SEC senior-level positions

For junior level positions, see previous post.

The U.S. Securities and Exchange Commission is seeking financial professionals at all levels of experience who strive to make a difference in capital markets to join its newly-formed Division of Risk, Strategy, and Financial Innovation

The Securities and Exchange Commission’s Division of Risk, Strategy, and Financial Innovation (“Risk Fin”) was created in the fall of 2009 with a broad mandate to support the Commission across a wide variety of economic, financial, and legal issues using a collaborative, multi-disciplinary approach. Risk Fin staff includes economists, financial engineers, lawyers, statisticians, and capital markets practitioners.

The Division consists of an Office of Policy and Development, an Office of the Chief Counsel, and five specialized Offices each supporting the Division’s goals with respect to analyses in the following areas:

Office of the Sell Side:
broker-dealers, prime brokers, rating agencies, and banks

Office of the Buy Side:
asset managers, hedge funds, and mutual funds

Office of Markets:
exchanges, clearinghouses, and market structure

Office of Litigation Support:
enforcement, damages, and distributions

Office of Data and Data Analytics:
financial data processing and analysis


The activities of each office as a whole include:
-
providing detailed, high-quality economic and other analyses, specific subject-matter expertise, and general guidance to the Commission and other Divisions/Offices
-
identifying and analyzing issues, trends, and innovations in the marketplace
-
where consistent with the Commission’s mission, writing internal white papers and thought-leadership pieces, and keeping abreast of, and contributing to, external peer-reviewed research on pertinent issues
-
developing analytical approaches, methods and models in order to identify trends, risks or potential securities law violations in the capital markets
-
working with outside experts in academia and industry to strengthen the Commission’s foundation of market knowledge
-
managing and analyzing public and private data to support relevant initiatives and projects
-
collaborating with other Offices in the Division, as well as other Divisions and Offices across the Commission
-
participating in, and contributing to, industry conferences


Risk Fin is currently hiring for each of its five specialized Offices. Qualified candidates should possess advanced quantitative and/or financial skills similar to those found in degrees such as: Masters of Finance or Financial Engineering, Masters of Business Administration (with a quantitative/finance concentration), or a Ph.D. in finance, economics, or statistics.

Work experience is not required. However, direct market knowledge gained through positions held at banks, exchanges, asset managers, broker-dealers, hedge funds, mutual funds, or related financial institutions would be helpful.


In addition to subject-matter expertise, candidates should demonstrate that they can:
-
effectively communicate (orally and in written form)
-
interact effectively, professionally, and courteously with a variety of personnel inside and outside of the Commission
-
exercise independent judgment, manage priorities and complete complex projects
-
collaborate with others and foster a productive and enjoyable team environment


Terms
−
Candidates will be hired with an initial appointment of two years, extendable to a maximum of four years in total.
−
All positions are located at the SEC’s headquarters in Washington, DC.
−
Salaries are competitive and based on education and relevant work experience.
−
Positions are open to US citizens and non-US citizens* from any treaty-allied nation (see http://www.state.gov/s/l/treaty/collectivedefense/). To apply, please send a cover letter and resume to riskfinrecruiting@sec.gov

*Please include your country of citizenship in either your resume or cover letter.

SEC jobs

I received the following from the SEC. If you are interested, let me know. There may be more senior positions available with the possibility of real input into future reform. If you are a math / science student or postdoc, and have ethical qualms about working on Wall Street, this might be something to consider.

Quantitative Risk Analyst at US Securities and Exchange Commission

Location: Washington, DC (Washington D.C. Metro Area)
Compensation: $80,000 - $130,000 Per Year


Job Description


The SEC's new Division of Risk, Strategy and Financial Innovation is looking for junior quantitative risk analysts. This division is of great importance given proposed legislation, which may increase the SEC’s responsibility for derivatives, hedge funds and credit agencies, as well as give the SEC a seat at the systemic Risk Council. Although the division is only a few months old and is quite small, it already has high visibility both within and outside the SEC.

The pay scale obviously is less than that in private industry; it will be between $80K and $130K depending on your experience and training. Another important difference in the government is that we hire full-time, permanent employees through an open competition - which means a significant process. We can also bring employees on board relatively more quickly if employment is limited to a period of a few years. For those who are interested in public service or making an important contribution to the markets, there will be the opportunity to be involved in any number of efforts that will affect the investment regulatory environment for many years to come.


Skills


Here is the kind of person we are looking for:

# Strong undergraduate training in math, computer science and statistics.
# An ability to program and code. This will include areas like:
Database knowledge, because a lot of what you do will require that you manage or pull in the data you use.
Developing prototype systems and models. In our case, not models for trading but for targeting where to focus our examinations and investigations.

Do analytical work, such as regressions and factor analysis. Part of this will be to contribute to the broad agenda on systemic risk.
# Two or more years of experience in the financial industry.
However, we can look at people fresh out of school if you are very strong technically and are willing to work at the lower pay scale.
# A masters degree in financial engineering or similar field is valuable but not necessary.
# The work will be in DC, and we would like to have a two year commitment on your part.


Company Description


The mission of the U.S. Securities and Exchange Commission is to protect investors, maintain fair, orderly, and efficient markets, and facilitate capital formation.

The new division of Risk, Strategy and Financial Innovation combines the Office of Economic Analysis, the Office of Risk Assessment, and other functions to provide the Commission with sophisticated analysis that integrates economic, financial, and legal disciplines. The division’s responsibilities cover three broad areas: risk and economic analysis; strategic research; and financial innovation. With the creation of the new division, the SEC now has five divisions, including the Division of Corporation Finance, the Division of Enforcement, the Division of Investment Management, and the Division of Trading and Markets. This is the first new division at the SEC since 1972.

