Pessimism of the Intellect, Optimism of the Will     Archive   Favorite posts   Twitter: @steve_hsu

Friday, March 31, 2006

Robot Genius invasion

A new version of our software is available here. Please try it out and help us find more bugs!

Did you ever wonder about the hundreds or thousands of files and registry changes that accompany any Windows software install? Our product tracks all of these changes and lets you uninstall any software (even malware or rootkits) in a few clicks. We operate at the driver layer between the OS and the physical hard drive, so all file creations and modifications must pass through our filter.

Tuesday, March 28, 2006

Non-residential net worth

Here it is by age group -- net worth excluding primary residence. Note the data is from 2001, so the top 1% or 5% will have gained while everybody else will be about the same (see Krugman discussion of growing wealth and income inequality here). The latest surveys suggest about 8 million US families (out of 110 million) have more than $1 million in liquid net worth. See here for more data from an IRS study.

Gordon Gekko, from the movie Wall Street: "The richest one percent of this country owns half our country's wealth, five trillion dollars. One third of that comes from hard work, two thirds comes from inheritance, interest on interest accumulating to widows and idiot sons and what I do, stock and real estate speculation. It's bullshit. You got ninety percent of the American public out there with little or no net worth. I create nothing. I own."

Monday, March 27, 2006

The end is near

Stephen Roach, Chief Economist at Morgan Stanley, speech at the China Development Forum in Beijing on March 19.

“There is a very simple and extremely powerful macro point that is being overlooked in this debate: America no longer has the internal wherewithal to fund the rapid growth of its economy. Suffering from the greatest domestic saving shortfall in modern history, the United States is increasingly dependent on surplus foreign saving to fill the void. The net national saving rate the combined saving of individuals, businesses, and the government sector after adjusting for depreciation fell into negative territory to the tune of -1.2 per cent of national income in late 2005. That means America doesn't save enough even to cover the replacement of its worn-out capital stock. This is a first for the United States in the modern post-World War II era, and I believe a first for any great power over a much longer sweep of world history.

...America, in general, and its consumers, in particular, treat rapid economic growth as an entitlement. That leaves the United States with little choice other than to pursue the second option - drawing heavily on the global saving pool in order to fund economic growth. Once the United States started down the slippery path of consuming beyond its internal means, it got harder and harder to break the habit. Ironically, it has become exceedingly difficult for Washington to accept the consequences of that habit - a nation that has become beholden both to external funding and production. And yet that's exactly how China fits into America's macro equation.

That underscores a key attribute of the savings-short, deficit nation: It is forced to run current account deficits in order to attract the requisite foreign capital. And in the case of the United States, where external funding needs are so massive - now closing in on US$800 billion per year, or about US$3 billion per business day - most of the current account imbalance shows up in the form of a huge trade deficit. In 2005, the trade deficit in goods and services accounted for fully 93 per cent of the total current-account gap.

High skill immigration

Amid the recent debate over illegal immigration, little attention has been paid to high skill immigration. The paper below discusses the effects of foreign immigration on the compensation and career prospects of PhDs in the US. I discussed this issue previously, for example here. The conclusion: what is bad for Joe or Jane PhD is good for Sergei, Sanjay and Song, and good for the US economy and rate of innovation. Benefits accrue to lawyers and financiers, and society in general, but not to US scientists :-(

(Via Economist's View.)

Immigration in High-Skill Labor Markets: The Impact of Foreign Students on the Earnings of Doctorates

Introduction The rapid growth in the number of foreign students enrolled in American universities has transformed the higher education system, particularly at the graduate level. In 1976, 72.4 thousand foreign students were enrolled in graduate programs, making up 5.5 percent of total enrollment. By 2000, 232.3 thousand foreign students were enrolled, or 12.6 percent of enrollment. The impact is even greater at the doctoral level. For example, the fraction of doctoral degrees awarded to foreign students rose from 11.3 to 24.4 percent during the same period, with nonresident aliens receiving a remarkably high share of the doctoral degrees awarded in the physical sciences (36.5 percent of all doctorates awarded in 2000), engineering (50.7 percent), and the life sciences (25.7 percent).

Many of these newly minted doctorates remain in the United States after receiving their doctoral degrees... Despite the large size of the supply shock and despite the importance of the labor market for doctorates in determining technological change and economic growth, there has not been any study of how the foreign student program affects labor market conditions for high-skill workers. This paper provides an initial attempt to address a question...: Has the foreign student influx into doctoral programs harmed the economic opportunities of competing native workers?

