Friday, December 28, 2007

Bitter defeat

Some more reading recommendations, this time on the immediate aftermath of WWII: the Russian occupation of Berlin and the US occupation of Tokyo. It's interesting that (at least in English translation) both the Germans and Japanese referred acerbically to their conquerors as the Victors.

Why this subject? I'm interested in how people behave when the framework of civilization has all but collapsed (i.e., the Hobbesian natural state, where life is nasty, brutish and short), and also in how order and the social contract are subsequently reimposed.

A Woman in Berlin: this memoir is interesting because of the period it covers -- 8 weeks beginning in April 1945, including the siege of Berlin, the early Russian occupation, and emerging postwar normalcy -- but also because of the unmatched clarity and insight of the 34 year old German author, a Russian-speaking journalist with a cosmopolitan background. Most readers will be shocked at the behavior of the Russian occupiers (in particular the mass rapes), but if the recollections of the author are correct the Russians seemed to have behaved much better than their German counterparts in the earlier eastern campaigns. I was quite impressed by the literary quality of the writing and feel that this memoir should someday make an excellent film.

Related recommendations: the early short fiction of Nobel laureate Heinrich Boll. Until I read this memoir I felt Boll's stories ("Trümmerliteratur"—the literature of the rubble) best captured the feel of occupied Germany. See also After the Reich: the Brutal History of the Allied Occupation for a historian's summary with none of the literary merits of Boll or the memoir (which seems to be used as a primary source by a number of historians).

Tokyo Year Zero: this is a novel, and a bit over the top, but captures well the deprivation and despair of Tokyo in 1946 (Guardian review). The title evokes Rossellini's film Germany Year Zero, which I also recommend. See also John Dower's Embracing Defeat, which seems to have provided some of the historical background.

Thursday, December 27, 2007

Anatomy of a CDO

Click the image for the original flash graphic, which illustrates quite clearly the re-bundling of securities. The process may sound technical, but in fact it appears the main variable is just an estimate of default probability and recovery fraction for various categories of mortgages. It will be interesting to see, when the dust settles, how much of the (massive) error in prediction came from outright fraud on the mortgage application (essentially, miscategorization) versus error in estimates of default probability by category.

WSJ: ... Mr. Ribotsky says the selling required little effort, as Merrill drummed up interest from its network of contacts. "That's what they get their fees for," he says.

Norma sold some $525 million in CDO slices -- largely the lower-rated ones with higher returns -- to investors. Merrill declined to say whether it kept Norma's triple-A rated, $975 million super-senior tranche or sold it to another financial institution.

Many investment banks with CDO businesses -- Citigroup Inc., Morgan Stanley and UBS -- frequently kept or bought these super-senior pieces, whose lower returns interested few investors. In doing so, they bet that the top CDO slices, which typically comprised as much as 60% of the whole CDO, were insulated from losses.

By September, Norma was in trouble. Amid a steep decline in house prices and rising defaults on mortgage loans, the value of subprime-backed securities went into a free fall. As increasingly worrisome delinquency data rolled in, analysts upped their estimates of total losses on subprime-backed securities issued in 2006 to 20% or more, a level that would wipe out most triple-B-rated securities.

Within weeks, ratings firms began to change their views. In October, Moody's downgraded $33.4 billion worth of mortgage-backed securities, including those which Norma had insured. Those downgrades set the stage for a review of CDOs backed by those securities -- and then further downgrades.

Mezzanine CDOs such as Norma were the hardest hit. On Nov. 2, Moody's slashed the ratings on seven of Norma's nine rated slices, three all the way from investment-grade to junk. Fitch downgraded all nine slices to junk, including two that it had rated triple-A.

Worse Performances

Other mezzanine CDOs, including some underwritten by other investment banks, have had worse performances. Around 30 are now in default, according to S&P. Norma is still paying interest on its securities. It is not known whether it has had to make payouts under the credit default swap agreements.

Ratings companies say their March opinions represented their best read at the time, and called the subprime deterioration unprecedented and unexpectedly rapid. "It's one of the worst performances that we've seen," says Kevin Kendra, a managing director at Fitch. "The world has changed quite drastically -- and our view of the world has changed quite drastically."

By mid-December, $153.5 billion in CDO slices had been downgraded, according to Deutsche Bank. Because banks owned the lion's share of the mezzanine CDOs, they bore the brunt of the losses. In all, banks' write-downs on mortgage investments announced so far add up to more than $70 billion.

For larger banks, holdings of mezzanine CDOs could account for one-third to three-quarters of the total losses. In addition to the $9.4 billion fourth-quarter write-down Morgan Stanley just announced it would take, Citigroup has projected its fourth-quarter write-down could reach $11 billion. UBS said this month it would take a $10 billion write-down after taking a $4.4 billion third-quarter loss.

Merrill, for its part, took a $7.9 billion write-down on mortgage-related holdings in the third quarter. Analysts expect it to write down a similar amount in the current quarter, which would represent the largest losses of any bank. News of the losses have led to the ouster of CEO Stan O'Neal and Osman Semerci, the bank's global head of fixed income. Mr. Margolis left this summer.


I haven't read this yet, but am looking forward to it. (OK, maybe I just like the title and cover :-)

Here is the author's web page; he's a psychology professor at Bennington.

Q: Do other countries have this problem?

A: The idea that it is unattractive or unappealing to be intelligent is not a universal concept. Not in Asia, certainly not in India. There's no concept in India that being good at math and science and technology has negative social consequences. That's the reason there are so many Indian engineers.

Q: Why did this grow out of American culture?

A: Historically, America is a place for men of action, for men who discover things, make things with their hands, have practical intelligence as opposed to book learning. Book-learning was suspect — the musty old European way, as opposed to practical, snazzy America. I think this tradition has never gone away.

The problem is that now it just doesn't work anymore. You can't do anything unless you pay attention in school. You can't invent things without knowing calculus. If you don't study math, it won't work. Benjamin Franklin was an American genius, a model of the American tinkerer, but the Ben Franklin model is not working anymore.

Q. Wasn't everyone talking about the need for better math and science education back in the days of Sputnik?

A: What's new is the sexualization of it. Kids live in such a sexualized world. ... (If you are called a nerd or a geek, it's) not just creepy or weird, you're labeled as someone who is never going to get laid. There's a lot more at stake because kids are so much more exposed to a culture that's all about being attractive, having sex early.

The nerds and the geek stereotype is that if you're doing well in math and science, you are completely unattractive to the opposite sex.

All the nerd and geek self-tests, what they ask you is: Are you good at science and math? Are you unwashed? Have you never had a date? You don't know anyone's phone number except your mothers?

Sunday, December 23, 2007

On earth peace, good will toward men

For years, when asked what I wanted for Christmas, I've been replying peace on earth, good will toward men :-)

No one ever seems to recognize that this comes from the bible, Luke 2.14 to be precise!

Linus said it best in A Charlie Brown Christmas:

And there were in the same country shepherds abiding in the field, keeping watch over their flock by night.

And, lo, the angel of the Lord came upon them, and the glory of the Lord shone round about them: and they were sore afraid.

And the angel said unto them, Fear not: for, behold, I bring you good tidings of great joy, which shall be to all people.

For unto you is born this day in the city of David a Saviour, which is Christ the Lord.

And this shall be a sign unto you; Ye shall find the babe wrapped in swaddling clothes, lying in a manger.

And suddenly there was with the angel a multitude of the heavenly host praising God, and saying,

Glory to God in the highest, and on earth peace, good will toward men.

Merry Christmas!

Friday, December 21, 2007

Vacation reading: Gregory Clark's A Farewell to Alms

A Farewell to Alms: A Brief Economic History of the World

Clark's book is an ambitious look at world economic history. The first half of the book is an excellent discussion of the Malthusian trap, in which increases in standard of living only lead to increases in population, which then (over generations) lead to declines in standard of living. The only stable point of this dynamics is at subsistence-level income. I need to think more about it, but I suspect Clark overstates the case for how well the Malthusian model applied in early human history. My impression is that there were wide disparities in levels of development that can't be easily explained in that context.

The second half of the book concerns the industrial revolution, and advances his (controversial) thesis that one of the main causes for this qualitative shift in the rate of human advancement was genetic. By analyzing historical demographic data he argues that by 1800 almost all residents of England were descended from previous generations of wealthy strivers -- reproduction rates correlated highly with family wealth in the previous Malthusian era, and the wealthy literally replaced the poor over time (less favored offspring of the rich often become the poor of future generations). Therefore, traits which are positive for commerce, long term planning, wealth accumulation, market organization, etc. had become much more widespread thanks to natural selection. I find this effect plausible -- it is consistent with recent genetic data on accelerated human evolution -- but am not as convinced that it dominates over cultural factors (at least, the two would work hand in hand). His case that it was a priori likely for England to be the first to have an industrial revolution doesn't seem particularly convincing (see Kenneth Pomeranz's Great Divergence for another set of arguments based on geography and natural resources).

