Felix Salmon

Sensible data

Felix Salmon
Jan 14, 2014 20:40 UTC

I have an essay in the January issue of Wired about the limits of quantification. In the magazine it’s headlined “Why Quants Don’t Know Everything”, but online it’s been retitled “Why the Nate Silvers of the World Don’t Know Everything” — which is a little unfortunate, since the whole essay is deeply indebted to Silver’s book, which makes substantially the same point.

The initial idea behind the essay was the concept of priests vs quants — the seemingly eternal (but actually only quite recent) distinction between people who trust numbers, on the one hand, and, on the other, people who rely instead on their personal expertise and experience. Think of the difference between Billy Beane, using dispassionate analysis to outperform in baseball, and Bobby Fischer, whose gifts at chess seemed almost god-given. Nowadays, in a world of quasi-infinite data, it seems the quants are in the ascendant, while the priests are reduced to fighting a rearguard action, clinging desperately to some vestige of relevance. Today, if you want to change someone’s mind, you don’t appeal to authority: instead, you bring numbers.

The result is a deep societal disruption, in which quants take on priests and win: the Oakland A’s against the Yankees, the Obama team against Romney’s. It doesn’t take long before the war is won — we’ve all seen this particular movie before, especially the kind of people who sit on boards of directors. Thus does the priesthood wither away, taking with it a huge amount of valuable institutional knowledge.

The rise of the quants is, unsurprisingly, one of the driving forces behind Silicon Valley venture capital: if you start a small company which competes with a huge company, and the small company gets a few significant wins, then there’s almost no sum the bigger company won’t pay to acquire its smaller foe and use those skills against its competitors. The small company never needs to make money: all it needs to do is show disruptive potential, and it becomes enormously valuable.

But the data-rich narrative — the idea that science is taking over the world — has bred its own counter narrative for some 200 years now, ever since Mary Shelley published Frankenstein in 1818. People don’t like the idea that the computers are in control: for a prime example, look at the way Twitter exploded with privacy concerns as soon as it was announced yesterday that Google was buying Nest.

Or, just look at popular entertainment. The dweeby Q, in Skyfall, tells James Bond that “I can do more damage on my laptop sitting in my pajamas before my first cup of Earl Grey than you can do in a year in the field.” But Q isn’t the hero: Bond is. Similarly, it’s incredibly easy to paint Wall Street quants as the big villains in financial-crisis stories.

When the quants come into an industry and disrupt it, they often don’t know when to stop. They’re young, they’re arrogant, they’re rich and powerful – and they tend not to have decades of institutional knowledge about the field in which they have found themselves. They don’t work for the people who know such things, and they don’t listen to them, either. They’re winners, what do they have to learn from dinosaurs?

Put like that, the risks are obvious. Quants are just as blinkered, in their own way, as the priests they replace – even more so, in fact, since they can be quite Spock-like in their inability to understand the deep role that certain institutional functions are playing. Quants are great at coming up with clever ways of analyzing the world. But that doesn’t mean they’re great at managing institutions, or understanding how their employees might end up gaming the systems that they’ve created.

The solution to such problems is not to disdain the quants, as Bond does Q. Rather, it’s to synthesize the best of both worlds. Look at Southwest Airlines, for example: it has some of the most sophisticated operations geeks in the business, governing everything from fuel-price hedging to the most efficient way to board an airplane. But that doesn’t stop it having a much more human face than its larger competitors. Apple, too, is a prime example: quantitative to its toenails, it nevertheless is clearly governed by overarching principles of human-focused design. And why did Nest succeed where Google Power Meter didn’t? Just because it was designed in a way which made it desirable as a consumer product.

The secret ingredient, I think, is to ensure that managers have a deep understanding of the science being used in the organization — and also of its limitations. Great technology is all well and good, for instance, but if you want it to be broadly adopted, then you need a whole other set of packaging skills as well. And you can’t take technology further than its natural limits, either. It wasn’t really the Gaussian copula function which killed Wall Street, nor was it the quants who wielded it. Rather, it was the quants’ managers — the people whose limited understanding of copula functions and value-at-risk calculations allowed far too much risk to be pushed out into the tails. On Wall Street, just as in the rest of industry, a little bit of common sense can go a very long way.


You can’t discuss this question without talking about short-term vs. long-term thinking.

Most of the problems in the world today, including those related to quantification, can be traced to the glorification of short-term thinking at the expense of long-term thinking.

In essence….maximize profits and damn the consequences.