The new division will perform all of the functions previously performed by the Office of Economic Analysis and the Office of Risk Analysis, along with the following: (1) strategic and long-term analysis; (2) identifying new developments and trends in financial markets and systemic risk; (3) making recommendations as to how these new developments and trends affect the Commission's regulatory activities; (4) conducting research and analysis in furtherance and support of the functions of the Commission and its divisions and offices; and (5) providing training on new developments and trends and other matters.

See you in court

The best reporting on the SEC charges against Goldman that I've found so far is by Felix Salmon at Reuters. See the post I linked to and several more he's made in the past day or two.

I still think the SEC will have a tough time with this case. In terms of public opinion (no nuance), this is definitely a black eye for Goldman. But a court of law is a different thing, one hopes.

Below are two comments from Felix's blog that sum up the issue nicely. ABN/IKB are the large European banks that were big buyers of the senior tranches of the Abacus deal in question (more correctly, ABN was an insurer of the Abacus risk). ACA is the Abacus CDO manager that collected millions of dollars in fees and put its reputation on the line. The records show that they rejected many of Paulson's suggestions for reference names to be placed in the structure.

The bottom line is that when a trade occurs the buyer and seller typically have different views on the likely future of the security. See my previous post for some color on what the world looked like when the Abacus deal was being done. If ABN/IKB had done well with their investment they might be chuckling today: "Yeah, we took a lot of money from those short chumps back in 2007. Can you believe we got 100 bps for free for AAA! Who is John Paulson?"

Somehow it's easier to blame this disaster on evil fraudsters than on plain stupidity.

Well, that’s all very nice, but what is it that ACA was supposed to have done for the fee it was paid? Just accept Paulson’s selections? Were they not supposed to be you know actually evaluating the structure and underlying assets?

When a firm gets millions of dollars to validate an investment, doesn’t it have some obligation to do some research?


*********************


I'm still at a loss to understand why sellers of credit protection would have changed their mind if they knew the buyer of the protection structured the transaction.

Do ABN/IKB assume all market participants are lazy and gullible as they are? Popeye the Sailor could have selected the collateral but it was incumbent on ABN and IKB to perform their own due diligence on the underlying assets.

Paulson/GS also had no inside or non-public information that these bonds would default. They structured the CDO based on their view that the bonds were likely to default. This information was also available to ACA, ABN, IKB.

Paulson/GS certainly didn’t force the homeowners to default on their payments or cause the bond trustee to declare the bonds in default.

If I sell you a share of Apple stock, chances are I am less optimistic than you are about the future price. However, it's also possible I'm selling it just for risk management or asset allocation reasons; maybe I have too much concentration in my portfolio so I'm selling it just to rebalance. But we usually don't require that the buyer has to know my reasons for selling the share. The SEC is more or less saying Goldman committed fraud by not letting the buyers know that Paulson really thought subprime was a bubble, and wasn't simply hedging his mortgage positions. Yes, Paulson suggested names for the structure, but ACA was there to vet all of them.

The WSJ has a nice analysis of ACA's role, and its centrality to the case, here.

This NYTimes article describes the internal dynamics at Goldman -- until very late there was no consensus that CDOs built from subprime assets would melt down.

How it looked to Paulson

Why did Paulson and Magnetar need to participate in the creation of new CDOs? Apparently Paulson had trouble finding counterparties at the time willing to take direct billion dollar bets on CDS indices. If I had had a decent way to short subprime in 2005 or 2006 I would have done it too -- see post1 and post2 from 2004!

Regarding the SEC charges against Goldman: the buyers of synthetic CDOs should have realized that there must have been short interest on the other side of the deal, and at the time Paulson wasn't a prominent figure. His short strategy was very contrarian and took enormous guts. I think this is going to be a very tricky case for the SEC.

Here's a revealing excerpt from WSJ reporter Gregory Zuckerman's book The Greatest Trade Ever. Read the whole excerpt!

WSJ: ... Mr. Paulson traveled to Boston to meet with Mark Taborsky, who helped pick hedge funds for Harvard's endowment. Mr. Taborsky was wary. Mr. Paulson's fund was willing to lose 8% a year to buy the mortgage insurance, which seemed like a lot. Mr. Taborsky also thought Mr. Paulson might be excessively gloomy about the housing market. Mr. Taborsky turned him down, too.

Even some investors who agreed with Mr. Paulson's view that housing prices would tumble doubted he would make much money because there was relatively little trading in the investments he was buying. He might have a hard time selling his investments without sending prices tumbling, shrinking any profits, they said.

"It looked like a dangerous game, taking one single bet that might be difficult to unwind," said Jack Doueck, a principal at Stillwater Capital, a New York firm that parcels out money to funds. He, too, said no to Mr. Paulson's fund.

Mr. Paulson's growing fixation on housing began to spark doubts about his business. One long-time client, big Swiss bank Union Bancaire Privée, received an urgent warning from a contact that Mr. Paulson was "straying" from his longtime focus, and that the bank should pull its money from Paulson & Co., fast. The bank stuck with Mr. Paulson but turned down his new fund.

... By the summer of 2006, Mr. Paulson had managed to raise $147 million, mostly from friends and family, to launch a fund. Soon, Josh Birnbaum, a top Goldman Sachs trader, began calling and asked to come by his office. Sitting across from Mr. Paulson, Mr. Pellegrini, and his top trader, Brad Rosenberg, Mr. Birnbaum got to the point.