...This paper uses data drawn from the Survey of Earned Doctorates and the Survey of Doctoral Recipients to analyze the impact of the influx of foreign students on the earnings of doctorates. ... The empirical analysis ... clearly shows that a foreign student influx into a particular field at a particular time has a significant and adverse effect on the earnings of competing doctorates in that field who graduated at roughly the same time. A 10 percent immigration-induced increase in the supply of doctorates lowers the wage of competing workers by about 3 to 4 percent—remarkably similar to the elasticity estimates reported in Borjas (2003)... Because the magnitude of the immigrant supply shock in particular fields has been sizable, this elasticity implies that many doctorates employed in the United States, whether native-born or foreign-born, have experienced a substantial wage loss.

These results have implications in a number of different policy contexts. For instance, there has been a long-standing debate about whether immigration affects labor market conditions for native workers at all. This study ... seems to suggest that the supply-demand textbook model is correct after all: increases in labor supply do move the labor market along the demand curve and lead to lower wages for competing workers.

It is also the case that economic opportunities in high-skill labor markets are among the key determinants of the career decisions made by the native-born student population. The increase in the number of foreign doctorates has clearly reduced economic opportunities in some fields relative to others, and may be an important factor driving native students to enter particular occupations and avoid others. For example, the wage that could be earned by native postdoctoral workers employed in research biology labs is much lower than it would have been in the absence of the immigrant influx, perhaps motivating bright U.S.-born undergraduates to pursue professional occupations that have not been targeted by immigration. The low wage paid to postdoctoral workers in these biology labs, however, still offers a very attractive opportunity when contrasted to the compensation available in other countries, so that the incentives for even more foreign students to enter the United States are not greatly reduced. ... In the resulting equilibrium, research labs find that they must keep recruiting from abroad because “natives do not want to do the type of work that immigrants do...

Finally, although the foreign student program grew rapidly in the past three decades, this growth occurred without any systematic study of the costs and benefits that such a program entails for the native-born population. This paper addressed an important component in such a cost-benefit analysis—the cost borne by doctorates in the U.S. labor market. There is an equally important component that has not yet been analyzed carefully, namely the benefits of the program, such as the possibility that the sizable increase in the skill endowment of the workforce accelerates the rate of scientific discovery. These benefits could be very large and accrue to particular parts of the population, so that high-skill immigration may have significant efficiency and distributional effects that have yet to be analyzed.

Saturday, March 25, 2006

Huawei, Lucent and Alcatel

Behind the Lucent-Alcatel merger: competition from Chinese companies like Huawei and ZTE.

WSJ: One key factor fueling the merger talks between France's leading telecommunications-equipment maker and its longtime U.S. rival is a looming competitive threat: China.

In a plan made public late Thursday night, Alcatel SA of Paris is pursuing a more than $13 billion merger with Lucent Technologies Inc., whose roots stretch back to Alexander Graham Bell's first telephone. The two came close to merging five years ago but talks fell apart.

Now, there is renewed urgency for a deal. Chinese upstarts such as Huawei Technologies Co. and ZTE Corp. are beginning to enter their turf with cheaper products, threatening to seize market share, take jobs and set industry standards for the huge Chinese market.

Alcatel, Lucent and its main competitors Telefon AB L.M. Ericsson, Nokia Corp. and Nortel Networks Corp., make technology that is largely invisible to consumers but underlies many of the services they use: telecommunications networks transmitting voice, video and Internet traffic.

...Last year, for instance, BT Group PLC of the United Kingdom signaled Huawei had arrived on the global stage by awarding it a sizeable piece of its multibillion-dollar project to upgrade its infrastructure. Huawei, based in the southern Chinese city of Shenzhen, started by concentrating its export strategy on developing countries, partly to avoid competing with the industry's giants as it built up its expertise, touting deals in Thailand, Indonesia and Bulgaria.

The pressures Lucent and Alcatel are facing in such markets was visible at an industry trade show in New Delhi this week. Just inside the main entrance, the first sight for many attendees was the giant booth of China's ZTE, where it demonstrated equipment for next-generation cellphone networks, wireless broadband and video cellphones.

For ZTE and Huawei, India is a second big opportunity to prove themselves against global competitors. Both companies played starring roles in the rapid buildup of China's wireless networks during the past six years.

Meanwhile, China is expected to start awarding licenses for so-called third-generation, or 3G, networks in the next six months in order to give its own state-owned phone companies time to prepare services for the 2008 Olympics in Beijing. China already is the largest mobile-phone market, and the upgrade may form the world's biggest high-speed wireless network.

To be sure, Chinese vendors including Huawei currently make up a fraction of the world market. They hold about 5% of the wireless market, including new contracts within China, and about 10% of the market for optical equipment such as switches, according to analyst Paul Sagawa of Sanford Bernstein.