Posted by mfw13 | Report as abusive

The incredible shrinking company

Felix Salmon
Aug 22, 2013 17:26 UTC

Christopher Mims has a good piece on Meg Whitman’s Hewlett-Packard today, pointing out that the company’s success (at least as measured by its stock price) over the past year or so is in large part due to her cost-cutting abilities.

Mims doesn’t mention that this strategy is hardly new for HP: it was executed with just as much success under Mark Hurd. The really big errors at HP have taken place at the board level, both in terms of hiring and firing CEOs and in terms of approving ridiculous over-ambitious acquisitions. When it puts its mind to it, HP turns out to be pretty good at shrinking. It’s growth which turns out to be where the pitfalls lie.

Which is why I’d take issue with Mims here:

A more efficient, better-managed HP doesn’t mean much in the long run if the company cannot move beyond the stagnating businesses that make half of its current revenue—PCs and printers…

Without any innovative consumer products to speak of, HP is essentially at the mercy of big businesses’ appetite for technology. And they seem to be wanting less and less of whatever HP is selling…

While she’s managing HP’s consolidation and retreat admirably, is she the CEO who can get the company to once again break the mold? Given her background as head of eBay, a company that changed little under her decade-long rule, we have to be dubious.

I can understand where Mims is coming from. HP is a technology company, and under the unspoken rules of the US stock market, all public companies, and especially all technology companies, should constantly be growing as fast as possible. It’s the inexorable mathematics of discounting: if a company can deliver consistent growth which is faster than the prevailing discount rate used to calculate net present value, then its stock price should, by rights, be infinite. Consequently, given that infinite upside, it’s worth risking quite a lot to achieve growth.

But the facts are pretty plain: (a) HP is very good at producing excellent products in the shrinking markets which make up most of its business right now; (b) HP has in recent years shown no particular ability to produce excellent products in other markets; (c) Meg Whitman is not by nature a visionary innovator. Given those facts, it makes perfect sense for HP to run its existing businesses as efficiently and as profitably as it can, and to extract as much value out of them as possible, in the knowledge that all companies are mortal. In fact, it makes more sense to do that than it does to follow the Tim Armstrong playbook, where AOL’s CEO decided to take his enviable dial-up revenue stream and invest it in doomed content plays like Patch.

John Kay had an excellent column on this subject yesterday, under the headline “sometimes the best that a company can hope for is death”. He finds the genesis for the Mims view in the work of marketing guru Theodore Levitt, who said 50 years ago that companies can and should reinvent themselves in imaginative ways. Levitt was wrong, says Kay: what really matters is not imaginative executives, but rather competitive advantage. HP has a competitive advantage making PCs and printers and servers; that doesn’t mean that it can readily apply that competitive advantage elsewhere.

Kay’s example of JC Penney is a good one. Bill Ackman looked at Penney and saw it as a company in the “retail” bucket, just as Mims looks at HP and sees a company in the “technology” bucket. Ackman then looked for the biggest retail success story he could find — Apple — and tried to inject Apple-style success into JC Penney by hiring Apple’s Ron Johnson as Penney’s CEO. That didn’t work:

The outlets of both JC Penney and Apple are shops but the age group and disposable incomes of their customers – and their reasons for visiting the stores – were entirely different: JC Penney and Apple were not really in the same business.

Penney was not going to succeed by looking at successes elsewhere in the retail space and trying to copy them with little regard for its own existing strengths. And similarly, if HP decides to “break the mold” again, the inevitable consequence will be yet further billions in avoidable write-downs. Once upon a time, HP had a legendary R&D operation — but that operation fell victim to Hurd’s cost-cutting, and in any case no R&D shop can maintain excellence forever. The HP board knows that, which is why they were open to ambitious acquisitions; they simply failed to notice that the acquisitions in question were fundamentally really bad deals.

In general, the stock market’s bias towards growth makes it very difficult for a public-company CEO to execute a strategy of shrinking profitably. Maybe at some point, as a result, Whitman will decide to take HP private, or will accept a takeover offer from a private-equity shop. But for the time being, an ambitious growth strategy is the last thing that HP needs.

It’s not easy to excel in a shrinking business: as Kay says, it goes against human nature to accept that there might be a natural life cycle for a company. But if you work with the cycle, rather than against it, it’s still possible to get extremely impressive results. Just look, for example, at Lehman Brothers International Europe — the bankrupt entity which has been spending the past five years quietly trying to repay its creditors as much money as it can. The latest news? Those creditors are going to be repaid 100 cents on the dollar — including statutory interest at 8%. As a result, some Lehman claims are trading at as much as 120 cents on the dollar.