Not only were Mr. Birnbaum's clients eager to buy some of the mortgages that Paulson & Co. was betting against, but Mr. Birnbaum was, too. Mr. Birnbaum and his clients expected the mortgages, packaged as securities, to hold their value. "We've done the work and we don't see them taking losses," Mr. Birnbaum said.

After Mr. Birnbaum left, Mr. Rosenberg walked into Mr. Paulson's office, a bit shaken. Mr. Paulson seemed unmoved. "Keep buying, Brad," Mr. Paulson told Mr. Rosenberg.

[Yes, in hindsight everyone knows that subprime loans were "toxic waste" -- but at the time lots of smart money didn't think so. Mr. Market needs a little help sometimes ... Steve]

Months into their new fund, Mr. Paulson and Mr. Pellegrini were eager to find more ways to bet against risky mortgages. Accumulating mortgage insurance in the market sometimes proved slow. They soon found a creative and controversial way to enlarge their trade.

They met with bankers at Bear Stearns, Deutsche Bank, Goldman Sachs, and other firms to ask if they would create securities—packages of mortgages called collateralized debt obligations, or CDOs—that Paulson & Co. could wager against.

The investment banks would sell the CDOs to clients who believed the value of the mortgages would hold up. Mr. Paulson would buy CDS insurance on the CDO mortgage investments—a bet that they would fall in value. This way, Mr. Paulson could wager against $1 billion or so of mortgage debt in one fell swoop.

Paulson & Co. wasn't doing anything new. A few other hedge funds also worked with banks to short CDOs the banks were creating. Hundreds of other CDOs were being created at the time. Other bankers, including those at Deutsche Bank and Goldman Sachs, didn't see anything wrong with Mr. Paulson's request and agreed to work with his team.

At Bear Stearns, however, Scott Eichel, a senior trader, and others met with Mr. Paulson and later turned him down. Mr. Eichel said he felt it would look improper for his firm. "On the one hand, we'd be selling the deals" to investors, without telling them that a bearish hedge fund was the impetus for the transaction, Mr. Eichel told a colleague; on the other hand, Bear Stearns would be helping Mr. Paulson wager against the deals.

Some investors later would argue that Mr. Paulson's actions indirectly led to the creation of additional dangerous CDO investments, resulting in billions of dollars of additional losses for those who owned the CDO slices.

At the time, though, Mr. Paulson still wasn't sure his trade would work. He simply was buying protection, he said. "We didn't create any securities, we never sold the securities to investors," Mr. Paulson said. "We always thought they were bad loans."

Enough Magnetar, now on to Paulson, Goldman and the SEC

See Felix Salmon for a discussion of the SEC complaint against Goldman. They allege that Goldman violated its fiducial responsibilities in marketing a synthetic CDO called ABACUS. The SEC allegations are very similar to the story told by ProPublica and This American Life, with John Paulson's fund playing the role of Magnetar.

This is being covered aggressively by the NY Times and WSJ as well.

The Magnetar trade part one, two, three.

The Magnetar trade, part 3

An MBS guy writes:

You wrote: "... there may have been motivation for Magnetar to pressure the underwriters to include particularly toxic mortgages in the CDOs they sponsored." No, I doubt it. Magnetar pressured the CDO managers (*not* the underwriter, which were the banks) to put in HIGHER YIELDING assets -- now, higher yielding assets are probably crappier, but they didn't care about the credit quality per say, only the high coupon with which they can achieve the positive carry -- i.e., getting paid while making the trade.

Furthermore, after reading that Magnetar asked Janet to provide advice, I must say, these guys aren't that sophisticated. Janet's books are great, but are NOT technical books and are horrible places for seeking trade ideas! You're better off asking the research desk of any investment bank. Please see the attached CDS primer from Lehman circa 2005, which shows you the pricing discrepancies that exist in the market, and how you can profit from it (page 10, under "valuation considerations"). They even provide a nice table (see figure 11 on page 12) on the returns given different assumptions of leverage and borrowing cost. This is just one document however, and plenty of other guides available that time will tell you how to hedge the default risk by shorting a more senior tranche - which implies that you believe correlation is high (you're long correlation -- see page 8 on the correlation trading guide from Merrill in 03 - this is for corp bond CDOs, but whatever, same concept).

So yeah, I doubt these guys knew enough to be able to figure out which loans were crappy even if they tried. If they were so good at picking out bad loans, they'd have just taken a bet (trading a view) on the crappy loans via an ABS-CDS on that crappy bond pool. Why go through all the trouble of setting up a CDO?

So in summary - these guys didn't purposefully set up crappy CDOs because

1) they frankly weren't that smart and didn't seem that knowledgeable (they made money because they were lucky - Howie Hubler blew up and lost $8 billion doing a correlation trade, and he was the smartest bond trader at Morgan Stanley), and therefore

2) couldn't have picked out which bonds were crappy even if they tried.

Here is the Lehman CDS primer (20 pages) and here is a Merrill primer on correlation trading (37 pages).

Hmm... what else could these brainy guys have been doing instead of working on "financial innovation"?

See A reallocation of human capital and The bubble algorithm for human capital allocation.








The Magnetar trade part one, two, three.

The Magnetar trade, part 2

Janet Tavakoli pointed me to this WSJ article which covered the Magnetar strategy already in 2008. Tavakoli, and another correspondent (anonymous) who works in MBS, both believe Magnetar's claim that their strategy was market neutral (in a sense; see below). It was actually a common correlation trade at the time: assuming a certain correlation between the fate of the risky equity tranche and the AAA senior tranches, one could guarantee a positive return by investing in the former and shorting the latter using a CDS contract. The strategy would make money regardless of the default rate (assuming the correlation), but it seems the upside was larger from the CDS contracts. So there may yet have been motivation for Magnetar to pressure the underwriters to include particularly toxic mortgages in the CDOs they sponsored (as alleged).