Clark makes the interesting connection between modern man's descent from the strivers of previous generations and the hedonic treadmill: our happiness seems to correlate more with our position relative to perceived peers than with absolute levels of wealth.

I like the following quote from the final chapter of the book. I always found it very amusing that modern economic models can't do much better than to treat technological change as an exogenous, stochastic variable. (Yes, I know about Romer and growth theory, but would lump that in the "can't do much better" category.)

God clearly created the laws of the economic world in order to have a little fun at economists' expense. In other areas of inquiry, such as the physical sciences, there has been a steady accumulation of knowledge over the past four hundred years. Earlier theories proved inadequate. But those that replaced them encompassed the earlier theories and gave practitioners greater ability to predict outcomes across a wider range of conditions. In economics, however, we see instead that our ability to describe and predict the economic world reached a peak around 1800. In the years since the Industrial Revolution there has been a progressive and continuing disengagement of economic models from any ability to predict differences of income and wealth across time and across countries and regions.

Wednesday, December 19, 2007

How to run a hedge fund

Via Mark Thoma, this nice summary of how to run a hedge fund, taking advantage of the fee structure and the asymmetry of compensation and risk. If the fund goes up, the manager gets paid. If the fund goes down, the manager still gets paid the management fee, if not the performance component. This means he is incentivized to take asymmetric bets like selling insurance policies, which have a high probability of a decent return from the premium, but a small chance of blowing up due to a rare event. In the latter case the investors lose their money but not the manager. What looks like a decent return by the fund might in fact reflect crazy risk taking -- it's hard for investors to know.

But don't be too critical. These people are making our economy more efficient ;-)

Most managers don't behave this way, although it's an example of the more general agency problem, which arises whenever the incentives and interests of a principal differ from those of an agent (e.g., hired to manage a company or financial portfolio owned by the principal).

Washington Post: ...It is extremely difficult to tell, based on past performance, whether a fund is being run by true financial wizards, by no-talent managers who happen to get lucky or by outright scam artists.

To illustrate how easy it is to set up a hedge fund scam, consider the following example. An enterprising man named Oz sets up a new fund with the stated aim of earning 10 percent in excess of some benchmark rate of return, say 4 percent. The fund will run for five years, and investors can cash out at the end of each year if they wish. The fee is the standard '2 and 20': 2 percent annually for funds under management, and a 20 percent incentive fee for returns that exceed the benchmark.

Although he has no investment track record, Oz has a smooth manner, a doctorate in physics and many rich acquaintances. He raises $100 million and opens shop. He then studies the derivatives market and finds an event on which the market places fairly long odds, say 9:1. In other words, it costs $.10 to buy an option that pays $1 if the event occurs and $0 otherwise. The nature of the event is unimportant: it might be a large fall in the stock market, Florida getting hit by a Category 5 hurricane or Russian President Vladimir Putin dying before the end of the year.

Next Oz writes some covered options on this event and sells 110 million of them in the derivatives market. This obligates him to pay the option holders $110 million if the event does occur and nothing if it does not. He collects $11 million on the options. To cover his obligations in case the 'bad' event occurs, he uses the investors' money plus the proceeds from the options to buy $110 million in one-year Treasury bills yielding 4 percent, which he deposits in escrow. This leaves $1 million in "pocket money," which he uses to lease some computer terminals and hire a few geeks to sit in front of them, just in case his investors drop by.

The probability is ninety percent that the bad event does not occur and Oz owes nothing to the option holders. With a gross return (before expenses) of $15,400,000, the investors are thrilled, and so is Oz. He collects $2 million in management fees (of which he has only spent $1 million), plus a performance bonus equal to 20 percent of the 'excess return', namely, 20 percent of $11,400,000. All in all, Oz nets over $3 million for doing absolutely nothing.

Oz can then repeat the same gambit next year. When the fund terminates after five years, the chances are nearly 60 percent that the unlucky event will never have occurred. Oz looks like a genius and gets paid like a genius. ...

Sunday, December 16, 2007

Miami 2007: 10 years of AdS/CFT

I found this tag and notes when I sat down for the afternoon session :-)

Miami micro-climate is killing me: outside it's 80 and humid (with small rainstorms rolling through), but indoor temperatures seem to vary from 60-70 degrees! Sometimes I'm shivering under blasts of cold air in the lecture hall -- I've discovered you have to carry a jacket around with you all the time just in case the AC is on overkill.

Friday, December 14, 2007

More mutants

From the Economist, a nice summary of the Hawks et al. paper on a possible recent speedup in human evolution.
Economist: ...Dr Moyzis's paper suggests is a wider phenomenon—that Homo sapiens is continuing to undergo local evolution. He and his colleagues reckon they can both estimate the rate of evolution and identify many of the evolving genes, by using a trick with the clumsy name of linkage disequilibrium.

Genes are linked together in cell nuclei on structures called chromosomes. These come in pairs, one from each parent. However, when sperm and egg cells are formed, the maternal and paternal chromosomes swap bits of DNA to create a new mixture. The pieces of DNA swapped are complementary—that is, they contain the same types of gene. But they may contain different versions of the genes in question, and these different versions can have different biological effects.

Over the generations this process of swapping mixes the genes up thoroughly, and an equilibrium emerges. If a new mutation appears, however, it will take quite a while for that thorough mixing to happen. This means recent mutations can be spotted because they are still linked to the same neighbouring bits of DNA as they were when they first appeared. Moreover, the size of these neighbouring blocks gives an indication of how long ago the mutation in question emerged; long blocks suggest a recent mutation because the mixing process has not had time to break them up.

All this has been known for decades, but it is only recently that enough human DNA sequences have become available for the technique to be used to compare people from different parts of the world. And this is what Dr Moyzis and his colleagues have now done.

What they have found is that about 1,800 protein-coding genes, some 7% of the total known, show signs of having been subject to recent natural selection. By recent, they mean within the past 80,000 years. Moreover, as the chart shows, the rate of change has speeded up over the course of that period. (The sudden fall-off at the end is caused because the linkage-disequilibrium method cannot easily detect very recent mutations, rather than by a sudden reduction in the rate of evolution.) The researchers put this acceleration down to two things. First, the human population has expanded rapidly during that period, which increases the size of the gene pool in which mutations can occur. Second, the environment in which people find themselves has also changed rapidly, creating new contexts in which those mutations might have beneficial effects.

That environmental change itself has two causes. The past 80,000 years is the period in which humanity has spread out of Africa to the rest of the world, and each new place brings its own challenges. It has also been a period of enormous cultural change, and that, too, creates evolutionary pressures. In acknowledgment of these diverse circumstances, the researchers looked in detail at the DNA of four groups of people from around the planet: Yoruba from Africa, Han Chinese and Japanese from Asia, and Europeans.

Various themes emerged. An important one was protection from disease, suspected to be a consequence of the increased risk of infection that living in settlements brings. In this context, for example, various mutations of a gene called G6PD that are thought to offer protection from malaria sprang up independently in different places.

A second theme is response to changes in diet caused by the domestication of plants and animals. One example of this is variation in LCT, a gene involved in the metabolism of lactose, a sugar found in milk. All human babies can metabolise lactose, but only some adults can manage the trick. That fact, and the gene involved, have been known for some time. But Dr Moyzis's team have worked out the details of the evolution of LCT. They suspect that it was responsible for the sudden spread of the Indo-European group of humanity about 4,000 years ago, and also for the more recent spread of the Tutsis in Africa, whose ancestors independently evolved a tolerant version of the gene.

The pressures behind other changes are less obvious. In the past 2,000-3,000 years, for example, Europeans have undergone changes in the gene for a protein that moves potassium ions in and out of nerve cells and taste buds. There have also been European changes in genes linked to cancer and Alzheimer's disease. Chinese, Japanese and Europeans, meanwhile, have all seen changes in a serotonin transporter. Serotonin is one of the brain's messenger molecules, and is particularly involved in establishing mood.

The finding that may cause most controversy, however, is that in the Asian groups there has been strong selection for one variant of a gene that, in a different form, is responsible for Gaucher's disease. A few years ago two of the paper's other authors, Gregory Cochran and Henry Harpending, suggested that the Gaucher's form of the gene might be connected with the higher than average intelligence notable among Ashkenazi Jews. The unstated inference is that something similar might be true in Asians, too.

The Ashkenazim paper caused quite a stir at the time. It was merely a hypothesis, but it did suggest a programme of research that could be conducted to test the hypothesis. So far, no one—daring or foolish—has tried. Eventually, however, such questions will have to be faced. The paper Dr Moyzis and his colleagues have just published is a ranging shot, but the amount of recent human evolution it has exposed is surprising. Others will no doubt follow, and the genetic meaning of the term “race”, if it has one, will be exposed for all to see.