The best thing for Whitman to do, then, is to accept that HP’s core business is in decline, but to still execute that core business as efficiently as possible. That takes a very different style of management than Silicon Valley is used to, but it can be done. And it’s much more likely to work than some Hail Mary acquisition attempt.


What percentage of total Lehman Bros. was LB International Europe? It looked like their liabilities were $8B which would be in the single-digit percentage of the full Lehman Brothers company.

Posted by winstongator | Report as abusive

Why Mark Zuckerberg shouldn’t listen to management gurus

Felix Salmon
Feb 10, 2012 20:54 UTC

This is why Mark Zuckerberg was smart to stay in complete control of Facebook and not listen to anybody telling him that a multi-billion-dollar company needed a seasoned, professional CEO in charge.

Jack and Suzy Welch are onto something when they diagnose a potential class problem at Facebook, post-IPO.

After its IPO, Facebook is going to have two classes of citizens. That’s just reality. Some of its 3,000 or so employees — several hundred in number by some counts — will have significant riches in the hand. Newer hires, though — well, they’ll mostly have options in the bush.

Where they go hilariously wrong is in their proposed solution to the problem. There’s a carrot, which as far as I can tell involves Silicon Valley manager-geeks suddenly transforming themselves into motivational speakers. And then there’s a stick:

With all this exultant “barking,” there also needs be bite — in the form of frequent, rigorous performance reviews. The facts are, if Facebook wants urgency, speed and intensity around its mission, those behaviors must be explicit values that, when demonstrated, result in bonus money and upward mobility — or not.

Any company, in the wake of an IPO, finds itself growing new and previously-unnecessary layers of management, especially in areas like the general counsel’s office, investor relations, and public relations. But for the Welches, that’s not enough: extra management also has to be marbled throughout the organization, to be found everywhere as “frequent, rigorous performance reviews”.

There is absolutely zero evidence that frequent, rigorous performance reviews ever do any good, and quite a lot of reason to believe that they actually do harm. And what’s true of business professionals in general is especially true of Silicon Valley engineers — a culture where pretty much everybody knows exactly who’s hot and who’s not, without any need for formal, frequent, or rigorous performance reviews.

What’s more, it’s far from clear that the best way to motivate a Silicon Valley engineer is to dangle an annual bonus in front of his face and tell him that if he works hard he could get an extra couple of months’ salary at the end of the year. Rather, the best way to get the most out of engineers is to surround them with other great engineers, in a collegial atmosphere where everybody works hard and everybody does really well building great products that everybody is proud of.

Performance reviews are horrible, divisive things which create a whole other set of class distinctions within a company, between the “high performers” who get money and promotions and the grunts who live in fear that their review will be used to punish or fire them. (And of course if bonus-greed isn’t a great motivator of computer engineers, fear is even worse, especially in the context of Silicon Valley, where there are multiple jobs permanently being dangled in front of just about anybody who can code.)

Managing a company like Facebook is all about creating a magnetic culture — a place where employees love to work, and where they’ll tell their friends that they’re having a great time and that they should come join them. Meanwhile, there has never been a company in the history of capitalism where managers really love the performance-review process and tell all their friends that they would hugely enjoy going through it themselves on a frequent and rigorous basis.

In other words, the mere existence of such things would probably be enough to put off many talented potential employees with a wide choice of possible employers. At a company like GE, a CEO like Jack Welch tends not to worry about such things. But at Facebook, the ability to continue to attract Silicon Valley’s best coders is very high up Mark Zuckerberg’s list of priorities and concerns.

Which is reason number 1,452 that all of Facebook’s shareholders should be very happy indeed that the company is being run by Mark Zuckerberg and not by Jack Welch.


You’re absolutely right!

Posted by killad1508 | Report as abusive

How Rajat Gupta corrupted McKinsey

Felix Salmon
May 19, 2011 15:00 UTC

Just how corrupted was the culture of Rajat Gupta’s McKinsey? Suzanna Andrews reports:

McKinsey had a culture of superiority, says one longtime client, who declined to be identified, adding that consultants at the firm really seemed to think they were better than anyone else in the business world. This CEO is still shocked recalling an incident in the late 1980s, when a McKinsey team offered to provide him with a road map of what his competitors were doing. When asked how they could produce such information, he was told that McKinsey also worked with his competitors, but he could trust McKinsey to know what was confidential information and what was to be kept private. He says arrogance permeated the firm. The usual rules seemed not to apply. When this CEO listened to a wiretapped phone call from July 2008, in which Gupta relayed to Rajaratnam the details of the Goldman board’s discussions about buying a commercial bank, it sounded to him just like Gupta consulting a client.