"The biggest impediment to Huawei on a global stage is the lack of a service and support infrastructure," Mr. Sagawa says. "Carriers don't want to have to rely on engineers in Beijing to solve problems. They want people on the ground."

...Some analysts wonder why Mr. Tchuruk is plunging the company into a risky deal when Alcatel just now seems to have gotten its own house in order. "It would take a herculean feat to put two companies together," said Richard Windsor, an analyst at Nomura Securities. "You likely end up with a value-destroying merger."

Large buyers of telecom equipment signaled their support over the prospect of a deal. William Smith, chief technology officer of BellSouth, says that "there is an onslaught of new Chinese technology" that poses a threat to companies such as Lucent and Alcatel. By pooling research resources, the telecom industry could see the development of more new technologies, he adds. "You haven't seen any Nobel Prizes out of those organizations recently," he said.

Thursday, March 23, 2006

A night with hollywood

Some thoughts on last night's panel discussion.

We had a pre-dinner with the speakers and moderator at the Palomino in Westwood. Calacanis is a funny guy, as expected, but so is Zittrain (a law professor, go figure!). WIRED editor Jeff O'Brien did a fair job of not picking on Yahoo or AOL (Calacanis is at AOL since they bought his blogging company) as collaborators with bad governments.

The panel was held in Lawrence Bender's beautiful Bel Air home. The crowd was about 70 people, mostly young film and media people of all types, including actors, producers, directors, agents. I joked to someone that I thought you should be able to tell the actors and actresses from across the room, since they had to be exceptionally good looking, but then even the executives here seem to be good looking. I may have met more documentary film makers than anything else (Bender is currently working on a documentary film on global warming with Al Gore). I guess people interested in social causes tend to be interested in that type of project.

Although the topic was rather geeky the audience was very engaged. The panel went for 90 minutes, with lots of questions, but it seemed to me like it was over in an instant. People got most interested when the topic turned to privacy issues here in the US. Both Calacanis and Scott Moore (Yahoo VP of content, formerly in charge of MSN content like Slate) thought there would be a big market for privacy services going forward. In principle I agree, but the devil is in the details. Since the average person's understanding of their digital privacy is so amorphous it is hard to know what exactly they are willing to pay for.

They recorded the whole panel using pretty fancy equipment, including what seemed to be professional lighting and a nearly high-definition camera. I think a compressed version will be available from the Hollywood Hill site in a short while.

Saturday, March 18, 2006

No startup culture in Europe

The lack of a culture supportive of risk taking is one of the worst handicaps Europe faces in economic competition with the US and developing countries like India and China. Americans are descended from a population of immigrant risk takers, and the US may produce more hypomanic entrepreneurs per capita than other countries. We also continue to attract many from abroad, thanks to our forgiving attitudes toward failure and repeated attempts at success.

For those who know portfolio theory, risk adjusted return is optimized by having a certain balance of assets, some more risky, some less. Perhaps counterintuitively, return can sometimes be enhanced at no cost to overall risk by increasing the volatile component of the portfolio -- if the behavior of the additional assets is uncorrelated with the others. A country without entrepreneurs can't be near its "efficient frontier" in economic or technological development. It is not placing enough risky bets.
NYTimes: Skype survived its early tests as a European start-up to become a world leader in Internet telephony but the region's aversion to risk means many other fledging companies are doomed, Skype's founder said.

Ahead of meeting next week of European Union leaders to discuss long-delayed reforms to make Europe more competitive, Niklas Zennstroem told business and EU officials that Europe still has the wrong culture when it comes to entrepreneurship.

``In the U.S. for example, if you have a start-up and it doesn't work out, you have gained an experience,'' he told the conference. ``In Europe, you have made a mistake.''

Skype became one of the hottest takeover stories of last year when it was snapped up by the Internet auctioneer eBay for $4 billion, less than three years since the launch of its software enabling free phone calls over the Internet.

Zennstroem, a 40-year-old Swede, and co-founder Janus Friis of Denmark are a rare success story in a continent that lags behind the United States in research spending and offers little help to young innovators.

EU policymakers have long highlighted the need to accelerate research and innovation and support small and medium-sized enterprises (SMEs) to make the bloc more competitive against the United States and Asia.

Yet industry leaders and politicians admit the bloc has fallen short of its own 10-year Lisbon Agenda programme for a knowledge-based economy driven by research and development.

Technology start-up companies still face obstacles, chief among them scarce funding, Zennstroem said.

``We went round Europe trying to raise money for one year to close our first round. If we were a Silicon Valley company, it probably would have taken us one month,'' he said, referring to the area in California that is home to many high-tech start-ups.