Tavakoli and my other correspondent both believe that ProPublica and This American Life have the story wrong. It may be appealing to think that a single hedge fund was capable of exacerbating the credit crisis through its participation in the creation of $30-$40 billion in CDOs, but the real story is probably more complex.

WSJ: ... CDOs are sliced based on risk, with the riskiest pieces having the highest yield but the greatest chance of losing value. Less-risky pieces have lower yields and some pieces were once considered so safe that they paid only a bit more than a U.S. Treasury bond.

Magnetar helped to spawn CDOs by buying the riskiest slices of the instruments, which paid returns of around 20% during good times, according to people familiar with its strategy. Back in 2006, when Magnetar began investing, these were the slices Wall Street found hardest to sell because they would be the first to lose money if subprime defaults rose.

For the Wall Street firms underwriting the deals, selling the riskiest pieces was "critical to getting the deals done because they were designed to act as a cushion for other investors," says Eileen Murphy, principal at Excelsior CDO Advisors LLC, a structured-finance consultancy.

Magnetar then hedged its holdings by betting against the less-risky slices of some of these same securities as well as other CDOs, according to people familiar with its strategy. While it lost money on many of the risky slices it bought, it made far more when its hedges paid off as the market collapsed in the second half of last year.

Here is Tavakoli's comment on the ProPublica story -- she was actually asked for advice by Magnetar on how to put on the trade.

Tavakoli on the missed opportunity to grill Prince and Rubin under oath on Citi's CDO activities. Was senior management negligent (understood the risks, and approved them) or incompetent (didn't understand their own CDO business)?


The Magnetar trade part one, two, three.

The Magnetar trade

Just when I thought I was out, they pull me back in...

I thought I was done discussing all of this horrible CDS, CDO, credit crisis stuff, but here we go again. Yesterday I listened to this podcast from This American Life, which details the activities of a large hedge fund called Magnetar. See here for the original reporting done by ProPublica.

The Magnetar team ("Smart. Very smart"), it is claimed, put on a very clever trade in 2006, after spending 2005 surveying the MBS market and realizing that subprime was an irrational bubble. Magnetar actually created demand for CDO structures by agreeing to fund the riskiest equity tranches. By putting up $10 million (according to the reporting) they could help initiate the construction of a $1 billion dollar security. (If the models show that the expected loss is small, and someone is willing to expose themselves to it, other investors can come in later and take the senior tranches, which could have AAA ratings.) Magnetar apparently lobbied the banks assembling the CDOs to include the riskiest subprime mortgages in the structure, because they intended to bet against the more senior tranches using CDS contracts. The underwriters and asset managers went along with this in order to book fees (also in the tens of millions of dollars) collected at completion of the transation. These transactions generated fake alpha -- short term profits for the banks, at the cost of taking on hidden tail risk. The guys who did these deals made huge bonuses. When the CDOs tanked, Magnetar made huge profits through its CDS contracts. The senior tranches were typically sold to other banks, pension funds, investors, but a large portion were actually kept on the books of the banks that underwrote them. The CEOs of these banks were often unaware of the tremendous risks (potential losses of tens of billions of dollars, hidden in the AAA senior tranches) they were taking on in order to book hundreds of millions of dollars in fees during the bubble.

Comments:

1. Cognition is bounded. Magnetar understood what it was doing. Some of the structurers at the banks understood what was going on, but their incentives were to take the short term fees and do the transaction (who cares? IBG YBG = "I'll be gone, You'll be gone"). The counterparties who bought the senior tranches didn't know what was going on, and neither (I'll bet) did the CEOs of the banks that did the structuring. Certainly the shareholders of these firms didn't understand the risks. Mr. Market priced all these deals terribly.

2. Nobody in 2006 said "Paulson / Bernanke / Obama / Geithner / ... will definitely bail us out if this gets too hairy. We're too big to fail!" Most of the MBS guys lost their jobs in the crash. But they made plenty of money while the getting was good.

3. Some people violated their fiducial responsibilities in this game. But it appears they're going to get away with it.

For more, see Naked Capitalism and Yves Smith's book Econned.

Small correction for the NPR guys: a magnetar is a neutron star (pulsar) with very high magnetic field. It's not a black hole.


Note added: In letters to ProPublica Magnetar claims it was employing a market neutral stat arb strategy, and that the CDOs they sponsored were not built to fail. See here and here. My interest in this story is not Magnetar, per se, but rather what it reveals about the MBS industry in general, leading up to the credit crisis.

Here is what Yves Smith has to say about the stat arb strategy:

... While Magnetar paid roughly 5% of the total deal value for its equity stake, it took a much bigger short position by acting as a protection buyer on some of the credit default swaps created by these same CDOs. This insurance in turn was artificially cheap because over 80% of the deal was rated AAA. Most investors did not understand what Magnetar recognized: this concentrated exposure to the very riskiest type of bond associated with risky mortgage borrowers, each of these CDOs was a binary bet. It would either work out (in which case Magnetar would still show a thin profit) or it would fail completely, giving Magnetar an enormous profit and wiping out even the AAA investors who mistakenly believed they were protected by having other investors sit below them and take losses first. Thus the AAA investors were only earning AAA returns for BBB risk.

The Magnetar trade part one, two, three.

