The manner in which Jeff Flowers and David Friend, Carbonite’s founders, expanded beyond their well-tested software skills as they entered the small business market exemplifies sensible risk-taking. Sara Campbell, too, applied the tenets of sensible risk when she launched new retail stores as a destination for finished product that she was contractually obliged to produce. We have also seen how ignoring or violating those same tenets—as Genzyme did when it postponed construction of a new plant or as Sara Campbell did when she trusted Talbots’s sales projections—can unravel years of solid strategic planning. We can observe how the tenets apply to the challenges facing a wide range of enterprises, despite the obvious differences among both the types of risk and the organizations.
The following chapter showcases several risk-related dilemmas. After posing the problem in the first several sections, we’ll revisit the scenarios to observe the outcomes and how they might have been affected if the individual decision makers had properly evaluated the risks by applying the four tenets.
When George Bell accepted the job as CEO of Excite.com, an Internet portal and search engine, in 1996, he was a wildlife and adventure filmmaker with deep experience in traditional media production. Although George was a Harvard-educated athlete with considerable energy, wit, and enthusiasm, he knew little about the Internet revolution he was being asked to join and help lead. He told me that he thought Kleiner Perkins, the venture capital powerhouse behind Excite, was the name of a law firm.
George Bell was not the only outsider moving into a starring role in this wild new world. The entire Internet was brand new, and there was no one already endowed with the technological, strategic, and marketing skills required for success. The managements of these new ventures, whether it was AOL, Yahoo, eBay, Lycos, or multitudes of others, were scrambling to grow as quickly as possible to preempt competition, grab the best talent available, close partnerships, ensure content for their websites, and promote like crazy.
The challenges for George were to focus on the correct market, right size the resources to capture and retain that business, and remain grounded in reality. The risk was that, without any definition or concept of what long-term success would look like in this arena, all the bushels of money investors were pouring into dot-com start-ups might not save them.
In 2004, Larry Summers, president of Harvard University, participated in discussions with members of the ruling Harvard Corporation and his financial staff on how to finance the expansion of the Allston campus, across the Charles River from the central campus in Cambridge. Somehow, they reached a consensus to lock in the prevailing 4.7 percent interest rate for bonds that they would issue four years in the future to finance this new construction project. The vehicle that allows for future rate setting is called a “forward swap,” arranged by investment banks that will exchange whatever rate Harvard would have incurred without the swap four years later with other borrowers. Swaps are not uncommon in debt financing, but both the $2.3 billion size and the four-year duration were very unusual.
This case has been covered in great detail by the press, although not from the perspective of the major participants and how they might have viewed the risk if interest rates turned down rather than moved upward as expected by almost all experts. I interviewed several people who were employed by Harvard at the time, most of whom spoke on condition of anonymity. It seems unlikely from my conversations that anyone in power was particularly worried that interest rates would plummet; if they were, they were afraid to voice that opinion and confront President Summers or his close allies on the swap proposal.
What is clear is that if the school were able to save 2 percent per year for thirty years on a $2.3 billion loan, that would represent $46 million a year or almost $1.4 billion over the life of the bond. However, Harvard and other institutions typically issued long-term bonds with a call provision, allowing the university to redeem the debt after a fixed number of years and issue it again at lower rates. So the benefit might be worth five or ten years’ worth of lower interest payments, still a savings, but nothing close to $1.4 billion.
What would Harvard do if the direction of interest rates did not follow the script and climb over the next few years? This was the risk, however remote, that the administration might need to address.
No one would equate the risk of hedging billions in debt with the saga surrounding one employee at a regional real estate company, but the lessons about anticipating and reacting to risk are quite similar. When Steve Starr* hired Eve Ferrante† as the commercial leasing manager for his company in Austin, Texas, she had strong recommendations and appeared to have the requisite experience in the market. Starr Realty was growing quickly, and he was happy to have Eve.
Within six months, Steve realized something was amiss. There were irregularities in Eve’s e-mails, missed appointments, and sloppy reports. In addition, Eve began to take very long lunch hours and behave erratically toward her colleagues. Steve took her to lunch to talk things over as well as to try a test that he had heard a friend describe as effective: let the other person drink whatever and however much she wants. You’ll hear whatever is on her mind and will know if you’re dealing with a problem drinker. Eve had too much to drink at lunch and expressed considerable enthusiasm for the fantastic waitstaff at a favorite after-work hangout near the office.