Then, as the company began to be a success -- by last count Skype had 68 million users worldwide -- and was looking for a buyer, ``not a single European company was interested in even talking to us,'' Zennstroem said.

But it is not just the venture capitalists who are wary of risk in Europe. With unemployment high, young people are demanding job security in some countries.

Thursday, March 16, 2006

Hollywood Hill panel

When WIRED magazine calls, I can't resist. I'll be participating in a panel discussion on Internet censorship with some hollywood social activists next week. I was involved with Internet security and privacy issues during my SafeWeb days, even architecting an encrypted P2P packet rerouting system for defeating government censorship. At its peak our system was used over 5 million times per day, including by people in Iran and China! (Please don't ask me, or remind our investors, how much all that bandwidth cost.)

The event is in the Bel Air home of well-known producer Lawrence Bender (Pulp Fiction, Good Will Hunting, Kill Bill...). I hope I'll get the chance to pitch my screenplay idea :-)


Speakers: Jonathan Zittrain, Scott Moore, Jason Calacanis, Stephen Hsu, Jeff O'Brien

Sunday, March 12, 2006

China research and development

From WSJ. In many research areas salary is the largest cost, so if China can get to 2% of GDP in R&D expenditures (by 2010), that will make them a powerhouse. By comparison the US is spending 2.7%, but that number is unlikely to increase. Note the other metric mentioned: 1M science and engineering grads per year, which is, I believe, several times the US number. (Some PhD data here.)

Multinational companies, drawn by a huge and inexpensive talent pool, are pouring money into research and development in China -- a trend that promises to broaden the country's huge role in the global economy.

The total number of foreign-invested R&D centers in the country has surged to about 750 from 200 four years ago, according to China's Ministry of Commerce. And in a survey of multinationals published in September by the United Nations Conference on Trade and Development, China was by far the most frequently cited location for R&D expansion, well ahead of the U.S. and third-place India, China's chief rival as an emerging innovator.

Still, China's growth as a global R&D hub faces some constraints. Among them is the country's weak protection of patents and other intellectual-property rights. That has encouraged some foreign companies, fearful of risking their trade secrets, to keep more cutting-edge research out of China, analysts say. But others have rushed to expand the scope of their development efforts here.

Whereas R&D investment in China initially focused on adapting existing products and technologies to the Chinese market, companies such as Procter & Gamble Co., Motorola Inc. and International Business Machines Corp., among many others, have been investing to expand their Chinese R&D operations to develop products for the global market.

...Giving impetus to the R&D expansion in sectors from biotechnology to pharmaceuticals to semiconductors is China's government. Having enlisted foreign investment to transform China into a manufacturing powerhouse over the past few decades, Beijing now is mounting a campaign to strengthen domestic innovation that could help push the country into more advanced niches of the global economy.

In his annual report at the National People's Congress in Beijing, which ends tomorrow, Chinese Premier Wen Jiabao said the central government will increase spending on science and technology by nearly 20% this year. "China has entered a stage in its history where it must increase its reliance on scientific and technological advances and innovation to drive social and economic development," he said.

China's State Council, or cabinet, recently said the country would seek to boost R&D investment to 2% of gross domestic product in 2010 and 2.5% by 2020. At a news conference Friday, senior officials outlined tax breaks and other tools they plan to use to meet that target. Last year, total R&D spending in China -- not including foreign investment -- reached $29.4 billion, rising steadily from $11.13 billion in 2000, according to the government.

China faces numerous obstacles to joining the ranks of the world's innovation leaders -- beyond its weak intellectual-property protections. Research spending is still small compared with that of developed countries; the U.S., for example, spends about 2.7% of GDP on R&D, compared with 1.3% of GDP in China last year. And much of what is spent in China still comes from foreign companies: Less than a quarter of Chinese midsize and large enterprises had their own science and technology institutions in 2004. Of China's high-tech exports, valued at $218.3 billion last year, nearly 90% was produced by foreign-invested companies, according to the Ministry of Commerce.

Still, the R&D trend is bolstering China's position relative to other developing countries, particularly India, which is also seeking to build its innovation abilities. India's total domestic spending on R&D rose an estimated 9.7% to $4.9 billion, or 0.77% of GDP, in the fiscal year ended March 2005, according to India's Ministry of Science and Technology.

India is also trying to build R&D, "but the scale of investment [compared with China] is not much" because of budgetary constraints, says V.S. Ramamurthy, a top official at the ministry. Foreign investment in Indian R&D has also lagged behind that of China, he says. And while Mr. Ramamurthy argues that the amount of investment isn't the only way to measure R&D success, "it is a concern for us."