Janet Tavakoli interview

This was recommended by a commenter on my previous post. It's definitely worth watching if you have an interest in the financial crisis.



Here's a recommendation for her book:

"[T]he book’s real strength is the sub-plot that emerges as Tavakoli tugs vigorously at the seemingly disparate threads of the current financial crisis, naming names, citing cases and leaving no schmuck — whether investment bank, credit rating agency, monoline insurer, mortgage brokers, regulators and their ilk — unspared. Based on more than 20 years in the derivatives arena, and having served time at Salomon Bros, Bear Stearns and Goldman Sachs, she knows that of what and who she speaks. Should anyone ever display the slightest interest in criminalizing the criminals who led us down this path, a prosecutor could do worse than ordering up copies for the grand jury."

Here's a nice figure from her web site:

Moral hazard and the financial crisis

You may have noticed that I've written little about the financial crisis recently, despite having written a great deal about it before and during the events (see here for slides from my talk on the crisis). I suppose it's because I'm 1. tired of thinking about it, and 2. pessimistic about what will be learned (by regulators, economists, the general public) from the experience. The subject is just too complex and requires a combination of detailed knowledge of financial markets together with an appreciation of theoretical issues such as market efficiency, agency problems and regulatory capture. On a positive note, some economists who were, at the time of the crisis, familiar with the latter, but (shockingly) had little knowledge of the former (what is a CDS? is that like a CDO?), are finally getting up to speed.

See here for a link to what I consider one of the best papers characterizing systemic risks in the "New Financial Architecture". It seems that ideology still rules in certain places.

A number of talented people have written books on the crisis -- from economists like Simon Johnson to financial journalists like Gillian Tett and Michael Lewis. Note the absence of books by actual practitioners -- to my knowledge no CDO modeler or CDS trader has written a book [see comments for a correction to this statement]. I suppose Paulson's book or Congressional testimony by people like Rubin might give the CEO's perspective, but those people were plausibly, and by their own admission, clueless about what was being done in their own firms leading up to the crisis.

Yesterday I listened to this bloggingheads discussion: The Past, the Crash, and the Future, between two Brown economists. Ross Levine is a finance expert and Glenn Loury is an interlocutor with a gift for clarity. I think they got almost everything right except that Levine continues to blame moral hazard. At the naive level, moral hazard is a big issue because many of the bad actors (banks) were indeed bailed out by taxpayers like you and me. However, the story only works at the institutional level -- yes, Citi and JP Morgan Chase and Goldman all continue to exist as institutions. But the principals -- the actual people -- involved could never have been sure, at the time, or leading up to the crisis, about their own fates or those of their companies. I suspect most Goldman partners were scared out of their minds at the height of the crisis. And, certainly, people involved in the actual mortgage finance business within those companies have mostly lost their jobs. One friend of mine (a prop trader) noted, referring to the mortgage business, that an entire industry had disappeared, and that none of those jobs were coming back.

Compare this to the outcome if the traders and modelers and salesmen in the mortgage industry had produced real alpha, as opposed to fake alpha -- they'd still be gainfully employed. The guys collecting huge bonuses in 2010 are, by and large, not the same people who nearly destroyed the entire financial system leading up to 2008. (The public is upset because they belong to the same firms, share the same backgrounds, and do similar things. But a guy who trades FX isn't really responsible for the mortgage meltdown, except through guilt by association.) So where is the moral hazard story? It falls apart when you look at the incentives and outcomes for actual individuals.

To repeat, the real story behind the crisis (among many other factors, but not including moral hazard) was fake alpha, which isn't that different from a plain old-fashioned swindle:

How then can untalented investment managers justify their pay? Unfortunately, all too often it is by creating fake alpha – appearing to create excess returns but in fact taking on hidden tail risks, which produce a steady positive return most of the time as compensation for a rare, very negative, return. [The guys who bundled and sold CDOs in, say, 2005, got paid by their firms, and "sophisticated" counterparties at other banks, or at pension funds, bought the securities ... Mr. Market failed to detect the fake alpha. The idea with fake alpha is to be rich and retired, or in another job, when the sh*t hits the fan. It's not that someone is going to come and bail you out. There is no moral hazard.]

But many economists don't like the idea of fake alpha because it exposes a fundamental and scary kind of market inefficiency -- if markets were really good at pricing long term risks and complex uncertainties, then compensation committees and CEOs and shareholders would be able to detect the timebombs in fake alpha. The record shows they cannot.

Ideological capture (efficient markets!) of academic economists and political theorists, combined with political-economic capture of government, in the form of campaign finance, allowed lax regulation. And lax regulation allowed fake alpha schemes to create systemic risk. I should also mention, as always, bounded cognition: not even the CEOs really understood what was going on in their own firms, let alone directors and shareholders. (People are stupid -- look around!)

To be fair to Levine, moral hazard will be a huge problem going forward -- the remaining banks NOW know with some certainty that they are too big to fail. However, I'm unconvinced that moral hazard was one of the leading factors in the last crisis.

Perfect Rigor in NY Review of Books

Earlier I recommended Masha Gessen's book Perfect Rigor: A Genius and the Mathematical Breakthrough of the Century. In case you haven't read it, there is a nice review and summary at the NY Review of Books.

... Up until March of this year, there remained one more chapter to the Perelman saga. Would he accept the one-million-dollar prize promised by the Clay Mathematics Institute for solving one of the seven so-called Millennium Problems? While the rules say that a proof must appear in a peer-reviewed mathematics journal (not just in an Internet posting), the mathematicians mentioned above have published papers in such journals expounding and amplifying the proof. Surely Perelman deserves the prize, which he was finally and officially offered on March 18.