Goldman better than OK

It pays to be smart! +$4 B on the balance sheet from subprime bets (WSJ). The Times called this correctly in an earlier article I posted. Amazing that they let the reporter get so much color for the piece, but I guess it's especially flattering right now given magazine covers like this one:

Goldman = big prop trading hedge fund frontrunning clients? :-)

How Goldman Won Big On Mortgage Meltdown

A Team's Bearish Bets Netted Firm Billions; A Nudge From the CFO

December 14, 2007
The subprime-mortgage crisis has been a financial catastrophe for much of Wall Street. At Goldman Sachs Group Inc., thanks to a tiny group of traders, it has generated one of the biggest windfalls the securities industry has seen in years.

The group's big bet that securities backed by risky home loans would fall in value generated nearly $4 billion of profits during the year ended Nov. 30, according to people familiar with the firm's finances. Those gains erased $1.5 billion to $2 billion of mortgage-related losses elsewhere in the firm. On Tuesday, despite a terrible November and some of the worst market conditions in decades, Goldman is expected to report record net annual income of more than $11 billion.

Goldman's trading home run was blasted from an obscure corner of the firm's mortgage department -- the structured-products trading group, which now numbers about 16 traders. Two of them, Michael Swenson, 40 years old, and Josh Birnbaum, 35, pushed Goldman to wager that the subprime market was heading for trouble. Their boss, mortgage-department head Dan Sparks, 40, backed them up during heated debates about how much money the firm should risk. This year, the three men are expected to be paid between $5 million and $15 million apiece, people familiar with the matter say.

Under Chief Executive Lloyd Blankfein, Goldman has stood out on Wall Street for its penchant for rolling the dice with its own money. The upside of that approach was obvious in the third quarter: Despite credit-market turmoil, Goldman earned $2.9 billion, its second-best three-month period ever. Mr. Blankfein is set to be paid close to $70 million this year, according to one person familiar with the matter.

Goldman's success at wringing profits out of the subprime fiasco, however, raises questions about how the firm balances its responsibilities to its shareholders and to its clients. Goldman's mortgage department was a major underwriter of collateralized debt obligations, or CDOs, complex securities created from pools of subprime mortgages and other debt. When those securities plunged in value this year, Goldman's customers suffered major losses, as did units within Goldman itself, thanks to their CDO holdings. The question now being raised: Why did Goldman continue to peddle CDOs to customers early this year while its own traders were betting that CDO values would fall? A spokesman for Goldman Sachs declined to comment on the issue.

The structured-products trading group that executed the winning trades isn't involved in selling CDOs minted by Goldman, a task handled by others. Its principal job is to "make a market" for Goldman clients trading various financial instruments tied to mortgage-backed securities. That is, the group handles clients' buy and sell orders, often stepping in on the other side of trades if no other buyer or seller is available.

The group also has another mission: If it spots opportunity, it can trade Goldman's own capital to make a profit. And when it does, it doesn't necessarily have to share such information with clients, who may be making opposite bets. This year, Goldman's traders did a brisk business handling trades for clients who were bullish on the subprime-mortgage-securities market. At the same time, they used Goldman's money to bet that that market would fall. Some critics say that blending those functions creates a risk that traders could use information about client trades for their firm's own benefit.

Financial firms have good reason to keep a tight leash on proprietary traders. In 1995, bad bets by Nicholas Leeson, a young trader, led to $1.4 billion in losses and the collapse of Barings PLC. Last year, the hedge fund Amaranth Advisors shut down after a young Canadian trader lost more that $6 billion on natural-gas trades. But big trading wins such as George Soros's 1992 bet against the British pound, which netted more than $1 billion for his hedge fund, tend to be talked about for years.

The subprime trading gains notched by Messrs. Birnbaum and Swenson and their Goldman associates are large by recent Wall Street standards. Traders at Deutsche Bank AG and Morgan Stanley also bet against the subprime-mortgage market this year, but in each case, their gains were essentially wiped out because their firms underestimated how far the markets would fall. New York hedge-fund company Paulson & Co. also turned a considerable profit on the subprime meltdown this year, as did Hayman Capital Partners, a Dallas-based hedge-fund firm, say people familiar with the matter.

As recently as a year ago, few on Wall Street thought that the market for home loans made to risky borrowers, known as subprime mortgages, was heading for disaster. At that point, Goldman was bullish on bonds backed by such loans.

Hashing Out Risk

Last December, Mr. Sparks, a longtime trader of bond-related products, was named head of Goldman's 400-person mortgage department. That gave him a seat on the firm's risk committee, which numbers about 30 and meets weekly to hash out the firm's risk profile. It also gave him authority over the structured-products trading group, which then had just eight traders and was run jointly by Mr. Swenson and David Lehman, 30, a former Deutsche Bank trader.

Mr. Swenson, known as Swenny on the trading desk, is a former Williams College hockey player with four children and an acid wit. A veteran trader of asset-backed securities, he joined Goldman in 2000. In late 2005, he helped persuade Mr. Birnbaum, a Goldman veteran, to join the group. Mr. Birnbaum had developed and traded a new security tied to mortgage rates.

Mr. Swenson and Mr. Sparks, then No. 2 in the mortgage department, wanted Mr. Birnbaum to try his hand at trading related to the first ABX index, which was scheduled to launch in January 2006. Because securities backed by subprime mortgages trade privately and infrequently, their values are hard to determine. The ABX family of indexes was designed to reflect their values based on instruments called credit-default swaps. These swaps, in essence, are insurance contracts that pay out if the securities backed by subprime mortgages decline in value. Such swaps trade more actively, with their values rising and falling based on market sentiments about subprime default risk.

Messrs. Swenson and Sparks told Mr. Birnbaum the ABX was going to be a hot product, according to people with knowledge of their pitch.

They were right. On the first day of trading, Goldman netted $1 million in trading profits, people familiar with the matter say. But the index was tough to trade. In comparison to huge markets like Treasury bonds, there wasn't much buying and selling. That meant that Mr. Swenson's team nearly always had to use Goldman's capital to complete trades for clients looking to buy or sell.

Last December, David Viniar, Goldman's chief financial officer, gave the group a big push, asking it to adopt a more-bearish posture on the subprime market, according to people familiar with his instructions. Goldman had big exposure to the subprime mortgage market because of CDOs and other complex securities it was holding, Mr. Viniar told Mr. Sparks, these people say. Emerging signs of weakness in the market, he said, meant that Goldman needed to hedge its bets.

Mr. Swenson and his traders began shorting certain slices of the ABX, or betting against them, by buying credit-default swaps. At that time, new subprime mortgages still were being pumped out at a rapid clip, and gloom hadn't yet descended on the market. As a result, the swaps were relatively cheap.

Still, trading volume was thin, so it took months for the group to accumulate enough swaps to fully hedge Goldman's exposure to the subprime market. By February, Goldman had built up a sizable short position, and was poised to profit from the subprime meltdown.

The timing was nearly perfect. Goldman's bets were focused on an ABX index that reflects the value of a basket of securities that came to market in early 2006, known as the 06-2 index. Goldman bet that the riskiest portion of that index -- a sub-index that reflects the value of the slices of the securities with the lowest credit ratings -- would plunge in value. This January, as concerns about subprime mortgages grew, that sub-index dropped from about 95 to below 90. The traders handling the ABX trades were sitting on big profits.

Like other Wall Street firms, Goldman weighs its financial risk by calculating its average daily "value at risk," or VaR. It's meant to be a measure of how much money the firm could lose under adverse market conditions. Because the ABX had become so volatile, the VaR connected to the trades was soaring.

Goldman's co-president, Gary Cohn, who oversees the firm's trading business, became a frequent visitor, as did the firm's risk managers. More than once, Mr. Sparks was summoned to Mr. Blankfein's office to discuss the market. Goldman's top executives understood the group's strategy, say people with knowledge of the matter, but were uncompromising about the VaR. They demanded that risk be cut by as much as 50%, these people say.

Messrs. Swenson and Birnbaum, however, argued that the mortgage market was heading down, and Goldman should take full advantage by maintaining large short positions, people familiar with the matter say.

One day in late February, with the riskiest portion of the 06-2 index heading toward 60, the discussion about what to do grew heated, these people say. Mr. Birnbaum argued that Goldman would be leaving money on the table by unwinding some of the trades his group had used to bet on the mortgage market's decline.

"This is the wrong price" to close out the positions, Mr. Birnbaum snapped at a colleague assigned to help reduce risk, slamming down his phone receiver, these people say. He was overruled.

In March and April, the risky portion of the 06-2 index, which had taken a beating in February, bounced back from near 60 into the mid-70s. By then, the CDO underwriting business, which had been lucrative for Goldman, Merrill Lynch & Co. and other Wall Street firms, was grinding to a halt. Potential buyers had grown worried about the market.

Thanks to the wager that the ABX index would fall, Goldman's mortgage department earned several hundred million dollars during the first quarter, say people familiar with the matter. But the firm's concerns about risk had led the traders to unwind that bet. That left Goldman unhedged against further carnage, a worrisome situation for the second quarter.