Zhang Jun, director of the China Center for Economic Studies at Shanghai's Fudan University, says that given time, "China's advantages in this area will become more obvious...and its attractiveness will increasingly become stronger than India's."

Among China's draws, he says: the relatively low cost of hiring engineers and researchers; a huge talent pool, including five million university graduates annually (one-fifth majoring in science or engineering); and China's own huge market of 1.3 billion consumers. China offers its students abroad incentives to return once they graduate, including generous research grants and chances to run their own R&D projects.

One early returnee is Enge Wang. Mr. Wang, who had worked as a research associate at the University of Houston, decided to return to Beijing to conduct research under a Chinese Academy of Sciences program in 1995. At the time, he says, his U.S. colleagues and friends questioned his decision, but he says he is glad he made the move. Today, Mr. Wang is director of the Institute of Physics under the academy, one of China's top research organizations, which is engaged in several R&D cooperative ventures with foreign companies.

China's "research funding is getting much better," Mr. Wang says, and as a result, overseas Chinese are flocking back from top U.S. institutions like Harvard University and Lawrence Berkeley National Laboratory. Talented returnees can secure enough backing "to build up their own lab and extend their research in one direction for 10 years," he says. "It's hard to find such conditions elsewhere."

"There's been a paradigm shift among foreign companies in China," says Chen Zhu, a Chinese Academy of Sciences vice president. "Now, more foreign companies realize China is not just a market but a country with huge amounts of talent."

Motorola, which began investing in low-level R&D in China in 1993, now has 16 R&D offices in five Chinese cities, with an accumulated investment of about $500 million. The U.S. company has more than 1,800 Chinese engineers, and the number is expected to surpass 2,000 this year. They have recently begun developing new phones and other products for sale not only in China, but also overseas, executives say.

One phone developed in China, the A780, lets users write on the screen with just a finger, rather than a stylus. It's now available in the U.S. and Europe. Another phone that can scan contact information from business cards using a built-in camera and enter it into a contact database is expected to be marketed in the U.S. "China is moving from the manufacturing center into advanced R&D," says Ching Chuang, who heads Motorola's Chinese R&D operations.

Microsoft Corp.'s basic-research lab in Beijing was only its second outside the U.S. when it opened in 1998. That China lab now employs about 200 full-time scientists, and the software giant expects its total R&D headcount in China to double in this year to about 800 researchers.

At IBM's research lab in Beijing, Chinese scientists have led the development of several technologies now being used abroad. Among them: "voice morphing" software that can convert typescript or a recorded voice into another voice. "Our R&D now has a global mission," says Thomas S. Li, director of IBM China Research Lab.

Still evolving commotion

More from Nicholas Wade of the Times on the evidence that numerous human genes have been under selection pressure during historical (5-10k year) timescales. (See earlier post here.) The results imply that human populations as recently as a few thousand years ago might have had very different allele frequencies than current populations. (For example, a mutation present at the 1% level then might have become commonplace today.) The implications for history and society are fascinating. Note that different sets of genes have been under selection in different regions of the world -- leading to uncomfortable questions concerning group differences.

For the flabbergasted, ill-reasoned, politically-correct reaction, see Brad DeLong.

NYTimes: ...Humans have continued to evolve throughout prehistory and perhaps to the present day, according to a new analysis of the genome reported last week by Jonathan Pritchard, a population geneticist at the University of Chicago. So human nature may have evolved as well.

If so, scientists and historians say, a fresh look at history may be in order. Evolutionary changes in the genome could help explain cultural traits that last over many generations as societies adapted to different local pressures.

...In a study of East Asians, Europeans and Africans, Dr. Pritchard and his colleagues found 700 regions of the genome where genes appear to have been reshaped by natural selection in recent times. In East Asians, the average date of these selection events is 6,600 years ago.

Many of the reshaped genes are involved in taste, smell or digestion, suggesting that East Asians experienced some wrenching change in diet. Since the genetic changes occurred around the time that rice farming took hold, they may mark people's adaptation to a historical event, the beginning of the Neolithic revolution as societies switched from wild to cultivated foods.

Some of the genes are active in the brain and, although their role is not known, may have affected behavior. So perhaps the brain gene changes seen by Dr. Pritchard in East Asians have some connection with the psychological traits described by Dr. Nisbett.

Some geneticists believe the variations they are seeing in the human genome are so recent that they may help explain historical processes. "Since it looks like there has been significant evolutionary change over historical time, we're going to have to rewrite every history book ever written," said Gregory Cochran, a population geneticist at the University of Utah. "The distribution of genes influencing relevant psychological traits must have been different in Rome than it is today," he added. "The past is not just another country but an entirely different kind of people."