Five days later, on March 23, Perelman rejected the Clay prize. He reportedly said through the closed door to his spartan apartment, "I have all I want." The comments he made after rejecting the Fields Medal probably reflect his present state of mind as well:

I don't want to be on display like an animal in a zoo. I'm not a hero of mathematics. I'm not even that successful. That is why I don't want to have everybody looking at me.

Some might argue that monetary awards for mathematical work are inappropriate, or that the Poincaré Conjecture is of little practical value and not worth the one-million-dollar prize. The aesthetic and epistemic value of the proof is priceless, however, and it may eventually yield more earthly consequences as well. As for the size of the award—how many no-name hacks are there on Wall Street who make a million dollars or more not just once but every year, and contribute exactly what? Whether Perelman has practical need for the money or not, he could use it to help support his mother or mathematicians of his liking, or to advance the kind of education conceived by Andrei Kolmogorov, or for some purpose only he could imagine. Reconsider your decision, Grisha.

The horror of war

Pity the victims and their families. But also consider how brutalized are the soldiers who participate in this. The arriving ground troops look like an army of insects or small robots.

More information here.

... Upon hearing that one of the victims is a young girl, the pilots laugh, “Well, it’s their fault for bringing their kids to a battle”. Wrong.

The pilots fail to mention these two men walking into the building, nor do they mention another unarmed man (34:40) walking directly in front of the building as they shoot a Hellfire missile. Again, read FM 3-24 (Counterinsurgency Manual), Appendix F. Another obvious Counterinsurgency failure.

Free Will in Waking Life

I discovered this video while looking for some course material on determinism and free will. I haven't seen Waking Life, but it looks interesting.

Data Mining the University

Here is a draft of my paper with Jim Schombert on University of Oregon GPA and SAT statistics. I posted previously on this research: Cognitive thresholds , The value of hard work. Introductory slides on g, SAT and all that.

Much of our data is available in the plots here.

Why did I get interested in this stuff? Obviously, I have a long-standing interest in psychometrics. In teaching 100 level courses, both Schombert and I have been flummoxed at the large population of students who have trouble with what we would consider elementary concepts (e.g., "scaling", or even "area" or "volume"!), yet seem to be successful in their chosen major ("I just can't seem to do these problems, but I need this class to graduate (fulfill a science requirement). I always get A's in English/History/Sociology/ ..."!) I'm sure every physics professor hears these things. Because I invest a lot of time in helping students, e.g., solving lots of problems during office hours, I have a pretty close view of their learning abilities. I began to wonder how students could have trouble with these basic concepts ("Didn't you have to know that for the SAT?"), yet have high GPAs in their major. In talking to Jim, who is director of the General Science program, we realized the data was actually available to investigate these questions further.

Data Mining the University: College GPA Predictions from SAT Scores


From the Conclusions:

1. SATs predict upper GPA with correlations in the 0.35 -- 0.50
range.

2. Overachievers exist in most majors, with low SAT scores but
very high GPAs. These overachievers are disproportionately female.

3. Underachievers exist in all majors, with high SAT scores but
very low GPAs. These underachievers are disproportionately male.

4. Some majors, like math and physics, may exhibit a cognitive threshold -- mastery of the material is unlikely below an ability threshold (as measured by SAT-M), no matter how hard the student works.

5. Students at public universities, like UO, with high upper GPA (e.g., 3.7 or greater) likely have subject mastery similar to graduates of elite universities. Elite college students who transferred to a state university would likely average upper division GPAs of 3.7 or greater.

Figure 8 caption: Underachievers and Overachievers (red = females, blue = males) isolated by SAT and upper GPA (1.25 standard deviations from the green ridgeline). The overachievers are mostly female students (64%) and the underachievers are mostly male students (79%). (Click for larger version.)

April Fools

Today at the preschool I bumped into an old friend who has her son in the same class as my kids. She's a professor in a social science department here, and she couldn't wait to tell me about the prank she had pulled on her colleagues at their weekly seminar. She was in charge of refreshments, and decided to offer -- get this -- human milk cheese and human milk with coffee and tea! The milk was supposed to be hers (she is breast feeding a new baby) and she used Adobe Illustrator to create a label for the cheese. (I asked her to send me a copy of the label, which I'll post here. My suggestion would have been: "People's Dairy -- real cheese from real people's milk!")

My friend is solidly on the left, but she laughed as she pointed out that none of her colleagues dared speak out against the strange refreshments. It would have been politically incorrect to criticize someone else's dietary choices :-) They looked uncomfortable, but they ate the yummy "human cheese" and used the "human milk" in their coffee and tea. The People's Republic of Eugene!

What would you have done?

Update from the prankster:

... the one thing to correct is that most people didn't eat it, and most seemed pretty appalled... but they largely kept it to themselves. But when I finally said April Fools (about 30 minutes into it), there was palpable relief in the room. (Only 2-3 grads tried the cheese).

The price of eggs

Eggs purchased from female donors at elite universities are several times more expensive.



From Aaron D. Levine, "Self-Regulation, Compensation, and the Ethical Recruitment of Oocyte Donors," Hastings Center Report 40, no 2 (2010): 25-36. More.

Crete talk: black hole information and decoherence

I finally posted my contributions to the proceedings of the meeting I attended last summer in Crete (see here for talk slides). I waited until now because the contributions were refereed and only recently went to the publisher.

