dudley is the right person to replace Tony Hayward as CEO, we think BP is in serious need of an extreme makeover to change its culture and the way it conducts business. In order to achieve that, many key managers may have to be replaced.”

MSCI’s ESG Research group revised the manner in which it analyzes risk. It now focuses on risks associated with regulatory risk in respect to health and safety, especially in offshore operations. Notably, its framework for analyzing health and safety involves examining management systems, budgets, track records for oil spills and fatality/injury rates. In September, the research group noted that BP had the highest offshore regulatory risk among its peers and was ranked at the bottom for risk management systems in this dimension.

Nevertheless, some analysts appear to have been unconvinced about the centrality of BP’s risk management, especially in connection with well design. In a report dated September 9, 2010, UBS analyst Jon Rigby and associate analyst Caroline Hickson made the following comment, based on the release of a BP report (BP, 2010) released on September 1, “Report offers some rebuttal to critics – well design not to blame… BP says ’multiple’ parties and causes involved in Macondo… BP’s internal (although independent) report into the Macondo incident (the Bly report) indicates there was no one clear cause or culprit of the disaster… The committee stressed in a conference call that the well design was ’robust’ and not unusual for the area, and also defended the use of only six centralizers and the limited circulation of the drilling mud. These decisions had all been listed as possible causes of the blowout by the US Congress…”

Rigby appears to have based his assessment on BP’s report without adjusting for potential self-interest bias on the part of BP, and assessments of competitors such as Shell and Exxon. One possible explanation for doing so is the issue we described earlier, whereby analysts write reports that are excessively favorable in order to curry favor with the companies they follow, in the hope the management of these companies will engage the firms for which they work. This motivation has been suggested in the popular financial press. See Pressman (2010), who discusses this possibility.26

5.3 Investors and Regulators: Availability Bias and Confirmation Bias

Turning to investors, an interesting example is offered by the Corporate Responsibility (CR) community that seems not to have understood the incongruity between BP rhetoric and deeds. BP ranked first in the 2007 Fortune and AccountAbility’s list of world’s “Most Accountable Companies,” the annual ranking of business responsibility. In 2010 it was named runner-up for the category “Openness and Honesty” in the Corporate Register’s CR Reporting Awards (Sverjensky, 2010). Prior to the explosion of Deepwater Horizon, the Dow Jones Sustainability Indexes (DJSI) identified BP as leading its peer companies in corporate sustainability and commitment to shaping the industry in the social and environmental aspects of business (Statman, 2010). Of course, after the explosion, DJSI removed BP from its indexes.

In our view, the factors underlying investors’ underestimation of BP’s risk profile are the same as the factors underlying those of analysts’ judgments, namely availability bias and confirmation bias. Moreover, analysts’ judgments may well have impacted investors’ judgments. Even if based on publicly available data, most investors appear to rely on information provided by companies. This information often only includes companies’ commitments, i.e., what they state they plan to do, and are often based only on the CSR reports of those companies. Instead, rankings and indices should be driven by externally verified data coming from a variety of sources (Sverjensky, 2010).

In contrast to analysts and investors, the relationship between firms and regulatory agencies also features several potential agency costs. For example, a supervisory authority may face the conflicting duties of regulating supervised companies and collecting royalties from them. This seems to have been the case of the Minerals Management Service (MMS), which regulated US oil exploration. The federal agency had been criticized since long before the spill for its conflicting duties (CNN Wire Staff, 2010). On May 19, 2010, Interior Secretary Ken Salazar, whose department oversees offshore oil drilling, announced that he was dividing MMS into three divisions. He affirmed, “We inherited here what was a legacy of an agency that essentially was rubber-stamping whatever it was that the oil and gas industry wanted” (CNN Wire Staff, 2010).27 The new name of the agency “Bureau of Ocean Energy, Management, Regulation and Enforcement” has been chosen to stress the different duties of the distinct divisions.

The National Commission concluded that MMS lacked the resources and political support required to promulgate regulations necessary for risk reduction. The Commission noted that MMS lacked personnel with the expertise and training needed to enforce effective regulations. In this regard, they point to the approval within 90 minutes by the agency of BP’s April 16 plans for the replacement of so much mud by seawater in the temporary abandonment plan, without assessing of the full risks of the plan. They also note that MMS did not even require that negative pressure tests be part of the protocol for deepwater drilling. It was actually BP that proposed performing the negative pressure test in its April 16 application. Counterfactually, had BP correctly interpreted the signals from that test, the explosion might have been averted.