John McNeill, a historian at Georgetown University, said that "it should be no surprise to anyone that human nature is not a constant" and that selective pressures have probably been stronger in the last 10,000 years than at any other epoch in human evolution. Genetic information could therefore have a lot to contribute, although only a minority of historians might make use of it, he said.

The political scientist Francis Fukuyama has distinguished between high-trust and low-trust societies, arguing that trust is a basis for prosperity. Since his 1995 book on the subject, researchers have found that oxytocin, a chemical active in the brain, increases the level of trust, at least in psychological experiments. Oxytocin levels are known to be under genetic control in other mammals like voles.

Wednesday, March 08, 2006

Purity of thought in the midwest

I'm on my way home from St. Louis, after giving a colloquium at Washington University (where, I learned, Arthur Compton discovered Compton scattering :-) I am happy to find that the huge airport in Las Vegas (can't get to St. Louis directly from Eugene) has free WiFi.

In informal discussions with theory graduate students today I learned that academic purity is still unspoiled in some quarters. The issue of future career plans (indeed, most probable career paths) came up, and I was surprised that none of them knew much about finance, where most of their brethren have ended up in recent years! In response to a question about what exactly physicists might do in finance, I gave an example of an old collaborator of mine who worked on heavy quark effective theory and is now a well-known modeler of prepayment risk in mortgage backed securities.

Coincidentally, I found this article from today's WSJ quite apropos.

Why Students Of Prof. El Karoui Are In Demand. French Math Teacher Covers Structure Of Derivatives; Banks Clamor for 'Quants' A Lesson on 'Smile Risk'

March 9, 2006; Page A1

When Xavier Charvet applies for a job at an investment bank next year, he thinks he'll have an advantage. The 24-year-old French student's resume begins with the phrase: "DEA d'El Karoui."

That stands for the postgraduate degree he is studying for under Nicole El Karoui, a math professor in Paris. She teaches skills required to create and price derivatives, the complex financial instruments based on stocks, bonds or loans. "When I talk about El Karoui's master's, everyone knows" about the degree, says Mr. Charvet.

As derivatives have become one of the hottest areas for the world's biggest banks, Ms. El Karoui, 61 years old, has become an unlikely player in the business. Her courses at the prestigious Ecole Polytechnique and a state university, in such rarefied subjects as stochastic calculus, have become an incubator for experts in the field. A rsum with her name on it "is a shortcut because you don't need to train the person on the basics of derivatives," says Rachid Bouzouba, a former student who is now head of European equity trading at the London office of Lehman Brothers Holdings Inc.

The derivatives departments at banking giants J.P. Morgan Chase & Co., Deutsche Bank AG, Dresdner Kleinwort Wasserstein, and France's BNP Paribas SA and Societe Generale SA include many of her protégés.

The high demand for her students reflects big changes in the global banking industry. Investment banks used to make much of their money from underwriting and trading stocks and bonds, or providing mergers-and-acquisitions advice. They hired people with a wide range of academic experience, including liberal-arts and science graduates.

In recent years, profits from trading and selling derivatives have come to rival those from stocks and bonds at many banks. On average, revenue from derivatives based on stocks now accounts for about 30% of an investment bank's total revenue from stock-related businesses, according to a Citigroup Inc. report issued in January.

As a result, banks are hiring an increasing number of recruits who understand derivatives. Inside banks, they are known as "quantitative analysts," or "quants" for short. They are able to marry stochastic calculus -- the study of the impact of random variation over time -- with the realities of financial trading.

Derivatives are financial contracts, often exotic, whose values are derived from the performance of an underlying asset to which they are linked. Companies use them to help mitigate risk. For example, a company that stands to lose money on fixed-rate loans if rates rise can mitigate that risk by buying derivatives that increase in value as rates rise. Increasingly, investors are also using derivatives to make big bets on, say, the direction that interest rates will move. That carries the possibility of large returns, but also the possibility of large losses.

The 75 or so students who take Ms. El Karoui's "Probability and Finance" course each year are avidly sought by recruiters. Three years ago, Joanna Cohen, a specialist in quant recruitment at Huxley Associates in London traveled to Paris to meet Ms. El Karoui to ensure her search firm was in the loop when students hit the job market. Today, Ms. Cohen says she carefully checks résumés with Ms. El Karoui's name to make sure applicants aren't overstating their interaction with the professor.

"French quant candidates know that Nicole El Karoui's name has real clout, so many of them put her name on their [curriculum vitae] even if they've just taken one course with her. They want to give the impression that she has supervised their Ph.D.," Ms. Cohen says. "It'd be impossible for any one person to supervise the number of students who put her name on their CV."

Rama Cont, a former student and now a research fellow at the Ecole Polytechnique, describes a degree with Ms. El Karoui's name on it as "the magic word that opened doors for young people."