A reaction I sometimes get to this work is that: "Maldacena showed in AdS that measuring a certain correlator can determine whether black hole evolution is unitary (i.e., whether information is destroyed)". That is correct, but the measurement precision needs to be of order exp(-S), where S is the entropy of the hole. It turns out that is the same level of precision one needs to detect decohered Everett branches of a system with S degrees of freedom.

http://arxiv.org/abs/1003.5382

The black hole information paradox and macroscopic superpositions

We investigate the experimental capabilities required to test whether black holes destroy information. We show that an experiment capable of illuminating the information puzzle must necessarily be able to detect or manipulate macroscopic superpositions (i.e., Everett branches). Hence, it could also address the fundamental question of decoherence versus wavefunction collapse.

Comments: 10 pages, 4 figures, contribution to proceedings of the 1st Mediterranean Conference on Classical and Quantum Gravity.

SES and IQ

It is sometimes claimed that IQ is just a proxy for SES (Socioeconomic Status): high IQ kids are merely the beneficiaries of better home environments, and correlations between IQ and life outcomes are merely a proxy for correlations between childhood environment and outcomes. Of course, this claim does not address the significant variations in IQ within families. Does IQ have predictive power once SES is controlled for? The answer is obvious from anecdotal experience: we all know siblings who, by definition, shared the same family SES, but with different IQs and life outcomes.

Vanderbilt psychologist David Lubinski sent me a recent paper addressing this issue -- Cognitive Epidemiology: With emphasis on untangling cognitive ability and socioeconomic status, Intelligence 37 (2009) 625-633. Among the figures is the following one, based on NLSY longitudinal data (the analysis was originally done by Charles Murray). It compares life outcomes of paired siblings (one of each pair required to be in the "normal" reference range of IQ), thereby controlling for SES. As you can see, IQ has a strong impact on education, earnings and social status even after family SES is controlled for. (Click for larger version of figure.)



Similar analysis shows that the correlation between SAT scores and college grades only decreases slightly when SES is controlled for. See also related post on non-shared environmental effects.

Below is a useful SES-SAT (or SES-IQ) syllogism, well supported by the studies at the two links above.

SES does not cause SAT (weakly at most). ***
SES does not predict college success, SAT does.

*** Someone recently asked me about the non-genetic or environmental variation in adult IQ. If we use a heritability range of .7-.9 (as deduced in twin studies; applicable for IQ in late adulthood and over a large range of childhood environments), then the resulting distribution of IQ differences for monozygotic twins would have SD = 5-8 points. That is, clones raised in different families would typically have IQ differences of around 5-8 points. Note, though, that very little of this difference is accounted for by SES -- most of it is due to non-shared environmental effects. For a lower heritability value such as .5 (perhaps more appropriate for adolescents), the expected variation between clones would be about 10 points.

Photos and long term memory

Why do I take photos of my kids or while traveling? The answer is so obvious it's a cliche: for the memories.

Remarkably, long term human memory capacity is probably only on the order of a gigabyte. Ask yourself what specific things you remember from the year 2000. How about your first day of class in college?

While the brain processes huge amounts of data (dominated by visual input), the amount of information retained is quite small -- no more than a few bits per second. When I ask you what your date's prom dress looked like, you are probably reconstructing it from conceptual primitives like "blue", "prom dress", or "velvet" :-) Do you think you could picture all of the girls you ever dated if you hadn't looked at their photographs over the years?

Looking at old photographs allows me to recapture and enhance otherwise elusive memories. Also, there's some selection bias in what one chooses to photograph: the memories triggered tend to be happy ones. Viva Espana! :-)

Human Memory: ... This makes the work of Thomas K. Landauer very interesting, for he has entirely avoided this hardware guessing game by measuring the actual functional capacity of human memory directly (See "How Much Do People Remember? Some Estimates of the Quantity of Learned Information in Long-term Memory", in Cognitive Science 10, 477-493, 1986).

Landauer works at Bell Communications Research--closely affiliated with Bell Labs where the modern study of information theory was begun by C. E. Shannon to analyze the information carrying capacity of telephone lines (a subject of great interest to a telephone company). Landauer naturally used these tools by viewing human memory as a novel "telephone line" that carries information from the past to the future. The capacity of this "telephone line" can be determined by measuring the information that goes in and the information that comes out, and then applying the great power of modern information theory.

Landauer reviewed and quantitatively analyzed experiments by himself and others in which people were asked to read text, look at pictures, and hear words, short passages of music, sentences, and nonsense syllables. After delays ranging from minutes to days the subjects were tested to determine how much they had retained. The tests were quite sensitive--they did not merely ask "What do you remember?" but often used true/false or multiple choice questions, in which even a vague memory of the material would allow selection of the correct choice. Often, the differential abilities of a group that had been exposed to the material and another group that had not been exposed to the material were used. The difference in the scores between the two groups was used to estimate the amount actually remembered (to control for the number of correct answers an intelligent human could guess without ever having seen the material). Because experiments by many different experimenters were summarized and analyzed, the results of the analysis are fairly robust; they are insensitive to fine details or specific conditions of one or another experiment. Finally, the amount remembered was divided by the time allotted to memorization to determine the number of bits remembered per second.

The remarkable result of this work was that human beings remembered very nearly two bits per second under all the experimental conditions. Visual, verbal, musical, or whatever--two bits per second. Continued over a lifetime, this rate of memorization would produce somewhat over 10^9 bits, or a few hundred megabytes.

Hangzhou photos

Below are some photos from Hangzhou (click for larger versions). Earlier in the week we had more meetings at Alibaba concerning the establishment of a social science data observatory sponsored by Alibaba Cloud Computing, making use of data from Taobao, Alipay, Alibaba.com, etc. Related post.