Gupta seems to have been directly violating internal McKinsey rules for years:

As the managing director and then as senior partner of McKinsey for four more years before he retired, he ran his own consulting business on the side — a violation of McKinsey rules…

While Gupta was devoted to his philanthropy in India, his quest to amass great wealth led him to lapses in judgment, says Bala Balachandran, dean of the Great Lakes Institute of Management in Chennai, India, and a friend for almost three decades.

“He wanted a billionaire’s life and the question for him was how could he become a billionaire in a short time,” Balachandran says.

If Andrews’s CEO was shocked by what a presumably-representative McKinsey team would do in the course of normal business, one can only imagine what Rajat Gupta and Anil Kumar got up to at Mindspirit, the company set up by their wives (!) to consult for InfoGroup CEO Vin Gupta — or what services Gupta provided to Genpact which resulted in him getting 81,405 stock options at a strike price of less than a buck apiece. Genpact, whose sole client is GE, another McKinsey client, is currently trading at $16.88 per share, which means that Gupta’s options in the company are worth well over a million dollars.*

And Andrews does put forward one explanation of why Gupta might have been giving such valuable information to Raj Rajaratnam. It’s not because he was being paid for it directly, but rather because he had much bigger ambitions: he was negotiating with Rajaratnam for a 10 percent to 15 percent stake in the Galleon International Fund in exchange for attracting investors and becoming the fund’s chairman. And the wiretaps don’t end there:

Listening to the pleading tone in Gupta’s voice as he pitches himself to Rajaratnam is almost painful. “I can be helpful in Galleon International, by the way—not Galleon International, Galleon Group,” he says, apparently angling for a bigger job. “I mean you’ve given [me] a position in Galleon International, that’s good enough. I, I … ,” he breaks off.

It seems clear that Gupta had various personal crises and issues which obviously aren’t reflective of McKinsey as a whole. But at the same time, this quiet man who described himself as a servant to McKinsey’s partners was clearly never going to crack down on any of them if and when they started pushing the envelope in terms of fishing for clients by trading in confidential information. Unless and until some heads start rolling within McKinsey, it’s fair to assume that many of the most successful Gupta-era partners remain. And we’ve been given precious little reason to trust in their integrity.

Update: Gail Marold of Genpact responds in the comments, saying that Gupta provided no consulting services beyond his capacity as a director. And although Genpact did start life as a division of GE, it now gets 62% of its revenue from non-GE clients.


Felix, on behalf of Genpact, I must clarify the inaccuracies in your post. First, I have provided the following statement to Bloomberg Markets and Businessweek, the originators of these feature stories on Mr. Gupta:

“Rajat Gupta was a non-voting advisory director for Genpact from 2005-2007 and a voting director and chairman of our board from March 2007 to March 2011. Contrary to what is stated in these recent articles, Mr. Gupta was not at any time engaged by Genpact to provide consulting services in his individual capacity. Mr. Gupta’s association with Genpact has always been public information and all compensation paid to Mr. Gupta by Genpact for his services as a director has been fully disclosed as required by the rules of the US Securities and Exchange Commission. “

Further, the options that Mr. Gupta received from Genpact during his time as an advisory director were designed to not vest until he became a full-time director, as he did in 2007.

Second, as background, Genpact (NYSE: G) is a $1.2B global business services company that was founded as a division of GE in 1997, now with more than 400 clients around the world and more than 48,000 employees now that we recently made a $550 acquisition of Headstrong. Per our year-end 2010 financials, revenues from GE businesses comprised 38% with 62% coming from other global clients. We continue to increase our revenues from both GE businesses and global clients.

It is important that the facts are reported correctly. Thank you very much.

Gail Marold
Genpact Public Relations

Posted by gailmarold | Report as abusive

Annals of C-suite dysfunction, Goldman Sachs edition

Felix Salmon
Apr 8, 2011 21:45 UTC

Ian McGugan has a good review of Bill Cohan’s huge new book on Goldman Sachs which includes an intriguing quote about how Bob Rubin “encouraged a culture of undisciplined risk taking” — something which goes directly against the reputation he’s spent many years cultivating. It comes from Chapter 15, which starts in the dangerous year of 1994 and which is full of juicy gossip about the very human frailties of the people running Goldman. Here’s more of it:

“For a long time he just sort of sat in his office,” one partner said of Corzine. “He would sit in his office breaking out in tears at various times while the firm was losing all this money.”…