Headhunters say Ms. El Karoui's graduates can expect to earn up to about $140,000 a year in their first job, including a bonus, once they complete an internship that constitutes part of her course. After five years, they could be earning at least three times as much.


In BNP Paribas's offices in London, the fixed-income interest rates derivatives research team, which totals six, includes three of her former students. On a recent day, Fahd Belfatmi, who took Ms. El Karoui's course in 2003, was working at the bank on a model to predict long-term interest rates. For help, he keeps handy a beat-up, paperback copy of Ms. El Karoui's French-language textbook, "Stochastic Models in Finance."

Ms. El Karoui's only hands-on banking experience in her 38-year career was a six-month stint about two decades ago at a French retail bank. "I'm still a theoretician. My knowledge of markets is patchy and I've never spent a year in a trading room," she says. "On many counts, I probably have a fairly naive vision of things."

Carving Out a Niche

But she was one of the first in the world to carve out an academic niche studying the underpinnings of derivatives transactions, starting courses in the late 1980s. About two dozen universities have moved into that field, setting up their own mathematical-finance departments, including Stanford University, Carnegie Mellon University and the Massachusetts Institute of Technology.

One of eight children in a middle-class family, Ms. El Karoui grew up a Protestant in a predominantly Catholic town in eastern France. Today she attributes her nonconformity to that background. "Protestants are rebels by nature," she says. Though her mother thought France's elite colleges were better suited for boys, her father, an engineer, encouraged her to take the tough entrance exams for Ecole Nationale Superieure, where she was accepted to study math. In 1968, around the time she was protesting the Vietnam War, she married a Muslim Tunisian economics professor, Faycal El Karoui.

"If you'd told the left-winger that I was then that I was going to end up working in finance, I'd never have believed it," Ms. El Karoui says.

France, the land of Descartes and Fermat, has a storied tradition in the study of math. Over the years, its engineering schools, including Ecole Polytechnique, a 212-year-old institution transformed by Napoleon into a military academy, have produced a steady stream of math students. Louis Bachelier's work in 1900 at the Sorbonne is considered the earliest effort to grasp how the markets work.

Ms. El Karoui first branched into finance in 1987. The government had just closed down the elite Ecole Normale Superieure in Paris, where she had been teaching. She took a six-month sabbatical to work in the research department of consumer credit bank Compagnie Bancaire.

At the time, many French mathematicians tended to deem the world of finance beneath them. "Finance meant selling your soul to the devil," she says. Her break with the French math establishment "took a lot of courage," says Marek Musiela, a leading figure in financial mathematics and the global head of fixed-income quant research at BNP Paribas.

At first, Ms. El Karoui felt out of her depth. "I didn't even know what a bond is. I took a dictionary to look up the financial words," she recalls.

But she soon realized that employees on the bank's newly formed derivatives desk were facing problems similar to those of stochastics scholars in trying to build models to predict the impact of interest-rate changes.

After her time at the bank, she took a post teaching at the Paris VI, officially known as the University of Pierre and Marie Curie. She and another academic, Hlyette Geman, launched a postgraduate mathematical-finance course. Demand for know-how in derivatives was growing rapidly among banks at that time, sparked by the development of specialized exchanges that could trade derivative products, such as futures.

"I said 'That's beautiful mathematics and it's teachable as a theoretical course,'" Ms. El Karoui says.

Amine Belhadj, head of BNP Paribas's U.S. equity and derivatives department in New York, says Ms. El Karoui played a crucial role in finding interns when the bank began handling derivatives for clients in 1989. "There was nobody on the options desk with a mathematical-financial background," he says. "Having someone like Nicole who was making a specialty of it was pretty timely."

Today, four of her five children have pursued careers in math and sciences, two as academics and two still as students. In her spare time, Ms. El Karoui plays classical piano, with a preference for Brahms sonatas.

She earns about 80,000, or about $95,000, a year as a professor, plus a smaller amount for consulting fees -- a fraction of what her students can make. She drives around Paris in a small Renault.

A Warning

Lately, Ms. El Karoui has been vocal in warning students to use derivatives carefully. She says she is perturbed that an instrument that began primarily as a hedge for banks and financial firms against market risk is increasingly being used as a way to make a profit. Investors can profit, for example, by betting that the prices of stocks or bonds will increase. Ms. El Karoui worries that those looking for quick speculative gains could ramp up their bets on derivatives, but lose sight of the underlying financial instruments on which they're based, actually increasing their risk exposure.

"Some clients aren't mature enough to understand the risks of products that are too complex," she says. "It's better to do business with those people responsibly, either taking the time to teach them or selling them a less complex product."