In late April, Mr. Sparks, the mortgage-department chief, met with Mr. Cohn, the trading head, Mr. Viniar, the chief financial officer, and a couple of other senior executives. "We've got a big problem," Mr. Sparks told them as they paged through a handout listing the declining values of Goldman's CDO portfolio, according to people with knowledge of the meeting. Prices were heading straight down, he told them. He suggested that Goldman cancel a number of pending CDO deals, these people say, and sell whatever it could of the firm's roughly $10 billion in CDOs and related securities -- probably at a loss.

Led by Mr. Lehman, the co-head of the structured-products trading group, Goldman began selling off the majority of its CDO holdings. The losses pushed the mortgage group into the red for the second quarter.

By then, the subprime-mortgage market was cratering. Dozens of lenders had filed for bankruptcy protection, and legions of subprime borrowers were losing their homes. At Bear Stearns Cos., two internal hedge funds that had invested in risky portions of CDOs and other securities were struggling. Merrill, Citigroup Inc., and Morgan Stanley, among others, were sitting on billions of dollars in depreciating mortgage holdings.

Although it had become more expensive to wager against the ABX index, Messrs. Swenson and Birnbaum got a green light to once again ratchet up the firm's bet that securities backed by subprime mortgages would fall further. In July, the riskiest portion of the index plunged.

No Time for Breaks

The structured-products traders were working long hours. Mr. Swenson would leave his home in Northern New Jersey in time to hit the gym and be at his desk by 7:30 a.m. When Mr. Birnbaum arrived from his Manhattan loft, they'd begin executing large trades on behalf of clients. There was no time for breaks. They took breakfast and lunch at their desks -- for Mr. Swenson, the same chicken-and-vegetable salad every day from a nearby deli; for Mr. Birnbaum, an egg-white sandwich for breakfast, a chicken or turkey sandwich for lunch.

Mr. Sparks, the mortgage chief, climbed into his car at 5:30 each morning for the drive in from New Canaan, Conn. To calm his nerves, he'd stop by the gym in Goldman's downtown building to briefly jump rope and lift weights. Sometimes he worked past midnight, arriving home exhausted. He canceled a family ski trip to Wyoming. Although he loved to attend Texas A&M football games and owned a second home near the university, he decided not to join his wife and two children on more than one trip. (Mr. Sparks is a major donor to the university's athletic program.)

By late July, the Bear Stearns funds had collapsed and rumors were circulating of multibillion-dollar CDO losses at Merrill. Goldman was raking in profits.

But once again, concern was growing about VaR, the all-important measure of risk. At one point in July, senior executives called another meeting to demand the mortgage traders pull back, according to people familiar with the matter. The traders agreed.

Ratcheting Back

Around Labor Day, Mr. Birnbaum was asked to ratchet back one of his short positions by $250 million, according to Hayman Capital managing partner Kyle Bass, a client who had similar positions at the time. Mr. Bass says he made $100 million by relieving Goldman of that particular short bet. "It appeared to me that [the traders] constantly fought a VaR battle with the firm once the market started to break," says Mr. Bass.

In the first three quarters of its fiscal year, Goldman's VaR rose 38%, ending that period at $139 million per day, an all-time high, regulatory filings indicate.

During the third quarter ended Aug. 30, the structured-products trading group made more than $1 billion, say people knowledgeable about its performance. That helped the mortgage department notch record quarterly earnings of $800 million, these people say.

The subprime market continued to deteriorate through the fall. Both Merrill and Citigroup announced massive write-downs connected to the subprime mess, and their chief executive officers resigned.

Goldman pressed forward with its bearish bets on the ABX index, people familiar with its strategy say. In October, Goldman's mortgage unit moved from one downtown Manhattan office building to another. Despite their stellar year, traders were crowded into a low-ceiling floor where 150 employees shared one small men's room.

In late November, Mr. Sparks summoned Messrs. Birnbaum and Swenson to his office for separate visits. He thanked each trader for what he had done for the firm.

But there has been no time to relax. Two weeks into Goldman's new fiscal year, credit markets are looking bleaker than ever. Already, analysts are trimming their estimates of how much Goldman and other Wall Street firms will make in the coming year.

Greetings from Miami

I'm here for the Miami 2007 meeting. Everything is great, except they just charged us $7 for an espresso at lunch :-/

Tuesday, December 11, 2007

We are all mutants now

Some interesting new science suggests that human evolution has accelerated in the last tens of thousands of years. The study by Hawks, Wang, Cochran, Harpending and Moyzis (of UW Madison, Affymetrix, U Utah and UC Irvine) uses linkage disequilibrium tests on hapmap SNP data to determine that roughly 7% of all genes have undergone strong selection recently. The method looks for regions of DNA with similar SNP patterns. If an advantageous gene swept through a population in a relatively short time, replacing other variants, then the pattern of nucleotide polymorphisms in that area of the chromosome will be particularly uniform throughout the group. The results imply that we are all descended from mutants who, relatively recently, out-competed and replaced their contemporaries. The distribution of mutations is not uniform in different geographical populations (i.e., races). Recent evolution is causing genetic divergence, not convergence.

There is a good theoretical argument for why evolution may speed up due to population growth. Given a particular probability distribution for producing beneficial mutations, a large population implies a faster rate of incidence of such mutations. Because reproductive dynamics leads to exponential solutions (i.e., a slight increase in expected number of offspring compounds rapidly), the time required for an advantageous allele to sweep through a population only grows logarithmically with the population, while the rate of incidence grows linearly.

Note Cochran is a physicist who does evolutionary biology as a hobby :-) For some reason, this seldom happens in the opposite direction...

(original graphic from the Times; PNAS paper; University of Utah press release.)

NYTimes: Researchers analyzing variation in the human genome have concluded that human evolution accelerated enormously in the last 40,000 years under the force of natural selection.

The finding contradicts a widely held assumption that human evolution came to a halt 10,000 years ago or even 50,000 years ago. Some evolutionary psychologists, for example, assume that the mind has not evolved since the Ice Age ended 10,000 years ago.

But other experts expressed reservations about the new report, saying it is interesting but more work needs to be done.

The new survey — led by Robert K. Moyzis of the University of California, Irvine, and Henry C. Harpending of the University of Utah — developed a method of spotting human genes that have become more common through being favored by natural selection. They say that some 7 percent of human genes bear the signature of natural selection.

By dating the time that each of the genes came under selection, they have found that the rate of human evolution was fairly steady until about 50,000 years ago and then accelerated up until 10,000 years ago, they report in the current issue of The Proceedings of the National Academy of Sciences. The high rate of selection has probably continued to the present day, Dr. Moyzis said, but current data are not adequate to pick up recent selection.

The brisk rate of human selection occurred for two reasons, Dr. Moyzis’ team says. One was that the population started to grow, first in Africa and then in the rest of the world after the first modern humans left Africa. The larger size of the population meant that there were more mutations for natural selection to work on. The second reason for the accelerated evolution was that the expanding human populations in Africa and Eurasia were encountering climates and diseases to which they had to adapt genetically. The extra mutations in their growing populations allowed them to do so.

Dr. Moyzis said it was widely assumed that once people developed culture, they protected themselves from the environment and from the forces of natural selection. But people also had to adapt to the environments that their culture created, and the new analysis shows that evolution continued even faster than before.

The researchers took their data from the HapMap project, a survey designed by the National Institutes of Health to look at sites of common variation in the human genome and to help identify the genes responsible for common diseases. The HapMap data, generated by analyzing the genomes of people from Africa, East Asia and Europe, has also been a trove for people studying human evolutionary history.

David Reich, a population geneticist at the Harvard Medical School, said the new report was “a very interesting and exciting hypothesis” but that the authors had not ruled out other explanations of the data. The power of their test for selected genes falls off in looking both at more ancient and more recent events, he said, so the overall picture might not be correct.

Similar reservations were expressed by Jonathan Pritchard, a population geneticist at the University of Chicago.

“My feeling is that they haven’t been cautious enough,” he said. “This paper will probably stimulate others to study this question.”

University of Utah press release:



Media Contacts

Dec. 10, 2007 - Researchers discovered genetic evidence that human evolution is speeding up - and has not halted or proceeded at a constant rate, as had been thought - indicating that humans on different continents are becoming increasingly different.

"We used a new genomic technology to show that humans are evolving rapidly, and that the pace of change has accelerated a lot in the last 40,000 years, especially since the end of the Ice Age roughly 10,000 years ago," says research team leader Henry Harpending, a distinguished professor of anthropology at the University of Utah.

Harpending says there are provocative implications from the study, published online Monday, Dec. 10 in the journal Proceedings of the National Academy of Sciences:

"We aren't the same as people even 1,000 or 2,000 years ago," he says, which may explain, for example, part of the difference between Viking invaders and their peaceful Swedish descendants. "The dogma has been these are cultural fluctuations, but almost any Temperament trait you look at is under strong genetic influence."