As the losses in 1994 mounted, many partners became increasingly nervous that the firm was at risk… Some forty partners left Goldman at the end of 1994, the first time anything like that many partners had voted with their feet. “People resigned out of fear,” one partner said. “That should tell you something.”… Howard Silverstein, the partner in charge of Goldman’s Financial Institutions Group, left. “He was perceived as being an expert,” one partner on the Management Committee said. “And all he did was just do a simple calculation if this continues. You know: wiped out.”…

Paulson cut people, travel expenses, allowances for overseas living, and many of Goldman’s vaunted perks. He even cut back on the use of a corporate jet and recalled grueling overseas trips flying around Europe and Asia on commercial flights…

Goldman’s problems at that time weren’t only ones of cost and bad bets. A culture of undisciplined risk taking had built up over many years. “A lot of these practices were set up when [Rubin] was there,” one top partner said. “Okay? The lack of a risk committee, trusting individual partners, model-based analytics — that by God you can be smart and figure it all out — and letting traders become too important and being afraid to confront them if they’ve been big moneymakers. All that sort of stuff built up.” …

Aside from why Friedman had seemingly botched his departure, the other lingering question that remained among many of the Goldman partners was how Corzine could have emerged as the firm’s leader when he was leading the very division — fixed- income — that had lost hundreds of millions of dollars in 1994…

Goldman had selected as its new leader the very person who had just presided over a complete meltdown in Goldman’s fixed- income business and who, as a result, never fully had the trust and faith of the firm’s investment bankers. “That is one good question,” one Goldman trading partner said. “At a normal place, it would be discordant. You couldn’t imagine it. And I guess at this place, somehow you could.”

This, remember, is the world’s best investment bank. It’s worth bearing in mind when you see those eight-figure salaries and wonder whether they’re earned. And when you hear politicians bellyaching about the importance of keeping US banks “competitive” on the international stage. If this is competitive, it might well be best to just drop out of the competition all together.


Felix, by definition, those salaries are earned because the owners of the company have agreed to pay them.

Posted by TinyOne | Report as abusive

Youthful swearing, cont.

Felix Salmon
May 25, 2010 03:28 UTC

Remember the Global Business Oath of the Young Global Leaders at Davos? Let me remind you: it’s a terribly silly and earnest document which begins with “As a business leader I recognize that” and ends with “This pledge I make freely and upon my honor.” In between is a bunch of do-goody pablum. But thanks to Ben McGrath’s wonderful Talk piece in the latest New Yorker, I now know that the Global Business Oath has a rival: the MBA Oath. This one has a few small differences — it starts with “As a business leader I recognize my role in society” and ends with “This oath I make freely, and upon my honor.” But it also has a big difference: it’s a book, which can be bought at places like 800 CEO READ.com.

In hindsight, this was probably the biggest problem with the Global Business Oath, whose Twitter feed has only 264 followers, to the MBA Oath’s 568. For one thing, there are many more MBAs than there are Young Global Leaders, so the MBA Oath has a naturally larger constituency of smug young backstabbers to cultivate. And more importantly, the first thing that the inventors of the MBA Oath did was sell the book rights to the concept, with some unspecified “portion” of the proceeds going to “support the MBA Oath movement”. After all, in this world, if you’re not doing well by your idea, then you hardly count as even doing any good in the first place. So Max Anderson and Peter Escher first sold their book, and then parlayed their advance, before the book was even published, into well-paying jobs in asset management. Which I’m sure comports perfectly with the final principle of their oath:

I will invest in developing myself and others, helping the management profession continue to advance and create sustainable and inclusive prosperity.

So far, over 3,000 people have taken this oath, all of whom were presumably unfazed by the idea that they were pledging their personal honor to “help the management profession continue to advance”. If, that is, they actually read the thing.

John Paulson’s high-risk hubris

Felix Salmon
Jan 14, 2010 06:17 UTC

Malcom Gladwell is no particular expert on financial markets. But he has said, according to Moe Tkacik, that of all the people he has interviewed, he most identifies with Nassim Nicholas Taleb — in a 7,800-word profile which explains just how hard it is to invest in markets when your strategy involves losing money every day and waiting for a tail event.