Some big banks are being criticized for selling derivatives to institutions that may not understand the risks. Last year, for instance, Bank of America Corp. and Barclays PLC of the United Kingdom each agreed to settle claims that they had missold or mismanaged derivatives that were purchased by smaller banks in Italy and Germany. The banks said the matters were settled amicably.

One recent afternoon in her classroom, Ms. El Karoui ran through a series of dense formulas designed to price derivatives. In class were about 50 students studying for the DEA, or "Diplome d'Etudes Approfondies," as a French master's degree leading to a doctorate is known.

Ms. El Karoui talked softly toward the blackboard as much as she faced her students. There were few questions. Only near the end of the two-hour class did she raise a faint titter as she gestured to a full page of equations headed "General Pricing Formula." "There might be some of you brave enough to go through this," she said, then continued on, breezing through arcane jargon such as "smile risk," "volatility of volatility" and "Vega hedging."

To some, Ms. El Karoui has been almost too successful in placing her students in top international banks. Ryan Taylor, a headhunter specializing in quantitative-finance candidates at Napier Scott Executive Search Ltd. in London, says some investment bankers are now starting to question how many French-trained quants are in the field. "France has got what borders on a monopoly of quant candidate production and we'd love to hear from quants in other countries," he says.

Tuesday, March 07, 2006

Still evolving

Surprise! Selection pressures on different ethnic groups have been far from uniform during the period since humans left Africa.

I was shocked to learn over the last few years that most social scientists who teach about race don't know what HapMap is. Parents of college students may want to request some of their tuition back.

Related posts here and here.

NYTimes: Providing the strongest evidence yet that humans are still evolving, researchers have detected some 700 regions of the human genome where genes appear to have been reshaped by natural selection, a principal force of evolution, within the last 5,000 to 15,000 years.

The genes that show this evolutionary change include some responsible for the senses of taste and smell, digestion, bone structure, skin color and brain function.

Many of these instances of selection may reflect the pressures that came to bear as people abandoned their hunting and gathering way of life for settlement and agriculture, a transition well under way in Europe and East Asia some 5,000 years ago.

...The finding adds substantially to the evidence that human evolution did not grind to a halt in the distant past, as is tacitly assumed by many social scientists.

...The four populations analyzed in the HapMap project are the Yoruba of Nigeria, Han Chinese from Beijing, Japanese from Tokyo and a French collection of Utah families of European descent. The populations are assumed to be typical of sub-Saharan Africa, East Asia and Europe, but the representation, though presumably good enough for medical studies, may not be exact.

Dr. Pritchard's test for selection rests on the fact that an advantageous mutation is inherited along with its gene and a large block of DNA in which the gene sits. If the improved gene spreads quickly, the DNA region that includes it will become less diverse across a population because so many people now carry the same sequence of DNA units at that location.

Dr. Pritchard's test measures the difference in DNA diversity between those who carry a new gene and those who do not, and a significantly lesser diversity is taken as a sign of selection. The difference disappears when the improved gene has swept through the entire population, as eventually happens, so the test picks up only new gene variants on their way to becoming universal.

The selected genes turned out to be quite different from one racial group to another. Dr. Pritchard's test identified 206 regions of the genome that are under selection in the Yorubans, 185 regions in East Asians and 188 in Europeans. The few overlaps between races concern genes that could have been spread by migration or else be instances of independent evolution, Dr. Pritchard said.

Thursday, March 02, 2006

Success vs ability

The figure above illustrates the correlation between two variables, let us say success and ability. Each point represents an individual whose success and ability are shown on the vertical and horizontal axes, respectively. In the figure, the correlation is high, but not 100%.

For example, in American football the ability axis might represent the quantities obsessively tracked by NFL scouts: sprinting speed (40 yard dash time), natural strength (bench press), etc., while the vertical axis represents actual output, like passes caught or rushing yards gained. In real life, output is never purely determined by a single, or even several, input ability or abilities. If nothing else, luck ensures that the correlation is imperfect. Sports fans know that the fastest wide receiver isn't necessarily the best, nor the tallest basketball center the most productive, even if being fast or tall confer specific advantages. In the figure, the most able individual is not the most successful. They are seldom the same individual unless the correlation is 100%

In science or academia, we might take the horizontal axis to represent raw intellectual ability. The graph tells us to expect that the smartest person is not necessarily the most successful. It also suggests a population of successful but insecure people (the upper right dots above the fit line -- they are dumber than peers of similar accomplishment) and a population of smart people who are bitter about their unrecognized genius (dots on far right below the fit line -- they are smarter than peers of similar accomplishment).

Blog Archive


Web Statistics