A few days later, Steve had dinner at the restaurant in question and asked the manager, who was a friend, if he saw Eve there often. Colin, the manager, told him that she was very well known there, and frequented the bar at lunch or after work several days a week.
When Steve reviewed Eve’s work further, he discovered that she had failed to send contracts and related paperwork, jeopardizing several important deals. Steve took care of the problem but realized that his business’s reputation could be irreparably harmed if Eve’s negligence continued. When he confronted her, she became defensive about her behavior and disagreed that there were lapses in her work. Eve was not getting along with her colleagues or clients, she was irresponsible, and her work was unreliable—Steve knew that he had to let her go.
Steve’s challenge was to separate Eve from Starr Realty without her causing much damage. Based on comments she’d made around the office, he felt it likely that Eve would sue for gender discrimination. When he told his attorney about the situation, his lawyer agreed that the sooner Steve let her go, the better. He was not going to allow Eve’s problems at work destroy his company.
The risks ranged from the possibility that Eve’s unreliability would damage the business if she stayed on, to her potential retaliation upon dismissal. Steve feared a long and expensive lawsuit that would surely impact his reputation, budget, and time. He needed to decide what to do.
Like every new business, Aureus faced many hurdles in the first years, and we tried to apply a practical approach to these. Having seen many friends, colleagues, and acquaintances start all types of financial services companies, I had the opportunity to watch and learn from their successes and failures. I had not yet clearly identified the four tenets that I now apply to risk.
The goal of every financial services firm is to have enough assets under management that the fees more than cover expenses, including the salaries of the founders. The primary way to attract additional clientele is through strong performance. The biggest mistake I saw repeatedly was entrepreneurs creating a high-cost infrastructure before growing their revenues sufficiently to cover such expenses. Short-term, weak results due to a market slump or poor stock or investment selection would hurt the top line, and if the fixed costs were too high, the firm might never even cross that breakeven level.
David and I expected to take no salary for the first year, or maybe two. We right sized our rent and office space, which is very expansive and beautifully built, thanks to our landlord. The cornerstone tenant had left the building three years earlier, leaving 75 percent of the space vacant, and our landlord was desperate. In addition, we right sized our initial headcount, our market data services, trading and custody platform, and all our other service providers, once we sorted out our initial misstep with a very high-priced corporate law firm.
Fortunately, our formation allowed us to capture three more strong years in global stock markets. That was lucky, of course, but the economy was strong and both David and I felt that the momentum, over the next few years, was in our favor. Our collective experience among the original partners was in managing stock portfolios and, to a lesser extent, alternative investments, such as hedge funds. We diligently learned all about compliance, followed all the rules implicitly, and developed procedures for trading, reporting, security, and client service.
As for skepticism, my middle name should be “cynical.” At Fidelity, I learned that the great investors have tremendous capacity to visualize incredible new concepts, products, and services for the companies in which they invest, but they are always intensely clear on every detail of the model illustrating the road to such lofty goals. There is no unbridled optimism without deep inquiry and analysis.
The second way to grow and retain business in our industry is to develop a reputation for competence in all aspects of client service. In 2011, despite all our attention to process and execution, we hit a speed bump with damaging force. We manage assets for very wealthy people and sometimes they spend a bit of money. A large number of our clients have us send regular distributions from their accounts, but occasionally they ask for sums outside the norm—these may be for a charitable donation or a major purchase for themselves or a family member.
It was, therefore, not particularly noteworthy when Nicole,* our director of client services, received an e-mail from Christine,† a long-term customer who travels extensively, explaining that she needed funds for a few large Asian antiques while she was traveling in the Far East. They exchanged e-mails, and Christine sent Nicole the wiring instructions to a bank in South Korea from which Christine could access the cash; then Nicole loaded the transfer and hit “send.”
Almost immediately, Nicole received another e-mail from Christine, thanking her but also mentioning that she had found some other beautiful items to purchase and needed another wire. At this point, Nicole got goose bumps, and had what she later described as an “Oh shit!” moment, when she realized that she had perhaps been duped by a hacker.
