What Happened to Goldman Sachs: An Insider's Story of Organizational Drift and Its Unintended Consequences (19 page)
Authors: Steven G. Mandis
Outside Board of Directors
Some have argued that boards of directors of the banks must take more responsibility for oversight of conflicts, so it’s important to consider the role of the Goldman board here briefly.
The Sarbanes–Oxley Act, which became law in 2002, sought to improve the oversight of boards by requiring the appointment of more outside directors. In 1999, Goldman’s board of directors had seven members, including four executives of the firm; two outside, independent directors; and John L. Weinberg. In 2001, Goldman’s board of directors had eight members: four executives at Goldman and four outside directors (John L. had decided not to seek reelection to the board after his initial three-year term at the IPO). After Sarbanes–Oxley passed, and in compliance with new requirements of the NYSE, Goldman added two outside directors. Thus, the majority of its board (60 percent) was comprised of outside, independent directors.
While a board’s legal and fiduciary responsibilities are typically explained by legal counsel at board meetings for board members, according to those I interviewed, understanding, interpreting, and protecting Goldman’s principles, values, and ethics, and determining whether practices are consistent with them, would be very difficult for outside board members. Many of the partners whom I interviewed generally agreed that it is reasonable to conclude that the outside board members’ understanding of the firm’s principles, values, and ethics most likely differs from that of Whitehead and John L. Weinberg because the organizational and regulatory structures and legal liabilities the firm is subject to have changed since their time. At the same time, most of the Goldman “insiders” (executives) on the board have a trading background, and some partners suggested that this may mean that they are less sensitive to issues regarding managing client relationships than Whitehead and Weinberg were.
A Change of Meaning
While some at Goldman told me that the firm continues to abide by its first principle, I believe the evidence shows that that can only be considered true if the interpretation of the principle has changed. I believe the process of this change was one of social normalization, or rationalization, which was gradual, and this in part explains why they do not see it. This kind of rationalization happens because the pressures are so strong to grow and change, which can result in client conflicts. It is vital in evaluating how an organization’s culture has drifted to recognize that organizations generally must adapt to new competitive demands and other external pressures, and Goldman’s change to applying a legal standard in regard to serving clients’ interests is a good case in point. While the differences may now look to outsiders like a clear change of standards, discerning drift that is detrimental from healthy adaptation in the moment is a very difficult judgment for those inside an organization to make. That was all the more true for Goldman due to its increasing complexity and size. Also, this has contributed to making conflicts much more complicated to manage.
Part Four
GOLDMAN’S PERFORMANCE
Chapter 8
Nagging Questions: Leadership, Crisis, and Clients
T
HE MERE FACT THAT GOLDMAN SURVIVED AS AN INDEPENDENT
company in 2008 during the financial crisis when many of its peers did not can be considered a success, at least in a relative sense, but it is a precarious one at best. Goldman wrote in its proxy statement, “In 2008, we outperformed our core competitors due, in part, to the outstanding performance of our Named Executive Officers (NEOs).”
1
Goldman had a return on equity of 4.9 percent versus −5.0 percent for its peers in 2008, and 22.5 percent versus −1.8 percent, in 2009.
Whether Goldman would have survived without government intervention is debatable.
2
In fact, in the days immediately following the collapse of Lehman, it became apparent that both Goldman and Morgan Stanley could have shared the same fate as Lehman.
3
I focus here first on the aspects of Goldman’s culture that helped it relatively outperform its peers during the credit crisis. Then I’ll explore the accusations of wrongdoing leading up to and during the financial crisis that were made against Goldman and how the firm has responded. The nature of the accusations, and Goldman’s responses, offer much food for thought about the change in culture and the potential future risks that Goldman, and the whole banking system, face. Organizational drift doesn’t necessarily involve a total abandonment of prior values and principles. Elements of the culture may be retained, or retained in part, and that is true of Goldman. For simplicity I will use the word
residual
in front of a word describing an element of the culture that has changed but not enough that it is gone or unrecognizable.
4
In fact, the residual elements help obscure the changes and add to the process of organizational drift.