"Human races are evolving away from each other," Harpending says. "Genes are evolving fast in Europe, Asia and Africa, but almost all of these are unique to their continent of origin. We are getting less alike, not merging into a single, mixed humanity." He says that is happening because humans dispersed from Africa to other regions 40,000 years ago, "and there has not been much flow of genes between the regions since then."
"Our study denies the widely held assumption or belief that modern humans [those who widely adopted advanced tools and art] appeared 40,000 years ago, have not changed since and that we are all pretty much the same. We show that humans are changing relatively rapidly on a scale of centuries to millennia, and that these changes are different in different continental groups."

The increase in human population from millions to billions in the last 10,000 years accelerated the rate of evolution because "we were in new environments to which we needed to adapt," Harpending adds. "And with a larger population, more mutations occurred."

Study co-author Gregory M. Cochran says: "History looks more and more like a science fiction novel in which mutants repeatedly arose and displaced normal humans - sometimes quietly, by surviving starvation and disease better, sometimes as a conquering horde. And we are those mutants."

Harpending conducted the study with Cochran, a New Mexico physicist, self-taught evolutionary biologist and adjunct professor of anthropology at the University of Utah; anthropologist John Hawks, a former Utah postdoctoral researcher now at the University of Wisconsin, Madison; geneticist Eric Wang of Affymetrix, Inc. in Santa Clara, Calif.; and biochemist Robert Moyzis of the University of California, Irvine.

No Justification for Discrimination

The new study comes from two of the same University of Utah scientists - Harpending and Cochran - who created a stir in 2005 when they published a study arguing that above-average intelligence in Ashkenazi Jews - those of northern European heritage - resulted from natural selection in medieval Europe, where they were pressured into jobs as financiers, traders, managers and tax collectors. Those who were smarter succeeded, grew wealthy and had bigger families to pass on their genes. Yet that intelligence also is linked to genetic diseases such as Tay-Sachs and Gaucher in Jews.

That study and others dealing with genetic differences among humans - whose DNA is more than 99 percent identical - generated fears such research will undermine the principle of human equality and justify racism and discrimination. Other critics question the quality of the science and argue culture plays a bigger role than genetics.

Harpending says genetic differences among different human populations "cannot be used to justify discrimination. Rights in the Constitution aren't predicated on utter equality. People have rights and should have opportunities whatever their group."

Analyzing SNPs of Evolutionary Acceleration

The study looked for genetic evidence of natural selection - the evolution of favorable gene mutations - during the past 80,000 years by analyzing DNA from 270 individuals in the International HapMap Project, an effort to identify variations in human genes that cause disease and can serve as targets for new medicines.

The new study looked specifically at genetic variations called "single nucleotide polymorphisms," or SNPs (pronounced "snips") which are single-point mutations in chromosomes that are spreading through a significant proportion of the population.

Imagine walking along two chromosomes - the same chromosome from two different people. Chromosomes are made of DNA, a twisting, ladder-like structure in which each rung is made of a "base pair" of amino acids, either G-C or A-T. Harpending says that about every 1,000 base pairs, there will be a difference between the two chromosomes. That is known as a SNP.

Data examined in the study included 3.9 million SNPs from the 270 people in four populations: Han Chinese, Japanese, Africa's Yoruba tribe and northern Europeans, represented largely by data from Utah Mormons, says Harpending.

Over time, chromosomes randomly break and recombine to create new versions or variants of the chromosome. "If a favorable mutation appears, then the number of copies of that chromosome will increase rapidly" in the population because people with the mutation are more likely to survive and reproduce, Harpending says.

"And if it increases rapidly, it becomes common in the population in a short time," he adds.

The researchers took advantage of that to determine if genes on chromosomes had evolved recently. Humans have 23 pairs of chromosomes, with each parent providing one copy of each of the 23. If the same chromosome from numerous people has a segment with an identical pattern of SNPs, that indicates that segment of the chromosome has not broken up and recombined recently.

That means a gene on that segment of chromosome must have evolved recently and fast; if it had evolved long ago, the chromosome would have broken and recombined.

Harpending and colleagues used a computer to scan the data for chromosome segments that had identical SNP patterns and thus had not broken and recombined, meaning they evolved recently. They also calculated how recently the genes evolved.

A key finding: 7 percent of human genes are undergoing rapid, recent evolution.

The researchers built a case that human evolution has accelerated by comparing genetic data with what the data should look like if human evolution had been constant:

The study found much more genetic diversity in the SNPs than would be expected if human evolution had remained constant.

If the rate at which new genes evolve in Africans was extrapolated back to 6 million years ago when humans and chimpanzees diverged, the genetic difference between modern chimps and humans would be 160 times greater than it really is. So the evolution rate of Africans represents a recent speedup in evolution.

If evolution had been fast and constant for a long time, there should be many recently evolved genes that have spread to everyone. Yet, the study revealed many genes still becoming more frequent in the population, indicating a recent evolutionary speedup.
Next, the researchers examined the history of human population size on each continent. They found that mutation patterns seen in the genome data were consistent with the hypothesis that evolution is faster in larger populations.

Evolutionary Change and Human History: Got Milk?

"Rapid population growth has been coupled with vast changes in cultures and ecology, creating new opportunities for adaptation," the study says. "The past 10,000 years have seen rapid skeletal and dental evolution in human populations, as well as the appearance of many new genetic responses to diet and disease."

The researchers note that human migrations into new Eurasian environments created selective pressures favoring less skin pigmentation (so more sunlight could be absorbed by skin to make vitamin D), adaptation to cold weather and dietary changes.

Because human population grew from several million at the end of the Ice Age to 6 billion now, more favored new genes have emerged and evolution has speeded up, both globally and among continental groups of people, Harpending says.

"We have to understand genetic change in order to understand history," he adds.

For example, in China and most of Africa, few people can digest fresh milk into adulthood. Yet in Sweden and Denmark, the gene that makes the milk-digesting enzyme lactase remains active, so "almost everyone can drink fresh milk," explaining why dairying is more common in Europe than in the Mediterranean and Africa, Harpending says.

He now is studying if the mutation that allowed lactose tolerance spurred some of history's great population expansions, including when speakers of Indo-European languages settled all the way from northwest India and central Asia through Persia and across Europe 4,000 to 5,000 years ago. He suspects milk drinking gave lactose-tolerant Indo-European speakers more energy, allowing them to conquer a large area.

But Harpending believes the speedup in human evolution "is a temporary state of affairs because of our new environments since the dispersal of modern humans 40,000 years ago and especially since the invention of agriculture 12,000 years ago. That changed our diet and changed our social systems. If you suddenly take hunter-gatherers and give them a diet of corn, they frequently get diabetes. We're still adapting to that. Several new genes we see spreading through the population are involved with helping us prosper with high-carbohydrate diet."

Sunday, December 09, 2007

Meet the bots: love and the Turing test

Nowadays you never know when you are interfacing with an alien machine intelligence :-)

Online chat, poker and chess are all infested by bots or other machine intelligences. Even those seeking romance can't be sure who or what is on the other end...

CNET: A program that can mimic online flirtation and then extract personal information from its unsuspecting conversation partners is making the rounds in Russian chat forums, according to security software firm PC Tools.

The artificial intelligence of CyberLover's automated chats is good enough that victims have a tough time distinguishing the "bot" from a real potential suitor, PC Tools said. The software can work quickly too, establishing up to 10 relationships in 30 minutes, PC Tools said. It compiles a report on every person it meets complete with name, contact information, and photos.

"As a tool that can be used by hackers to conduct identity fraud, CyberLover demonstrates an unprecedented level of social engineering," PC Tools senior malware analyst Sergei Shevchenko said in a statement.

Among CyberLover's creepy features is its ability to offer a range of different profiles from "romantic lover" to "sexual predator." It can also lead victims to a "personal" Web site, which could be used to deliver malware, PC Tools said.

Friday, December 07, 2007

CDO update

Here's a good summary from the Washington Post. (Link via Brad DeLong's blog.)

You can skip this beginning part if you know what a CDO is or know a mezzanine tranche from junk:

By now, almost everyone knows that most mortgages are no longer held by banks until they are paid off: They are packaged with other mortgages and sold to investors much like a bond. In the simple version, each investor owned a small percentage of the entire package and got the same yield as all the other investors. Then someone figured out that you could do a bigger business by selling them off in tranches corresponding to different levels of credit risk. Under this arrangement, if any of the mortgages in the pool defaulted, the riskiest tranche would absorb all the losses until its entire investment was wiped out, followed by the next riskiest and the next. With these tranches, mortgage debt could be divided among classes of investors. The riskiest tranches -- those with the lowest credit ratings -- were sold to hedge funds and junk bond funds whose investors wanted the higher yields that went with the higher risk. The safest ones, offering lower yields and Treasury-like AAA ratings, were snapped up by risk-averse pension funds and money market funds. The least sought-after tranches were those in the middle, the "mezzanine" tranches, which offered middling yields for supposedly moderate risks.