Nicole ran to my office and recounted the story, and I asked her to show me the e-mails from “Christine,” who it seemed increasingly unlikely was our client. Upon careful reading, it was clear to me that the writer was not a native English-speaking person, but Nicole had many quick e-mails from Christine over the years about purchases in different parts of the world, sometimes hastily written, and she had not noticed the improper grammar. We called David to join us and placed a call to Christine, asking her to call me right away, then called Fidelity, our custodian, to see if the transfer had gone through (of course it had!), if they could intercept it before the recipient withdrew the money (already gone!), and if they would call their contacts at the FBI (they would, but good luck!).
We needed to decide immediately what to do when Christine called back. We had to tell her the truth, of course, and inform her she needed to change her e-mail password immediately. However, we had just wired a large amount of money into cyberspace and had to make that right. I needed a few minutes of quiet to think about the risks of different actions, so I asked Nicole to stay in touch with Fidelity for any news, and I would have my decision very soon.
For George Bell, the most critical risk was in choosing the direction for Excite. The explosion of the Internet made it very difficult to predict which model would succeed, and euphoria abounded wherever you looked. He and the venture capitalists who had helped fund the enterprise ultimately bet on building a portal and search engine. The search space was jammed with competitors, among them AltaVista, AOL, Yahoo, Infoseek, HotBot, Magellan, Lycos, WebCrawler, and Ask Jeeves. Most of them no longer exist.
The dominant player, as we now know, became Google, because its technology, superior to that of its predecessors, produced search results so accurately and quickly that it became the most popular. Famously, Sergey Brin and Larry Page, two Stanford students who founded Google, offered their company to Excite for $1 million in 1999, but George, in a decision that may always haunt him, declined. He was intent on preserving the cohesive Excite culture, which he felt would be disrupted by a Google acquisition.
In hindsight, the risk to Excite’s culture would probably have been worth taking, at least by one measure, based on Google’s current market value of over $360 billion. However, it was almost impossible, in the late 1990s, to quantify the scope of risk in this environment because of the insatiable demand by investors who bid up the prices daily on Internet public companies. George acquired competitors Magellan and Webcrawler with Excite stock and built a large campus of buildings to house the expanding Excite workforce. In early 1999, he accepted a $7 billion offer from @Home, a high-speed Internet service provider that partnered with cable companies such as ATT and Comcast, keeping a third of the monthly subscription fee. George was asked by the controlling board to become CEO of the recently merged companies, and he agreed.
The most significant tenet that Excite (as well as hundreds of other web start-ups) violated was lack of skepticism. As George Bell describes it, those in the industry were all undiluted optimists who believed that ad revenue would mushroom, cable deals would expand exponentially, and business growth was in the very early innings.
Nearly all participants—whether they were the executives of dot-com companies, mutual fund and hedge fund managers, or typical Americans watching their 401(k) statements—trusted the forecasts. The stocks surged every day, which seemed to be ample proof. Unfortunately, virtually none of the forecasts made by companies or analysts in 1999 became reality in the next few years.
At its peak, Excite@Home was worth nearly $35 billion in market value, with revenues of more than $600 million but an annual loss of $7.4 million. As a comparison, Colgate, Comcast, and United Technologies all traded at market capitalizations in the same region but had sales upward of $9 billion, $8 billion, and $26 billion, respectively, with net profit between $1 billion and $2 billion each.
During this bubble, companies such as Excite needed to keep expanding ahead of their competitors, adding people, acquisitions, and offices, but it was unclear where the strategy was leading. The pace of megadeals and acquisitions in 2000 barely allowed anyone time to safely jump off the ramp of the private jet and into the limo. Excite@Home was a victim and a participant in this land grab.
When Yahoo partnered with Google to power its search engine, however, it handed Google the keys to the search empire. Players across the industry were battling for dominance in their niche.