One of the most important elements of Goldman’s culture that has changed but is still recognizable is what I’ll call its residual dissonance. As discussed,
dissonance
is the term used by Columbia University sociologist David Stark to describe the ability of those in an organization to challenge one another, to ask questions and explain their own views. Dissonance of this type leads to more scrutiny of decisions as well as greater innovation and performance. The financial interdependence of partners during Goldman’s partnership days, as well as the social network of trust among them and the less hierarchical structure, encouraged this, and though the new organizational and incentive structure of the company has limited this residual dissonance, a strong enough social network among executives at the firm still exists that these were key factors that differentiated Goldman during the credit crisis. They helped to break through structural secrecy, and that, combined with more expertise at the top of the firm in trading and risk assessment, enabled Goldman to do a better job of perceiving and managing the risk that led to losses. Blankfein was an expert in trading, and David Viniar, Goldman’s chief financial officer, had vast experience outside traditional CFO responsibilities. The combination of expertise and residual dissonance at the top enabled Goldman to overcome structural secrecy.
Based on my interviews with executives at competitors, these cultural elements did not exist at other firms in the same way or intensity as they did at Goldman during the credit crisis. While discussing Goldman’s success with me, a widely respected consultant, who has experience working with many firms, explained that Goldman is exceptionally good at looking at overall risk and firmwide risk and understanding the aggregate size of the risk and correlations across the firm. He believes that Goldman had so many different proprietary desks in so many different asset classes with so many different correlations that it benefits from a diversification effect. When the corporate credit or equities businesses are doing poorly, then foreign exchange or interest rate businesses may be doing well. No other bank had invested as much in sophisticated, computer-driven quantitative systems to reveal the signals. And several senior people had the expertise to read the signals, ask the right questions, and then react.
5
Goldman was the only firm that had so many risk experts in the highest levels of management. As mentioned earlier, Goldman had learned from its 1994 experience.
Value at Risk, Models, and Risk Management
Models are widely used in risk management to synthesize risk and help analysts, investors, and company boards determine acceptable trading parameters under different scenarios. Value at risk (VaR) is a widely used measure of the risk of loss on a specific portfolio of financial assets, expressed in terms of a probability of losing a given percentage of the value of a portfolio—in mark-to-market value—over a certain time. For example, if a portfolio of stocks has a one-day 5 percent VaR of $1 million, there is a 0.05 probability that the portfolio will fall in value by more than $1 million over a one-day period. Informally, a loss of $1 million or more on this portfolio is expected on one day in twenty. Typically, banks report the VaR by risk type (e.g., interest rates, equity prices, currency rates, and commodity prices).
VaR may be an unsatisfactory risk metric, but it has become an industry standard. Wall Street equity analysts expect banks to provide risk (VaR) calculations quarterly, and they talk about risk increasing, or decreasing, depending on the output of the models. The models and VaR calculations, however, make numerous assumptions, some of which proved over time to be invalid, making it dangerous to rely on or extrapolate too much on VaR. Analysts and investors (and boards of directors) overrely on VAR as a measurement of risk, and therefore management teams do also, one of the external influences of being a public company. Yet the overreliance on VaR, one of the key measures employed in risk management, is controversial. Some of the claims made about it include that it “[ignores] 2,500 years of experience in favor of untested models built by non-traders; was charlatanism because it claimed to estimate the risks of rare events, which is impossible; gave false confidence; would be exploited by traders.”
6
Comparing VaR to “an airbag that works all the time, except when you have a car accident,” David Einhorn, the hedge fund manager who profited from shorting Lehman stock, charged that VaR also led to excessive risk-taking and leverage at financial institutions before the crisis and is “potentially catastrophic when its use creates a false sense of security among senior executives and watchdogs.”
7
Leading up to the crisis most of Wall Street essentially used the same models and metrics for risk management, particularly VaR (an effect of being public—analysts and investors compare VaR between firms in analyzing performance). But Goldman did not rely as heavily on it. Interviews confirmed the level of dissonance at Goldman, even as a publicly traded firm, in discussing and understanding that the output of the models was and is unique to Goldman, which meant the firm was not as dependent on the models as were other firms, and that, combined with what sociologists call a “heterarchical structure” (less hierarchy in the chain of command than in many firms) and the trading experience of its top executives, gave Goldman an edge.
8
The more intense scrutiny of the models and risk factors led Goldman’s top executives to pick up on market signals that other firms’ executives missed.
9
As Emanuel Derman, the former head of the quantitative risk strategies group at Goldman and now a professor at Columbia, wrote, at Goldman, “Even if you insist on representing risk with a single number, VaR isn’t the best one … As a result, though we [Goldman] used VaR, we didn’t make it our religion.”
10
(Meanwhile, at other firms, measures like “VaR [value at risk] … became institutionalized,” as the
New York Times’
Joe Nocera put it. “Corporate chieftains like Stanley O’Neal at Merrill Lynch and Charles Prince at Citigroup pushed their divisions to take more risk because they were being left behind in the race for trading profits. All over Wall Street, VaR numbers increased.”