Stick with me now, because this is where it gets interesting. For it is at this point that the banks got the bright idea of buying up a bunch of mezzanine tranches from various pools. Then, using fancy computer models, they convinced themselves and the rating agencies that by repeating the same "tranching" process, they could use these mezzanine-rated assets to create a new set of securities -- some of them junk, some mezzanine, but the bulk of them with the AAA ratings more investors desired. It was a marvelous piece of financial alchemy, one that made Wall Street banks and the ratings agencies billions of dollars in fees. And because so much borrowed money was used -- in buying the original mortgages, buying the tranches for the CDOs and then in buying the tranches of the CDOs -- the whole thing was so highly leveraged that the returns, at least on paper, were very attractive. No wonder they were snatched up by British hedge funds, German savings banks, oil-rich Norwegian villages and Florida pension funds.

Here is the money paragraph:

What we know now, of course, is that the investment banks and ratings agencies underestimated the risk that mortgage defaults would rise so dramatically that even AAA investments could lose their value. One analysis, by Eidesis Capital, a fund specializing in CDOs, estimates that, of the CDOs issued during the peak years of 2006 and 2007, investors in all but the AAA tranches will lose all their money, and even those will suffer losses of 6 to 31 percent. And looking across the sector, J.P. Morgan's CDO analysts estimate that there will be at least $300 billion in eventual credit losses, the bulk of which is still hidden from public view. That includes at least $30 billion in additional write-downs at major banks and investment houses, and much more at hedge funds that, for the most part, remain in a state of denial.

...and the rest: it's bad, going to get worse.

As part of the unwinding process, the rating agencies are in the midst of a massive and embarrassing downgrading process that will force many banks, pension funds and money market funds to sell their CDO holdings into a market so bereft of buyers that, in one recent transaction, a desperate E-Trade was able to get only 27 cents on the dollar for its highly rated portfolio. Meanwhile, banks that are forced to hold on to their CDO assets will be required to set aside much more of their own capital as a financial cushion. That will sharply reduce the money they have available for making new loans. And it doesn't stop there. CDO losses now threaten the AAA ratings of a number of insurance companies that bought CDO paper or insured against CDO losses. And because some of those insurers also have provided insurance to investors in tax-exempt bonds, states and municipalities have decided to pull back on new bond offerings because investors have become skittish.

If all this sounds like a financial house of cards, that's because it is. And it is about to come crashing down, with serious consequences not only for banks and investors but for the economy as a whole. That's not just my opinion. It's why banks are husbanding their cash and why the outstanding stock of bank loans and commercial paper is shrinking dramatically. It is why Treasury officials are working overtime on schemes to stem the tide of mortgage foreclosures and provide a new vehicle to buy up CDO assets. It's why state and federal budget officials are anticipating sharp decreases in tax revenue next year. And it is why the Federal Reserve is now willing to toss aside concerns about inflation, the dollar and bailing out Wall Street, and move aggressively to cut interest rates and pump additional funds directly into the banking system.

This may not be 1929. But it's a good bet that it's way more serious than the junk bond crisis of 1987, the S&L crisis of 1990 or the bursting of the tech bubble in 2001.

I told you so: here is a post of mine from 2005 predicting that things would end in tears in credit derivative markets.

Repeat five times: it's not a black swan event, it was predictable; it's not that statistical modeling and correlations are useless, but rather how they are used; don't blame the quants; the incentive structures were all wrong; traders selling naked puts; blah blah blah.

Meritocracy in China

Here's the latest partially insightful and also partially completely wrong column by David Brooks, whom I can't stand to read with any regularity (although Bobos in Paradise is a good book).

The point is not that the testing system in China is inordinately based on memorization. (Although learning the written language does require a lot of memorization.) I am pretty sure that success in the testing systems used in northeast asia (including Japan, Korea, Taiwan, etc.) is relatively correlated with IQ and not just rote memorization. The point is that *any* system of mass testing is going to be a flawed predictor. Ideally, societies should allow multiple paths to success so that there isn't any single bottleneck. Entrepreneurism, business creation and risk taking are huge in China, so that creative mavericks who don't test well still have ample opportunity for success there.

Having said this, I do think that Asian countries need to fine-tune their meritocracies, in particular by encouraging more individuality and creativity. A few years ago I visited Beida (Beijing University -- the Harvard of China, as they say) and Tsinghua (usually referred to as the MIT of China, although as a Caltech guy it pains me to type that), with the director of a major institute of theoretical physics and mathematics as my guide. I was quite impressed by the huge, newly constructed buildings for nanotechnology, computer science, molecular biology, etc., but he was very careful to repeat a saying that I've heard many times in Asia: hardware is easy, software is hard. By this, they mean that infrastructure (buildings, roads, machines) is easy, but social organization (markets, rule of law) and encouraging new ways of thinking are not.

NYTimes: Let’s say you were born in China. You’re an only child. You have two parents and four grandparents doting on you. Sometimes they even call you a spoiled little emperor.

They instill in you the legacy of Confucianism, especially the values of hierarchy and hard work. They send you off to school. You learn that it takes phenomenal feats of memorization to learn the Chinese characters. You become shaped by China’s intense human capital policies.

You quickly understand what a visitor understands after dozens of conversations: that today’s China is a society obsessed with talent, and that the Chinese ruling elite recruits talent the way the N.B.A. does — rigorously, ruthless, in a completely elitist manner.

As you rise in school, you see that to get into an elite university, you need to ace the exams given at the end of your senior year. Chinese students have been taking exams like this for more than 1,000 years.

The exams don’t reward all mental skills. They reward the ability to work hard and memorize things. Your adolescence is oriented around those exams — the cram seminars, the hours of preparation.

Roughly nine million students take the tests each year. The top 1 percent will go to the elite universities. Some of the others will go to second-tier schools, at best. These unfortunates will find that, while their career prospects aren’t permanently foreclosed, the odds of great success are diminished. Suicide rates at these schools are high, as students come to feel they have failed their parents.

But you succeed. You ace the exams and get into Peking University. You treat your professors like gods and know that if you earn good grades you can join the Communist Party. Westerners think the Communist Party still has something to do with political ideology. You know there is no political philosophy in China except prosperity. The Communist Party is basically a gigantic Skull and Bones. It is one of the social networks its members use to build wealth together.

You are truly a golden child, because you succeed in university as well. You have a number of opportunities. You could get a job at an American multinational, learn capitalist skills and then come back and become an entrepreneur. But you decide to enter government service, which is less risky and gives you chances to get rich (under the table) and serve the nation.

In one sense, your choice doesn’t matter. Whether you are in business or government, you will be members of the same corpocracy. In the West, there are tensions between government and business elites. In China, these elites are part of the same social web, cooperating for mutual enrichment.

Your life is governed by the rules of the corpocracy. Teamwork is highly valued. There are no real ideological rivalries, but different social networks compete for power and wealth. And the system does reward talent. The wonderfully named Organization Department selects people who have proven their administrative competence. You work hard. You help administer provinces. You serve as an executive at state-owned enterprises in steel and communications. You rise quickly.

When you talk to Americans, you find that they have all these weird notions about Chinese communism. You try to tell them that China isn’t a communist country anymore. It’s got a different system: meritocratic paternalism. You joke: Imagine the Ivy League taking over the shell of the Communist Party and deciding not to change the name. Imagine the Harvard Alumni Association with an army.

This is a government of talents, you tell your American friends. It rules society the way a wise father rules the family. There is some consultation with citizens, but mostly members of the guardian class decide for themselves what will serve the greater good.

The meritocratic corpocracy absorbs rival power bases. Once it seemed that economic growth would create an independent middle class, but now it is clear that the affluent parts of society have been assimilated into the state/enterprise establishment. Once there were students lobbying for democracy, but now they are content with economic freedom and opportunity.

The corpocracy doesn’t stand still. Its members are quick to admit China’s weaknesses and quick to embrace modernizing reforms (so long as the reforms never challenge the political order).

Most of all, you believe, educated paternalism has delivered the goods. China is booming. Hundreds of millions rise out of poverty. There are malls in Shanghai richer than any American counterpart. Office towers shoot up, and the Audis clog the roads.

You feel pride in what the corpocracy has achieved and now expect it to lead China’s next stage of modernization — the transition from a manufacturing economy to a service economy. But in the back of your mind you wonder: Perhaps it’s simply impossible for a top-down memorization-based elite to organize a flexible, innovative information economy, no matter how brilliant its members are.

That’s a thought you don’t like to dwell on in the middle of the night.

Tuesday, December 04, 2007


The New Yorker has an amazing piece on endogenous retroviruses -- retroviruses that succeeded in injecting their RNA into our DNA via germline cells such as sperm or eggs. These fragments of genetic code constitute 8 percent of our entire genome, whereas protein-producing genes are only 2 percent.