6. DEBIASING USING COGNITIVE REPAIRS

As we stressed in the previous section, one of the most important lessons for analysts from the BP-Macondo incident involves the need to assess how a firm’s culture impacts its risk profile, and by implication its prospects and value. In our view, analysts need to develop systematic procedures for assessing the extent to which firms are working to improve their cultures, particularly their susceptibility to psychological pitfalls. On the other side of the coin, one of the most important lessons for firms’ management teams is the importance of mitigating their vulnerability to psychological pitfalls, thereby improving their firms’ cultures.

In this section we apply the literature on debiasing and cognitive repairs to suggest ways that BP in particular, and other firms in general, can improve corporate culture. The 4x4 pitfalls-process framework described in section 2 provides the structure for our discussion, together with the cognitive repair approach described in Heath et al. (1998).

Shefrin (2008) argues that the starting point for instituting organizational debiasing is the recognition that psychologically induced mistakes are akin to addictive diseases. What behaviorally induced mistakes and addictive behaviors share in common is habituation. We know something about how to treat addictive diseases. We know that twelve-step programs are group programs that have truly helped many people combat their addictions. Indeed, “step one” of twelve is to acknowledge the problem. In the case of psychologically induced mistakes this means acknowledging susceptibility to phenomena such as confirmation bias, excessive optimism, overconfidence, and aversion to a sure loss. Shefrin suggests that successful debiasing often requires group interaction. Heath et al. make a similar point, stating that “many successful repairs will be social because individuals may not recognize the need to repair themselves” (p. 28).

The extent to which BP’s management has taken “step one” remains an open question. During appearances before both the US Congress (in May) and the British Parliament’s energy panel (in September), Hayward was aggressively questioned about his promise some years ago to focus like a laser on safety. In response, he stated that BP’s record was “better than the industry average.” In this regard, he went on to say that the blowout preventer on the rig “was fully compliant with the regulatory regime and it should have functioned.” When asked about a BP employee having described Macondo as a “nightmare well,” Hayward termed the description “unfortunate” and noted that the well “had been challenging—not unusually so for the Gulf of Mexico. The gulf is a more challenging drilling environment than the rest of the world.”


NYU Masters in Risk Management


To be sure, Hayward’s statements are intentionally crafted with the view of limiting BP’s financial liability for the explosion of Deepwater Horizon. The same statement applies to the report BP released on September 1, 2010 mentioned above, explaining the multiple

causes of the explosion. The BP report describes failures on the part of all parties involved, including contractor Halliburton and rig owner Transocean. Not surprisingly, both Halliburton and Transocean disputed BP’s findings: Transocean called them “self-serving” and Halliburton insisted that the source of the problem was BP’s well design. In this regard, we note that in January 2011, the National Commission reported to the President that all three parties had made serious mistakes. The report also pointed to difficulties at the government regulatory body (MMS) which was overseeing their operations.

Taken at face value, our view is that Hayward’s statements reflect confirmation bias. As the MSCI materials discussed in section 4 indicate, BP’s offshore risk management practices lie at the bottom of their peer group. This is the case, even though during 2009, BP’s history of spills and fatality/injury rates was better than others. As we discussed in section 3, BP’s well design treated the blowout preventer as a barrier, whereas alternative well designs such as those in use at Shell, treat the blowout preventer as a control. In this respect, we believe that Halliburton has a legitimate point. As for the characterization of the “nightmare well” designation as “unfortunate,” consider remarks from one of the mechanics who worked on Deepwater Horizon, “I’ve seen a lot of gas coming up from muds on different wells, and the highest I’ve ever seen in my 11 years was 1,500 units. And this well gave us 3,000. I’ve never been on a well with that high of gas coming out of the mud. That was kind of letting me know this well was something to be reckoned with.”

As the title of this paper indicates, under Tony Hayward’s leadership, BP failed to debias. In respect to the explosion of Deepwater Horizon, the National Commission identified “recurring themes of missed warning signals, failure to share information, and a general lack of appreciation for the risks involved. In the view of the Commission, these findings highlight the importance of organizational culture and a consistent commitment to safety by industry, from the highest management levels on down.” (p. ix) To us, missed warning signals, failure to share information, and a general lack of appreciation for the risks involved translates as confirmation bias, poor processes for sharing information, and overconfidence.

BP replaced Tony Hayward as CEO with his colleague Bob Dudley. On March 8, 2011, Dudley made a speech in which he said “BP is sorry. BP gets it. BP is changing.” Will Dudley be able to succeed where Hayward failed?

In the context of the 4x4 pitfalls-process framework, consider some cognitive repair techniques aimed at mitigating bias at BP. In the discussion below, we focus on processes first, identify pitfalls, and then suggest repairs. Notably, the discussion stresses the importance of interaction across processes.