With a few notable exceptions, like the few days when the market reopened after September 11th — Empirica has done nothing but lose money since last April. “We cannot blow up, we can only bleed to death,” Taleb says, and bleeding to death, absorbing the pain of steady losses, is precisely what human beings are hardwired to avoid. “Say you’ve got a guy who is long on Russian bonds,” Savery says. “He’s making money every day. One day, lightning strikes and he loses five times what he made. Still, on three hundred and sixty-four out of three hundred and sixty-five days he was very happily making money. It’s much harder to be the other guy, the guy losing money three hundred and sixty-four days out of three hundred and sixty-five, because you start questioning yourself. Am I ever going to make it back? Am I really right? What if it takes ten years? Will I even be sane ten years from now?” What the normal trader gets from his daily winnings is feedback, the pleasing illusion of progress. At Empirica, there is no feedback. “It’s like you’re playing the piano for ten years and you still can’t play chopsticks,” Spitznagel say, “and the only thing you have to keep you going is the belief that one day you’ll wake up and play like Rachmaninoff.”

So I was very puzzled to pick up this week’s New Yorker to find Gladwell write about John Paulson in very different terms. (The story is behind a firewall; it’s not particularly worth paying for, although the magazine as a whole is a fantastic value.)

Gladwell in this essay characterizes Paulson as “The most successful entrepreneur on Wall Street — certainly of the past decade and perhaps even of the postwar era”. I think this involves a very narrow criterion of what makes successful entrepreneur. Later in the essay Gladwell talks about how “people who work for themselves are far happier than the rest of us”, and in my experience hedge-fund managers — who do after all work primarily for their clients — are not in fact particularly happy people. They might not have a single boss telling them what to do, but the pressures of managing other people’s money are immense.

What’s more, any hedge fund manager playing a version of the negative-carry trade has it much worse than most of his peers. Warren Buffett says that the first rule of running other people’s money is don’t lose it; the second rule is “don’t forget the first rule”. One of the reasons Taleb gave for giving up running money day-to-day was precisely the incredible toll it takes when you’re losing money almost every day. Andrew Lahde, another huge winner from the subprime crisis, also quit the business, citing the way in which the stress of the job destroyed his health. Gladwell himself talks about how successful entrepreneurs will deliberately harm their own reputation if it means improving their risk-adjusted returns. That’s not a route to happiness.

And in any event, although success is often measured in dollars on Wall Street, even Wall Streeters don’t end the analysis there. Is Paulson really a more successful entrepreneur than, say, Charles Schwab, just because he arguably has more money? For that matter, Mike Bloomberg has not only founded a hugely successful company which can run very well without his presence; has also made more money than Paulson. Even if Paulson does count as an entrepreneur, it’s not at all obvious how he counts as being more successful than Bloomberg.

But I digress. The point is that Paulson, like Taleb, is a negative-carry kind of guy. Positive-carry investing can take you a very long way, and indeed it’s the foundation of the entire global banking industry, but it’s negative-carry trades which have the ability to score enormous home runs like Paulson’s. Many big hedge-fund managers avoid negative-carry trades, because they feel too much like a gambling habit: you pay out money every day in the hope of scoring a huge jackpot. That’s not the kind of strategy most investors in hedge funds particularly like, and indeed the likes of Taleb take great care to sell their funds as hedging devices — a place to put a small amount of your net worth for insurance purposes — much more than as absolute-return vehicles.

Paulson is no Taleb: his clients are pretty typical hedge-fund investors, including rich individuals who really hate losing money. Which means that his negative-carry trade — buying credit default swaps which obligated him to pay out millions of dollars in annual premiums, with no income attached at all — was extremely risky, from a business point of view. Gladwell quotes Greg Zuckerman explaining that “the most an investor could lose would be 8 percent a year”, while the upside (as we saw) turned out to be astronomical. But it doesn’t take many years of 8% losses — or any losses at all, for that matter — for clients to pull all their money out of your hedge fund.

Paulson was not actively trying to burst the bubble, in the way that George Soros pushed the pound out of the European exchange-rate mechanism with his legendary 1992 negative-carry trade. Instead, he was just the biggest of a long line of investors who saw that there was a housing bubble and tried to find a way to go short. Those who were right but too early disappeared into the footnotes of finance — if they were lucky to get even that. They learned the hard way that “the market can stay irrational longer than you can stay solvent”. Paulson was like them: he felt certain that the bubble was going to burst, but he didn’t — couldn’t — know when, and he simply had to pray that it would happen before his investors deserted him.

What’s more, there was no guarantee that even if the housing bubble did burst, that Paulson was going to make lots of money. To be sure, he had a lovely model, put together by his colleague Paolo Pellegrini, showing that if house prices stopped rising, subprime mortgages were going to suffer enormous losses. But on the other hand, all the banks and credit-rating agencies also had models, showing that the bonds that Paulson was betting against had almost no chance of defaulting. When your model shows one thing, and everybody else’s models show something else entirely, there’s a very good chance that your model is flawed.