Biologists can now reconstruct these long extinct viruses from the fragments, even correcting for mutations or copy errors by comparing different versions and using statistical models to guess at the original code. alter our genetic structure. That would require an organism to insinuate itself into the critical cells we need in order to reproduce: our germ cells. Only retroviruses, which reverse the usual flow of genetic code from DNA to RNA, are capable of that. A retrovirus stores its genetic information in a single-stranded molecule of RNA, instead of the more common double-stranded DNA. When it infects a cell, the virus deploys a special enzyme, called reverse transcriptase, that enables it to copy itself and then paste its own genes into the new cell’s DNA. It then becomes part of that cell forever; when the cell divides, the virus goes with it. Scientists have long suspected that if a retrovirus happens to infect a human sperm cell or egg, which is rare, and if that embryo survives—which is rarer still—the retrovirus could take its place in the blueprint of our species, passed from mother to child, and from one generation to the next, much like a gene for eye color or asthma.

When the sequence of the human genome was fully mapped, in 2003, researchers also discovered something they had not anticipated: our bodies are littered with the shards of such retroviruses, fragments of the chemical code from which all genetic material is made. It takes less than two per cent of our genome to create all the proteins necessary for us to live. Eight per cent, however, is composed of broken and disabled retroviruses, which, millions of years ago, managed to embed themselves in the DNA of our ancestors. They are called endogenous retroviruses, because once they infect the DNA of a species they become part of that species. One by one, though, after molecular battles that raged for thousands of generations, they have been defeated by evolution. Like dinosaur bones, these viral fragments are fossils. Instead of having been buried in sand, they reside within each of us, carrying a record that goes back millions of years. Because they no longer seem to serve a purpose or cause harm, these remnants have often been referred to as “junk DNA.” Many still manage to generate proteins, but scientists have never found one that functions properly in humans or that could make us sick.

Then, last year, Thierry Heidmann brought one back to life. ...

...The Nobel Prize-winning biologist Joshua Lederberg once wrote that the “single biggest threat to man’s continued dominance on this planet is the virus.” Harmit Malik, an evolutionary geneticist at the Fred Hutchinson Cancer Research Center, acknowledges the threat, yet he is confident that viruses may also provide one of our greatest scientific opportunities. Exploring that fundamental paradox—that our most talented parasites may also make us stronger—has become Malik’s passion. “We have been in an evolutionary arms race with viruses for at least one hundred million years,’’ he told me recently, when I visited his laboratory. “There is genetic conflict everywhere. You see it in processes that you would never suspect; in cell division, for instance, and in the production of proteins involved in the very essence of maintaining life.

“One party is winning and the other losing all the time,” Malik went on. “That’s evolution. It’s the world’s definitive game of cat and mouse. Viruses evolve, the host adapts, proteins change, viruses evade them. It never ends.” The AIDS virus, for example, has one gene, called “vif,” that does nothing but block a protein whose sole job is to stop the virus from making copies of itself. It simply takes that protein into the cellular equivalent of a trash can; if not for that gene, H.I.V. might have been a trivial disease. “To even think about the many million-year processes that caused that sort of evolution,” Malik said, shaking his head in wonder. “It’s dazzling.” Malik grew up in Bombay and studied chemical engineering at the Indian Institute of Technology there, one of the most prestigious technical institutions in a country obsessed with producing engineers. He gave no real thought to biology, but he was wholly uninspired by his other studies. “It was fair to say I had little interest in chemical engineering, and I happened to tell that to my faculty adviser,’’ he recalled. “He asked me what I liked. Well, I was reading Richard Dawkins at the time, his book ‘The Selfish Gene’ ”—which asserts that a gene will operate in its own interest even if that means destroying an organism that it inhabits or helped create. The concept fascinated Malik. “I was thinking of becoming a philosopher,’’ he said. “I thought I would study selfishness.” ...

Sunday, December 02, 2007

Some heresies from Freeman Dyson

See here, and this 2005 U Michigan commencement address.

Will Dyson get Watsonized, or are these heresies insufficiently dangerous?

..all the fuss about global warming is grossly exaggerated. Here I am opposing the holy brotherhood of climate model experts and the crowd of deluded citizens who believe the numbers predicted by the computer models. ... I have studied the climate models and I know what they can do. ... They do a very poor job of describing the clouds, the dust, the chemistry and the biology of fields and farms and forests. They do not begin to describe the real world that we live in.


...I say the United States has less than a century left of its turn as top nation. Since the modern nation-state was invented, about the year 1500, a succession of countries have taken turns as top nation. First it was Spain, then France, then and Britain, than America. Each term lasted about 150 years. Ours began in 1920 so it should end about 2070.

The reason why each top nation's term comes to an end is that the top nation becomes overextended militarily, economically and politically. Greater and greater efforts are required to maintain the number one position. Finally, the overextension becomes so extreme that the whole structure collapses. Already we can see in the American posture today some clear symptoms of overextension.

Who will be the next top nation? It might be the European Union or it might be China. After that it might be India or Brazil. You should be asking yourself not how to live in an America-dominated world, but how to prepare for a world that is not America dominated. That may be the most important problem for your generation to solve. How does a people who think of themselves as number one yield gracefully to become number two? So I'm telling you misfortunes are on the way.

Tuesday, November 27, 2007

Jim Simons is my hero

This is long, but worth the read! I just quoted the first part below. The whole thing can be found here.

Simons at Renaissance Cracks Code, Doubling Hedge Fund Assets 2007-11-27 00:13 (New York)

By Richard Teitelbaum
Nov. 27 (Bloomberg) -- On a hot afternoon in September, Renaissance Technologies LLC founder Jim Simons is too busy to take a phone call. It is, he says, from Cumrun Vafa, a preeminent Harvard University professor and expert on string theory, which describes the building blocks of the universe as extended one-dimensional filaments.

``Get another time when I can talk to him,'' Simons tells his assistant.

Then he mentions that the next day, he'll be meeting with Thomas Insel, director of the National Institute of Mental Health, to discuss autism research. And he's slated that Saturday to host a gala honoring Math for America, or MFA, a four-year-old nonprofit he started that provides stipends to New York City math teachers.

``I'm undoubtedly involved in too many things at the same time,'' Simons says in his 35th-floor office in midtown Manhattan. ``But you make your life interesting.''

String theory, autism, math education: It's fair to ask how Simons, 69, manages his day job overseeing the world's biggest hedge fund firm. The answer, judging from the numbers, is very well.

Renaissance is on fire: Its Medallion Fund -- which uses computers and trading algorithms to invest in world markets -- returned more than 50 percent in the first three quarters of 2007. It had about $6 billion in assets as of July 1.

Simons registered that performance as subprime and related markets were collapsing, sending two mortgage-related hedge funds run by Bear Stearns Cos. into bankruptcy. The turmoil pummeled the Goldman Sachs Global Alpha Fund, a rival to Renaissance's funds, which fell more than 25 percent during the same time. Morgan Stanley's computer jockeys lost $390 million in a single day in early August.

Life Story

Medallion's returns are no anomaly. The fund, which trades everything from soybean futures to French government bonds in rapid fire, hasn't had a negative quarter since early 1999. From the end of 1989 through 2006, it returned 38.5 percent annualized, net of fees.

More surprising than those returns is Simons's life story. At an age when hedge fund pioneers such as Michael Steinhardt have long since stopped managing other people's money, Simons is building on Medallion's success. He's adding funds and strategies and accumulating assets, which totaled $35.4 billion as of Sept. 28.

In August 2005, Simons started Renaissance Institutional Equities Fund, or RIEF, which invests in U.S. stocks. Through Sept. 30, it has returned 12.8 percent annualized. Unlike Medallion, which turns over its holdings dozens of times each year, RIEF keeps its positions for months or longer. Simons said at the time of the fund's inception RIEF could theoretically manage as much as $100 billion.

'New Possibilities'

In December 2006, he limited new investments in the fund to $1.5 billion a month. As of Sept. 30, 2007, it had $25.6 billion in assets.

In October, Simons started Renaissance Institutional Futures Fund, or RIFF, to invest in commodities. It's up 5.2 percent for the month. He says Renaissance's research shows the new fund can manage as much as $50 billion. Along with RIEF, it will promote cross-fertilization of ideas inside Renaissance, Simons says.

``Challenge is good,'' he says. ``It opens one's eyes to new possibilities.''

When not in Manhattan, Simons runs his empire from a 15-foot (4.6-meter) by 20-foot office in Renaissance's gated and guarded campus off Route 25A in East Setauket on New York's Long Island, some 50 miles (80 kilometers) east of the Empire State Building. With most of the trading automated, there's little of the hurly-burly of a typical hedge fund firm.

Doubling Assets

Along with routine personnel and marketing tasks, Simons makes time for the researchers and programmers who stop by his office to discuss mathematical and statistical issues they've encountered as they work on new trading strategies.