Excite@Home initially held an unassailable position as the provider of broadband services to the cable companies; however, a major regulatory issue, referred to as open access, began to send shudders through the cable industry. The FCC conducted hearings on whether broadband availability should be treated as a public utility rather than through exclusive contracts. Although open access was never legislated at the federal or state level, the presumption was that Excite@Home’s pricing structure would be challenged, undercutting the value of its franchise. The cable companies swiftly began offering their own broadband packages, bypassing Excite@Home.
George Bell needed to differentiate his company’s service, but he became increasingly frustrated by the intrusive directives from executives at the company’s cable customers. The backdrop of a crash of the technology and Internet stocks made his efforts much less likely to succeed. He decided to leave the firm in April of 2001 and move back to the East Coast with his family. George’s replacement was Patti Hart, a longtime veteran of the telecom industry, with close ties to ATT and other service providers. Excite@Home’s stock plummeted from a peak of $128 per share to $1 within two years, amid the Chapter 11 bankruptcy filing that occurred more than a year after George Bell had departed.
Excite@Home never exploited a unique expertise to preempt competition. Perhaps it could have developed a lock on a market, as Amazon, eBay, Yelp, TripAdvisor, Priceline, or PayPal had done, but the young creators of the Excite portal were innovators, competing concurrently with a handful of other groups, all aiming to attract new web users to their site.
Even if George Bell and his team had been skeptical of their own forecasts, it is not clear that they would have identified a new direction or prospered if they’d pursued it. Since returning to the Boston area, George has been CEO of two companies, his tenure at both ending with the successful sale of the enterprise. In addition, he has been a partner at General Catalyst, a venture capital firm.
The Harvard treasurer’s office kept a careful watch on the value of the swaps they had placed to ensure the 4.72 percent rate on over $2 billion of debt the university would issue four years into the future. Although long-term rates did inch over 5 percent briefly in 2006 and 2007, they began to slide in 2008, and then simply crashed to around 2.5 percent as a postscript to the demise of Lehman Brothers, Bear Stearns, and numerous other financial institutions.
Most borrowers would be thrilled with a weak interest rate environment, but Harvard had locked itself into a 4.7 percent borrowing cost for a very large absolute quantity of debt. In addition, the school had allocated 80 percent of its cash reserves, or $6 billion, to be invested with the Harvard endowment, another decision that defies right-sizing principles, considering that this is a school with massive, fixed operating costs. That bet was also crumbling, and Harvard authorities, gripped by panic at the end of 2008, chose to terminate the swap agreements and buy them back from J. P. Morgan, among other banks, for close to $1 billion over the next few weeks. Considering the already disastrous period for banks, this could have been their best business, by far, of the holiday season.
Rather than dissect why Harvard so cavalierly played these odds, it is worth considering the tenets as they apply to this case. No one involved with the decisions at the time has explained who exactly pushed for the size of the swap trade or the cash investment, which lost $1.8 billion by early 2009, and then pulled the trigger on both.
The sheer size of the swap, at $2.3 billion, was larger than Harvard or, probably, any university had ever issued previously. Given the uncertainly about the interest rate outcome, this was too large a bet, in the opinion of several economists and endowment managers with whom I spoke. Such an aggressive approach suggests arrogance, which can be harmless in small doses but often dangerous in larger ones. The four-year horizon was also much longer than that of typical swaps.
Larry Summers, the university’s president at the time, is a distinguished economist. Among his colleagues and advisers were fixed income experts at Harvard Management Company, Harvard Business School, and even the Harvard Corporation. However, they had no experience with the size and duration of this swap instrument. As trained academics and practitioners, they were well aware that predictions about bond rates several years in the future are very likely to be wrong.
The Harvard power elite, including Larry Summers, believed that the bond market would fall, rates would rise, and the value of the Harvard endowment would continue to climb. The models supporting these predictions were meticulously developed, highly complex, and incorrect. There was probably not enough skepticism about these models applied to the discussion before putting on the swaps as well a flawed understanding of the extent of possible damage to the university on such a large bet.
Even when the swap value plummeted and the corporation chose to repurchase and unwind the option, its timing was off. A few months later the cost would have been hundreds of millions less.
Though he was looking at a risk on a scale much smaller than that of the Harvard gamble, Steve Starr knew that he needed to let his commercial leasing manager, Eve Ferrante, go. She was unreliable, a problem drinker, and difficult with coworkers.