11
)
Even though VaR has flaws, it is the only relatively consistent risk data that is publicly reported from the various banks, which is why I analyzed it. When analyzing the publicly reported data from 2000 to 2010 for Goldman and its peers, what stands out is that Goldman’s standard deviation of VaR is higher (meaning that the level of the total VaR was more varied) than most other firms, implying that Goldman more dynamically managed risk than its peers over the time period.
Goldman also had the organizational structure and environment to complement and support risk management systems and procedures. During an interview for this book, an executive at a competing firm (who had earlier worked at Goldman) explained that his firm simply did not give the same attention to risk management. He did not think his firm had the capability to aggregate risk at a level similar to Goldman and said the top executives neither had the capability nor dedicated the time to interpret, discuss, and debate risk. He said that Goldman’s focus on proprietary trading and its profits had caused the firm to invest in systems and groom future leaders who understand it. He explained that Goldman’s biggest advantage was that its top people were real traders and risk takers or had access to and dialogue with such traders, as well as the culture to support investment in the systems and the dialogue. He explained that this was why he hadn’t been surprised when Hank Paulson and the board had bypassed John Thornton and John Thain (to whom Paulson allegedly had verbally promised the CEO position) and picked Blankfein to succeed him as CEO. Thornton and Thain did not have as much real time and extensive expertise in trading and risk. The firm gained trading expertise at its top with Blankfein, and that helped it navigate the credit crisis.
12
One executive at a competitor speculated that although his firm had people running around doing lots of things related to risk, including analyzing lots of models, no one knew how it all added up and, more important, what it meant. He questioned whether it was even possible to understand risk management in such large, complex organizations. But he said if one firm could, it was probably Goldman, which seemed to prioritize risk management because of its dependence on proprietary trading and because it had the cultural heritage of teamwork, as well as its near-death experience in 1994.
While some analysts of the crisis have pointed to the failure of the boards of directors at the banks to question risk-taking, most of the people I interviewed said that those on the boards of the banks had limited trading and risk expertise and that it would be almost impossible, anyway, for an outside, independent director who has an important full-time job and attends monthly or quarterly meetings to be able to take the time to understand and question such complex risks.
13
That made having executives at the top of the firm with trading and risk management experience all the more important, especially when there wasn’t a mechanism like financial interdependence of a partnership.
Communicating the Signals
Goldman’s relatively flat organizational structure and the relative strength of its social network are exemplified in a series of e-mails and memos from as early as the end of 2006 that were sent to top Goldman executives by traders. They discussed the firm’s exposure to mortgages and the top executives’ ability to understand the issues, concluding that Goldman should reduce risk as quickly as possible. There are e-mails from Blankfein and Viniar giving direct guidance related to risk.
14
In one e-mail, CFO Viniar wrote, “Let’s be aggressive distributing things because there will be very good opportunities as the markets [go] into what is likely to be even greater distress, and we want to be in a position to take advantage of them.” Blankfein wrote in one e-mail, “Could/should we have cleaned up these books before … and are we doing enough right now to sell off cats and dogs in other books throughout the division?” The communications show how involved the leaders were. (They also raise serious questions and concerns about what Goldman leaders knew, did, and instructed others to do, and the impact of these actions on the firm’s culture.)
Meanwhile, even as Goldman reversed course and tried to reduce risk in mortgages—and even allegedly shorted the market—its competitors were adding risk. For example, Chuck Prince at Citi was, as described and paraphrased in an interview, “dancing with elephants.”
15
A former Citi executive I talked to at the time expressed doubt that Prince (whose professional background was more legal and administrative in nature) had the same sort of direct communication, knowledge, and discussion that Viniar and Blankfein were having or the expertise to do so. One could argue that, at the time, Citi’s structure may have been more hierarchical and more complex than Goldman’s (Citi had more than 300,000 people in more than one hundred countries), making it harder for information to get to the top. Other elements, such as dissonance, may also have been missing, adding to structural secrecy and restricting information flow.
Goldman, by contrast, encourages the discussion and disagreement needed to arrive at the best answer. For example, when I was in proprietary trading in the early 2000s, I participated in meetings wherein the portfolio managers of proprietary trading groups presented their ideas and strategies to other proprietary portfolio managers in other groups. Once, a manager presented a trade that appeared to be a good investment based on models, but the other proprietary traders quickly ferreted out a weakness in the model: it was not properly accounting for the illiquidity and the severe downside that would happen in selected, albeit low-probability, scenarios. It is hard to believe that a risk management department or accountant expected to value the investment would be able to identify all the issues raised in that room of traders.