More than 200 employees, of whom about a third have Ph.D.s, work in East Setauket. Another 100 are based in Manhattan, San Francisco, London and Milan. ``He creates an environment where it's easy to be creative and works hard to keep the bullshit level to a minimum,'' says former managing director Robert Frey, who worked at Renaissance from 1992 to 2004.

Even without the new commodities fund, Renaissance's assets have more than doubled in a year from about $16 billion on Sept. 30, 2006. That growth has catapulted Renaissance past such titans as Daniel Och's Och-Ziff Capital Management Group LLC, Ray Dalio's Bridgewater Associates Inc. and David Shaw's D.E. Shaw & Co. to become the world's largest hedge fund manager, according to data compiled by Hedge Fund Research Inc. and Bloomberg.

Code Cracker

Medallion's 3.9 percent return during August, though that fund too was whipsawed by volatility, bolstered Simons's reputation as the silver-bearded wizard of quantitative investing.

In quant funds, mathematicians and computer scientists mine enormous amounts of data from financial markets looking for correlations among stocks, bonds, derivatives and other instruments. They search for predictive signals that will foretell whether, say, a palladium futures contract is likely to rise or fall.

``There are just a few individuals who have truly changed how we view the markets,'' says Theodore Aronson, principal of Aronson + Johnson + Ortiz LP, a quantitative money management firm in Philadelphia with $29.3 billion in assets. ``John Maynard Keynes is one of the few. Warren Buffett is one of the few. So is Jim Simons.''

Aronson credits Renaissance with validating the entire field of quantitative investing and proving that the freedom accorded to hedge fund managers to short stocks, borrow money and invest in myriad instruments can produce results that far outstrip typical market returns.

`Role Model'

Simons, standing just under 5 feet 10 inches tall and weighing 185 pounds (84 kilograms), has trod an unlikely path. A former code cracker for the U.S. National Security Agency, in 1968 he became chairman of the mathematics department at Stony Brook University, part of the New York state university system. He built the department into what David Eisenbud, former director of the Mathematical Sciences Research Institute in Berkeley, California, calls one of the world's top centers for geometry.

In 1977, frustrated with a math problem and eager for change, he abandoned academia to start what would become Renaissance, hiring professors, code breakers and statistically minded scientists and engineers who'd worked in astrophysics, language recognition theory and computer programming.

``All the quants in the world are trying to follow in Jim's footsteps because what he's built at Renaissance is truly extraordinary,'' says Andrew Lo, director of the Massachusetts Institute of Technology Laboratory for Financial Engineering and chief scientific officer of quant hedge fund firm AlphaSimplex Group LLC. ``I and many others look up to him as a tremendous role model.'' ...

Singularity summit

Someone once said that the Singularity is like the Rapture for geeks :-)

Nevertheless, some of the talks from the recent Singularity Summit are interesting.

So far I've enjoyed Omohundro, Jurvetson and Thiel. (Respectively, a physicist turned computer scientist, an engineer turned venture capitalist, and a derivatives trader turned PayPal CEO turned hedge fund manager.) Omohundro argues that AI beings will be much more rational than we are, using game-theoretic ideas of Von Neumann. Jurvetson thinks that we'll use evolutionary algorithms to create AI. But the resulting beings will be so complex that, while we understand the process of their creation, we won't understand how they really work -- any more than we understand evolved biological beings. Thiel's talk about investing for the singularity is goofy, but the comments near the end about Buffet's positions are quite interesting. The talk by robotics pioneer Rodney Brooks is surprisingly lame. There are quite a few talks I haven't yet had time to listen to...

Wednesday, November 21, 2007

Monster minds

There's a story in Surely You're Joking, Mr. Feynman, about the first seminar Feynman gives at Princeton. He's just a graduate student, working on a formulation of electromagnetism in terms of advanced and retarded potentials with his advisor Wheeler:

...So it was to be my first technical talk, and Wheeler made arrangements with Eugene Wigner to put it on the regular seminar schedule.

A day or two before the talk I saw Wigner in the hall. "Feynman," he said, "I think that work you're doing with Wheeler is very interesting, so I've invited Russell to the seminar." Henry Norris Russell, the famous, great astronomer of the day, was coming to the lecture!

Wigner went on. "I think Professor von Neumann would also be interested." Johnny von Neumann was the greatest mathematician around. "And Professor Pauli is visiting from Switzerland, it so happens, so I've invited Professor Pauli to come" - Pauli was a very famous physicist- and by that time, I'm turning yellow. Finally, Wigner said, "Professor Einstein only rarely comes to our weekly seminars, but your work is so interesting that I've invited him specially, so he's coming too."

By this time I must have turned green... Then came the time for the talk and here are these monster minds in front of me waiting!

Last Friday Sean Carroll emailed me to ask if I'd give a short blackboard talk at an informal cosmology meeting they have on Monday morning at Caltech. My real talk was in the afternoon, so I said, sure, no problem. I thought I'd mainly be talking to grad students and postdocs, so I didn't prepare anything. My plan was to give some background on entropy, information, black holes, etc. so that they could better follow the afternoon talk.

To my surprise, rather than a bunch of grad students I found monster minds arrayed around the big oak table in 469 Lauritsen: Carroll, Kamionkowski, Wise, Preskill, Politzer (Nobel laureate), Ooguri and Stanley Deser! No need for elementary background. I was kind of nervous at first, but we ended up having a lively discussion that lasted over 90 minutes -- I pretty much covered my whole talk using the blackboard, and ended up giving it again using slides later in the day. We had a funny moment when I first started discussing the ADM energy of the monsters. I looked over at Deser (the "D" in ADM) and smiled; he smiled back and nodded slightly :-) Having Stanley in the audience helped a lot because the entropy packing I described depends on using negative gravitational binding energy to nearly cancel the energy of the constituent matter. He was quite familiar with these constructions and helped convince the audience that I wasn't nuts.

Ten years ago I wrote a paper (unpublished) showing how to obtain a zero energy configuration in GR out of massive constituents. Particle theorists I discussed it with all thought I was crazy, but the referee was a very erudite relativist, who pointed out that a similar result (using different constructions) had been obtained by ADM, Novikov and Zeldovich, and others long ago. (So my result wasn't really new, but at least it wasn't wrong...) I had suspected Deser of being the referee but he said it wasn't him. He thought it might have been Wald... :-)

I had a wonderful visit, clouded only by the news (received by email on my cellphone while chatting with Sean) that Sidney Coleman had passed away on Sunday.

Sunday, November 18, 2007

Goldman OK?

According to the Times, Goldman unloaded their subprime exposure:

NYTimes: ... Late last year, as the markets roared along, David A. Viniar, Goldman’s chief financial officer, called a “mortgage risk” meeting in his meticulous 30th-floor office in Lower Manhattan.

At that point, the holdings of Goldman’s mortgage desk were down somewhat, but the notoriously nervous Mr. Viniar was worried about bigger problems. After reviewing the full portfolio with other executives, his message was clear: the bank should reduce its stockpile of mortgages and mortgage-related securities and buy expensive insurance as protection against further losses, a person briefed on the meeting said.

With its mix of swagger and contrary thinking, it was just the kind of bet that has long defined Goldman’s hard-nosed, go-it-alone style.

Most of the firm’s competitors, meanwhile, with the exception of the more specialized Lehman Brothers, appeared to barrel headlong into the mortgage markets. They kept packaging and trading complex securities for high fees without protecting themselves against the positions they were buying.

Even Goldman, which saw the problems coming, continued to package risky mortgages to sell to investors. Some of those investors took losses on those securities, while Goldman’s hedges were profitable.

When the credit markets seized up in late July, Goldman was in the enviable position of having offloaded the toxic products that Merrill Lynch, Citigroup, UBS, Bear Stearns and Morgan Stanley, among others, had kept buying.

“If you look at their profitability through a period of intense credit and mortgage market turmoil,” said Guy Moszkowski, an analyst at Merrill Lynch who covers the investment banks, “you’d have to give them an A-plus.”

This contrast in performance has been hard for competitors to swallow. The bank that seems to have a hand in so many deals and products and regions made more money in the boom and, at least so far, has managed to keep making money through the bust.

In turn, Goldman’s stock has significantly outperformed its peers. At the end of last week it was up about 13 percent for the year, compared with a drop of almost 14 percent for the XBD, the broker-dealer index that includes the leading Wall Street banks. Merrill Lynch, Bear Stearns and Citigroup are down almost 40 percent this year.

...At Goldman, the controller’s office — the group responsible for valuing the firm’s huge positions — has 1,100 people, including 20 Ph.D.’s. If there is a dispute, the controller is always deemed right unless the trading desk can make a convincing case for an alternate valuation. The bank says risk managers swap jobs with traders and bankers over a career and can be paid the same multimillion-dollar salaries as investment bankers.

“The risk controllers are taken very seriously,” Mr. Moszkowski said. “They have a level of authority and power that is, on balance, equivalent to the people running the cash registers. It’s not as clear that that happens everywhere.” ...

Blog Archive