See here for an Economist special report on the Big Data deluge. Imagine you could observe 100+ million users transacting on a platform with 400 million item listings -- Taobao volume equals about 2 percent of all Chinese retail sales, and growing fast :-)

Yesterday we gave a presentation to about 200 technical staff who were attending a company-wide retreat at a resort outside Hangzhou.

Today I have to turn my physics brain back on as I am giving a talk at Zhejiang University (Zheda). Until now I have been simulating a social / computational scientist :-)

Two book recommendations

Very high recommendations for the following books, which I stockpiled in advance of my trip to China, but ended up reading most of before departure...

The Lost Books of the Odyssey by Zachary Mason, a Borges-ian retelling of the Odyssey in 44 short stories.

Perfect Rigor by Masha Gessen, on Grisha Perelman and the proof of the Poincare / Geometrization Conjecture. I already knew a bit about the completion of Hamilton's program thanks to colleagues who work on Ricci flow. What I found most interesting in the book is the insider description of the Soviet math system, including the nurturing of young talents and the training for competitions like the IMO.

Cognitive thresholds

[See here for updated results.]



This graph (click for larger version) shows the upper division in-major GPAs of U Oregon physics and math graduates from a 5 year period. The math GPAs were computed using a specific set of rigorous courses taken by graduate school bound majors. One might reasonably associate "mastery" of a subject with GPA > 3.5 -- roughly, the minimum threshold to be admitted to graduate school (i.e., students who earn equal numbers of A's and B's in their upper division core courses are borderline for most graduate programs). We find that no student with SAT-M less than roughly 600 was able to attain this level of mastery.

This data provides at least modest evidence for a cognitive threshold required for physics and mathematics. That is, the a priori probability that a student with SAT-M < 600 will perform well enough to be admitted to graduate school in these subjects is extremely low. Note many poorly performing students switch majors, and hence do not populate the lower left corner of our graph.

A simple but plausible model for college performance might include two factors: 1. ability (measured by SAT) and 2. conscientiousness or effort (for simplicity, an uncorrelated random variable, probably normally distributed). The predicted GPA might depend linearly on the sum of the two factors, each measured in SD units. (See bottom figure here for a similar predictor of GPA.) Applying such a model to our data, we would conclude that even a student in the, e.g., 90th percentile of work ethic has a low probability of attaining mastery if their SAT-M score is below 600.

To reiterate, SAT-M 600 seems to be the lowest score at which even a very motivated student has a chance for mastery. From the data one might guess that only for SAT-M well above 700 do students have more than a 50 percent chance of obtaining GPA > 3.5. That is, a student with average motivation or conscientiousness probably needs SAT-M well above 700 to have a high probability of obtaining mastery.

We were unable to find any similar threshold (either in SAT-V or SAT-M) in other majors, including economics, sociology, history, philosophy, biology, chemistry, etc. (More about this data set, including data on other majors, here.)

The percentile equivalents of SAT-M scores for various groups can be found here. For the total SAT-taking population, 600 is about 75th percentile. For the overall population, it might be 85th percentile. SAT-M of 750 is roughly 98th percentile for the total SAT-taking population.

For comparison, here is the analogous plot for English (black) and History (blue) majors versus SAT-R (verbal). If a threshold exists it is probably at SAT-R of 450 or so.

Great Firewall and blogging

I will be behind a certain Great Firewall next week, so blogging might be interrupted :-(

I have taken some precautions, we'll see how it goes.

The perils of precocity

Last night at dinner my 4 year old daughter asked me about death.

"Do all princesses grow old?" she asked.

"But I don't want to die!" she said, knowing, it seems, full well what that meant.

"We might go on to something else after we die," I offered.

"Oh." She thought about that for a while.

My son doesn't seem concerned about any of this, although he does occasionally ask me what the biggest number is. (He also wants to know about the biggest dinosaur and the fastest rocket.) I usually ask him to name the largest number he can think of, and then point out that I can always add one to that to get an even larger number.

When I was a kid I spent a lot of nights pondering life and death and infinity. I still do.

James Cameron's machines

A nice essay on James Cameron and Avatar in the NY Review of Books.

The Wizard: ... The fascination with the seeming invincibility of sophisticated mechanical objects, and an accompanying desire to slough off human flesh for metal (and a celebration of flesh so taut it may as well be metal: Cameron's camera loves to linger on the tightly muscled bodies, male and female, of the soldiers so often featured in his violent films), is a recurrent theme in the techno-blockbusters that cemented the director's reputation in the late 1980s and throughout the 1990s. Aliens famously ends with Weaver's character, Ellen Ripley, battling the dragonish alien monster queen after strapping herself into a giant forklift-like machine whose enormous pincers she mechanically controls by maneuvering her own slender arms—a technology that puts the puny human, finally, on a par with her gigantic, razor-toothed, acid-bleeding adversary.

... A violent variation on the same mechanical bodysuits reappears, memorably, in Avatar, which culminates in a scene of bloody single combat between a Na'vi warrior and the evil Marine colonel, who has strapped himself into one such machine. If anything, the recurrent motif of humans inserting themselves into mechanical contraptions in order to enjoy superhuman powers reaches its fullest, most sophisticated expression in the new movie, whose characters can literally become other, superhuman beings by hooking themselves up to elaborate machines. All this seems to bear out the underlying truth of a joke that Linda Hamilton, the actress who played Sarah Connor in the Terminator movies, told about her first, unhappy interactions with the director (whom she later married and divorced): "That man is definitely on the side of the machines."