Gladwell’s thesis, in this essay, is that Paulson is actually very risk-averse, rather than being a big risk-taker. “Would we so revere risk-taking,” he asks right before introducing Paulson as Exhibit A, “if we realized that the people who are supposedly taking bold risks in the cause of entrepreneurship are actually doing no such thing?”

The fact is that Paulson did take bold risks, on factors which were entirely out of his control: When would the bubble burst? How long could he hold out before his investors deserted him? Indeed, Paulson’s strategy had a Ponzi aspect to it, where he would try to make up losses with new investments: “He bought CDS contracts by the truckload,” Gladwell writes, “and, when he ran out of money, he found new investors, raising billions of new dollars so he could buy even more.”

Warren Buffett has described his most recent mega-acquisition, that of Burlington Northern, as a huge bet on the long-term health of the US economy. That kind of bet has made him more billions than even Paulson can dream of, and it’s a bet made in the positive-sum game of the equity markets. Stocks can and do rise over time, and a well diversified stock-market investor has been able to reasonably expect to see some kind of profit over the long term.

Paulson, by contrast, was playing in the zero-sum derivatives markets. In order for him to make any money at all, somebody else had to lose. If I bought a random basket of derivatives contracts and held them over the long term, my expectation would be that I would end up with less money than when I started.

The amount of sheer hubris involved in Paulson’s trade, then, is enormous. He had to have an unshakeable faith in the infallibility of his own models, in a world where no model is infallible. He had to have entirely irrational confidence that the bubble would burst before he ran out of cash. And he had to do all of this with other people’s money: while he was already personally set for life when he entered into the bet, he couldn’t say the same thing about all of his clients, who didn’t necessarily share his shoot-for-the-moon risk profile.

After all, Paulson’s clients had invested their money with a manager whose returns, Gladwell quotes Zuckerman as saying, were solid, careful, and decidedly unspectacular. Did Paulson decide to put them into a risky derivatives trade with a negative carry just because he’d already made lots of money and was now aiming for posterity? I’m sure all those clients are very happy with Paulson today. But if you’d told them about his strategy while the bubble was still inflating, they might have had a very different opinion indeed.

Update: Zuckerman responds, in the comments.


I thought I’d weigh in on Salmon’s interesting piece, given that it concerns John Paulson, and my recent book, The Greatest Trade Ever.

Were there risks to Paulson’s trade? For sure. Losses of 8% a year for a few years certainly add up. Then there was the reputation risk—if the trade hadn’t worked, Paulson would have been know as the guy who bet foolishly against mortgages after the experts warned him not to. Paulson likely wouldn’t have been able to try anything similar ever again. Further, when the trade finally started to pay off in early 2007 and Paulson piled up billions, he held on to most of his positions rather than cash out, transforming the trade, in my view, into a riskier one. He suddenly sat on huge profits that easily could have evaporated (as some of them did when the subprime market rallied in the summer of 2007).

But the dangers to Paulson’s trade weren’t outsized and it’s fair to say that he wasn’t acting an “extremely risky” fashion. So I disagree with the thrust of the Salmon piece.

Paulson was using credit-default swaps, which have a much more limited downside than a short position on equities or many other negative bets. The embrace of CDS was a sign that Paulson was indeed risk-averse. And it is unlikely that Paulson ever would have faced 8% annual losses for an extended period. If he was wrong on his trade and housing held up or kept rising, most of those who took out risky mortgages would have refinanced their loans (most of which had 2-year teaser rates), ending his CDS trade.

Just as important, Paulson was smart enough — and risk-averse enough — to place most of his subprime bets in a separate fund and lock his investors up for two years in that fund. That way, if he was wrong, it wouldn’t cripple his entire firm. That was good business sense, but also another sign of watching the downside

It’s a misunderstanding to say that Paulson “simply had to pray that it would happen before his investors deserted him.” As I note, they were locked up for two years, at least those in Paulson’s credit funds. And they were well aware of the potential downside, it was all spelled out and quite obvious, since they were buying CDS contracts with set payments. I’m not sure betting billions on the health of the rail industry, a la Buffett, is less risky — or suggests less “hubris” — than entering into CDS contracts with set costs to buy insurance on toxic mortgages.

And to say there’s a “Ponzi aspect” to what Paulson was doing is a bit silly. It’s sort of like saying Pimco is running a huge Ponzi scheme because it takes in money from investors each day, and – get this — uses it to buy investments that the firm likes. Even if new money came it, the returns to Paulson’s investors would have been based on their initial investment and when it was made.