Steve orchestrated a meeting for the two of them in his office. After one sentence, she knew she was being fired, and let loose a string of profanities worthy of any construction crew, grabbed whatever she wanted in her office, and told Steve he’d be sorry. Two weeks later, Steve received notice from Eve’s attorneys that Starr Realty was being sued for several forms of discrimination. He met with his firm’s lawyer to discuss their strategy.
Eve was seeking a huge amount in damages. She had no case, but Steve needed to consider the potential downside. Her attorney was probably on a contingency, but Steve would be paying out of pocket and the legal bills could grow out of control. His firm’s reputation could be hurt for years to come. Importantly, this case could consume considerable time and energy, resources that were already in short supply.
Steve needed to right size the case and determine the most he should pay to settle. As a married mother of three, Eve, he assumed, was not anxious for anything to come to light about her poor judgment or lunchtime and after-work habits. The right timing of a settlement was as soon as possible. He was sure that Eve had nothing up her sleeve to throw at him; she was the one with spotty attendance, lack of follow-through, and a toxic attitude.
Steve’s lawyer made an offer that was slightly more than he wanted to pay but much less than the expense of a protracted court case. She took the money, and Starr Realty escaped with its reputation intact and its collegial office culture restored.
I sat in my office with the door closed, deciding what to tell Christine, the client from whose account we had mistakenly wired a six-figure amount (low six figures, honestly) to a thief, most likely in Asia. I remember looking at the clock on my phone, putting my head in my hands, and then two minutes later looking at it again.
I called David and told him that we needed to refund Christine’s account fully with our company’s cash reserves. He agreed that it was the right thing to do. We have always maintained a large cash balance just in case something unexpected happened, and now it had. If we recovered the funds, Aureus would keep them, but the chance of that was quickly dwindling to less than my likelihood of hitting the lottery. Christine would think much more highly of us this way than if we waited, and there was no reason to wait. If she decided we were idiots and pulled her account, she would not think less of us if we refunded her lost assets immediately.
No insurance company will cover hacked funds other than at a premium that makes the coverage irrelevant, and the fine print, which Stacey, my longtime assistant, and I read after this event, exempts claims for theft originating in many developing countries and most of Asia.
I called Christine and she picked up this time, having just heard my voice mail asking her to call me. I explained exactly what happened, and that we were refunding her account the exact amount lost on that day. We spoke for a short while longer, as I suspect Christine was in as much shock as I was, and I assured her that we were going to review and change our procedures for wiring funds to prevent this from ever happening again. She thanked me and we hung up. It was unclear whether Christine would stay with us or not.
Computer hacking is a thriving industry, and we were easily duped in a well-known scam. Despite our embarrassment, I am sharing this story because so many other companies have been the victim of hacking; I feel slightly less stupid in the company of major corporations like Target, Visa, and Anthem, as well as various branches of the U.S. government. However, we should already have instituted tighter rules about distributions from accounts.
We had been blind to the significant risk posed by the possible hacking of a client’s e-mail. Our policies about wiring assets changed that day, and now require a phone call and voice confirmation for any withdrawal of $5,000 or greater that is not a standing distribution for that account. Most banks and investment firms now have similar restrictions. These rules may be inconvenient, but they represent an appropriate safeguard.
Hopefully, the measure we have in place across the assets under our care will protect us against the next cyber attack. We have improved our diligence and security systems, and know our custodians have done the same.
Christine was appreciative of the rapid steps we took to reimburse her account, and has referred other clients to us. We applied right timing to the risk of further damaging our reputation and I have never regretted that move. The FBI investigation failed to uncover more than one vague lead, which was disappointing but not unexpected. Christine’s confidence in our firm confirms that we acted correctly to manage one risk, even though we totally failed to control another.
Major decisions at businesses or nonprofit organizations must include an evaluation of the scope of the exposure and the time frame involved. We must know as much as possible about the marketplace, the participants, and the customers, while remaining skeptical about research that seems overly optimistic or simplistic. None of this is easy, particularly in the heat of a crisis, which is often when we are forced to make our most important decisions. However, keeping in mind a framework to address risk can help assuage those challenges.