  1. Standards: BP appears to have framed the safety issue by using an excessively narrow definition of safety, a point made by Michaels (see section 4.1). The most important cognitive repair here involves reframing the notion of safety much more broadly. In addition, standards for risk management need to be appropriately framed. Gigerenzer et al. (2008) identify a series of framing pitfalls in connection with randomness. They emphasize the use of frequency statements in place of single-event probabilities, absolute risks instead of relative risks, mortality rates instead of survival rates, and natural frequencies instead of conditional probabilities.
  2. Planning: BP characterized the explosion of Deepwater Horizon as a rare bad luck event, stemming from the simultaneous occurrence of a series of unfavorable events. We think it plausible that BP’s management exhibited overconfidence in underestimating the probability of a failure, with anchoring and adjustment bias possibly being a contributing factor. Mitigating overconfidence is a challenge, particularly when it comes to overconfidence about knowledge, as measured by the use of stated confidence intervals. See Kaustia and Perttula (2010). Heath et al. (1998) discuss routine cognitive repairs that engineers use for addressing overconfidence. They state, “Fortunately, the engineering profession has developed a particular repair, called ’safety factors,’ that mitigate the overconfident reasoning of individual engineers.” (p. 4). These repairs involve the use of additional safeguards that would be excessive if not redundant, were engineers well calibrated.
  3. Incentives: On the surface, BP’s compensation appears to align the interests of managers and shareholders. The firm required employees to take a significant proportion of their compensation in company stock each year and to retain them. Therefore, many BP staff came to be heavily invested in the firm through employee share ownership plans. Shefrin (2006) argues that incentives alone do not align the interests of managers and shareholders. In particular, biases and framing effects associated with the other three processes standards and planning can undercut theoretically appropriate compensation plans. As was discussed in section 3.2, aversion to a sure loss is particularly dangerous, with its associations to sunk cost fallacy, escalation of commitment, and risk seeking decisions. See March and Shapira (1992). Shefrin (2006) describes several techniques for addressing aversion to a sure loss. One of these was coined “fire yourself” by Intel’s former CEO Andrew Grove. He suggested that the best way for an executive to deal with the reluctance to terminate losing projects is to pretend to fire him or herself, and act as he or she would were they to be the replacement, without the associated psychological pitfalls. Technically speaking, Heath et al. suggest that improving incentives is separate from instituting cognitive repairs. Nevertheless, firms like BP need to address how people are rewarded (or penalized) for resisting cost cutting measures they deem to be value destructive. Rewards and penalties can be nonpecuniary (praise and blame) as well as financial.
  4. Information sharing: This is a broad category, which involves (1) hypothesis generation, (2) information collection, and (3) the drawing of conclusions.

An example of a hypothesis in respect to the April 20 test results on Deepwater Horizon is that readings were caused by “U-tubing.” Heath et al. suggest that individuals often generate too few hypotheses, as may have been the case with the U-tubing hypothesis. They suggest two particular cognitive repairs for this tendency, known as “the Five Whys” and “single-case bore questions,” both of which are procedures for asking questions designed to address overly narrow search.

1 | 2 | 3 | 4 | 5 | 6 | Next Page

Hersh Shefrin and Enrico Maria Cervellati


26. Pressman also discusses alternative behavioral explanations such as groupthink. In addition, he provides examples from several analysts’ reports, including those from Credit Suisse, Citigroup, and Morgan Stanley.

27. This statement is consistent with both confirmation bias and agency conflicts.

UConn-ad

NYSSA Job Center Search Results

To sign up for the jobs feed, click here.


UPCOMING EVENT
MF16

NYSSA Market Forecast™: Investing In Turbulent Times
January 7, 2016

Join NYSSA to enjoy free member events and other benefits. You don't need to be a CFA charterholder to join!


CFA® EXAM PREP

CFA® Level I 4-Day Boot Camp
Midtown

Thursday November 12, 2015
Instructor: O. Nathan Ronen, CFA

CFA® Level II Weekly Review - Session A Monday
Midtown

Monday January 11, 2016
Instructor: O. Nathan Ronen, CFA

CFA® Level III Weekly Review - Session A Wednesday
Midtown

Wednesday January 13, 2016
Instructor: O. Nathan Ronen, CFA

CFA® Level III Weekly Review - Session B Thursday
Thursday January 21, 2016
Instructor: O. Nathan Ronen, CFA

CFA® Level II Weekly Review - Session B Tuesday
Thursday January 26, 2016
Instructor: O. Nathan Ronen, CFA