Oh, and Andrew Lahde did complain about his back, and the stresses of his job as a hedge-fund manager. But that’s not why he quit the business. He simply enjoys spending time on the beach with beautiful women. Thanks for taking the time to read my book and for the interesting discussion! Greg Zuckerman

Posted by GZuckerman | Report as abusive

CEOs: Founders beat out managers

Felix Salmon
Nov 5, 2009 16:33 UTC

We’re less than two months from a New Year’s where a 9 ticks over into a 0, and so that means all manner of decade retrospectives. (And still we haven’t come up with a name for this decade!) Fortune is getting into the game early, naming Steve Jobs its CEO of the decade, for his work at Apple.

What’s more interesting to me is the list of 12 “also-rans” for the title: Larry Page, Sergey Brin, Warren Buffett, Bernie Madoff, Carlos Slim, Ken Lay, Jeff Skilling, Andy Fastow, Bill Gates, Oprah Winfrey, Alan Greenspan, and Martha Stewart. Five of the 12 aren’t CEOs at all (Page, Brin, Skilling, Fastow, Greenspan); and not a single one of the 12 is a CEO who was hired to run a company by its board of directors.

Jobs, by contrast, is such a CEO, in a manner of speaking: although he did found Apple, he sold all his shares when he was ousted in the 80s, and was hired back by Apple’s board. (As a result, he’s made more money from Pixar than he has from Apple.)

It’s natural for company founders to give themselves the CEO job. But how come all of Fortune’s top CEOs seem to be founders, and none of them are in the much more common position of having been hired, by the board, to run the company?


Probably because there’s a vaccuum of true leaders in this planet now, evident from the crisis we’ve just been. No wonder founders/entrepreneurs have that intrinsic leadership in them that lifts them to Fortune 500 group.

How the Sidekick fiasco is Microsoft’s fault

Felix Salmon
Oct 12, 2009 15:13 UTC

Is there an M&A lesson to be learned from the Microsoft/Danger/Sidekick fiasco? Here’s Dave Methvin:

Any $500 million acquisition usually comes with some technical due diligence. When Microsoft bought Danger, didn’t they have someone take a look at how the company ran their servers? During the more than 18 months since the acquisition, didn’t anyone review how Danger was operating?

The implication here is that the meltdown would have happened if Microsoft hadn’t bought Danger, and that Microsoft’s biggest mistake was not managing its acquisition more diligently. But Danger seemed to be perfectly good at cloud computing until Microsoft bought it — and then buried its founders so far down the org chart that one could easily forgive them for becoming somewhat demoralized.

It’s pretty obvious that company founders aren’t going to act with the same drive and sense of ownership when they’re a tiny part of a monster organization as they did when they owned and ran their own shop. Microsoft should have been on top of what was (and, more importantly, what wasn’t) going on at Danger, and been alert to defections and any hints of dissatisfaction in the team. Instead, it’s managed to deliver the single largest blow yet to the whole concept of cloud computing. And there’s lots of indication that Microsoft is at fault here:

Microsoft’s takeover of Danger almost two years ago should have given the software giant the time to fortify and secure Danger’s online operations. Instead, it appears the company actually removed support to cut costs…

Microsoft’s accountability in supporting its acquired SideKick support obligations with T-Mobile was also shirked… Microsoft could get more money from T-Mobile for their support contract if T-Mobile thought that there were still hundreds of engineers working on the Sidekick platform. As we saw from their recent embarrassment with Sidekick data outages, that has clearly not been the case for some time.

That indicates that Danger’s high profile cloud services failure didn’t occur in spite of Microsoft’s ownership, but rather because of it.

It’s not just that company founders lose zeal once they’ve been acquired; it’s also that executives at the acquiring company are often suspicious of what they’ve bought, especially when that company used to be a direct competitor. If you’re going to be spending billions of dollars on acquisitions, you should certainly invest a chunk of time and money ensuring smooth integration. That clearly didn’t happen here.

Update: Andrew Leonard has another idea:

Maybe we should consider this a Machiavellian shot across Google’s bow? What better way to defend the Windows/desktop franchise than to create a sense of fear, uncertainty and doubt concerning the fundamental security of cloud computing?


For a multi-billion dollar company like Microsoft not to have a redundant backup is simply unforgivable and inexcusable! Whether Danger failed to make a backup before its upgrade or Microsoft’s oversight http://www.microsoft.com/en-us/office365  /, this incident is a total disaster especially for its customers.

Posted by MichaelHost | Report as abusive