Peter Levin’s Rethinking Markets

April 24, 2009

more aggrevaluation

Filed under: Culture, Markets, Technology — Peter @ 9:18 am

[this is a comment on Brayden's post, which got long enough that I'm reposting the comment here. Go read that thread though, it's quite interesting.]

Ha, I step out for an evening and realize I’m being called out by name, even praised, after I was so unkind to the orgtheory. Brayden you’re trying to make me reconsider…

A few things. First, I’m lumping together into ‘aggregation’ and ‘wisdom of crowds’ a number of different types of activities. Recommendation engines and the Hollywood Stock Exchange (where I’m currently ranked 47943th, with a lifetime ROI of +1,164.95%) are very different, and rather than go with the it’s complicated routine, I lumped a bunch together. I’d go a step further than Lena and say it’s a categorical error to suggest either expert opinion or crowd-sourced outcomes are generated with same logics. I’d like to hear more about the relative value of expert-vs-the crowd across these differences.

Second, design-wise, I might be wrong about exploitation v. exploration. I had in mind that if you put a movie on HSX and try to value it, you will be way off on movies that don’t already conform to existing kinds of movies. Or, if you base your decisions on what kinds of things people like/want, you end up with Playstation 3 and Xbox 360, and miss the Wii – because people didn’t ‘want’ it before it existed. Ugh, I’m getting muddy.

But I might be wrong – I mean, Paul DePodesta (of Moneyball, soon a movie with Brad Pitt and ironically undervalued at $31.75 on HSX) explicitly noted that he used data and no assumptions about existing scouting experts to come up with a new way to assess player value. Clearly exploration through data mining.

(and a propos Mike’s comment above, why 538 and Netflix does well, it seems as though it is something about the difference between regression analysis and Singular Value Decomposition/factor analysis; but that requires more explanation)

But my problems are theoretical, and here I enlist Hahn and Tetlock’s definitive Information Markets: A New Way of Making Decisions. Which for a theoretical basis starts with (on p2) “Why do information markets work as well as they do?” And then references….The Wisdom of Crowds. And then moves on to design. And Surowiecki’s theoretical answer?

“At heart, the answer rests on a mathematical truism. If you ask a large enough group of diverse, independent people to make a prediction or estimate a probability, and then average those estimates, the errors each of them makes in coming up with an answer will cancel themselves out. Each person’s guess, you might say, has two components: information and error. Subtract the error, and you’re left with information…With most things the average is mediocrity. With decision making, it’s often excellence. You could say it’s as if we’ve been programmed to be collectively smart” (p10-11).

Not really obvious at all. There is no answer why a market metaphor would result in something better than experts. But we have the technology to do it, and it seems experimentally to work, and in web 2.0 we can get users to do this all for free! And so screw it – off with the design team and on with the user testing.

Comments (0)

April 22, 2009

Speaking of reviews and recommendations

Filed under: Culture, Technology — Peter @ 1:18 pm

In another frame of mind, I would say that some crowd-sourced reviews do make me happy. ‘Skip S’ reviews Animal Collective’s Merriweather Post Pavilion on iTunes:

Okay. Imagine sitting back in the day with none other than the Buddha himself. The two of you are discussing life, love, and various other zen things. Then, all of a sudden a man approaches you both offering some fresh Indian rice. When the rice is finished cooking, you realize, “Oh my lord. I forgot my soy-sauce back in the future!” The Buddha tells you to calm down and he smiles and reaches into his woven napsack and pulls out his own personal bottle. You graciously accept his gift, and soon enough your stomach is full of a delicious medly of flavors and such. You lean back, completely content. A wide smile stretches across your face.

That’s what this album is like.

79 out of 95 listeners found this review helpful.

Comments (0)

Aggregation aggravation

Filed under: Culture, Markets, Technology — Peter @ 9:56 am

It seems to me that one of the fundamental advances and problems with web 2.0 is that it poses expertise against aggregation. The ‘old’ system (and here I would say that these are overlapping, not coterminous ways of doing things) is one of expert reviews, or critics. You want to know what movie to see, so you ask Roger Ebert (though his recent review and ongoing defense of Nicolas Cage’s Knowing strikes me as bizarre). If you want to know what music to listen to, you turn to Sasha Frere-Jones. For consumer goods, Consumer’s Guide. For electronics, David Pogue. And so on.

The point is fractal, incidentally. In this ‘old’ system, for policy advice you would call on sociological experts (naturally, though maybe other lesser social scientific experts if you’re interested in worse advice). In organizations, you would look for marketing advice from your marketing division, operations from operations, finance from finance. Obviously the more general the point I make the more fault you can find with it. And you would be right. But bear with me for a moment.

The ‘new’ system rests on a Wisdom of Crowds knowledge. That is, if you take a bunch of people and ask them their opinions, you can get a better fix on uncertain knowledge than you can with a small number of experts. Now, Surowiecki himself is not this simple: at minimum one must overcome problems of cognition, coordination, and cooperation. But this said, proponents of this kind of system point to rather stark indicators of success. Google’s PageRank (though I find the idea that they use 500 million variables and 2 billion terms absurd); Yelp; the Iowa Electronic Markets; Metacritic. And in the more general point, we see a substitution of ‘market’/crowdsourcing/datamining as a substitution for design, marketing, strategy. Here I mean the A/B testing ad absurdium as a substitute for design. Data-mining as a substitute for marketing. Quantitative finance as a substitute for market forecasting.

This whole edifice actually rests on a kind of efficient markets hypothesis, or more specifically a Friedrich Hayek-type consolidation of ‘adverse’ knowledge (meaning, in this context, private knowledge) via a market mechanism. While Hayek wanted to argue that market-based societies are better than centrally-planned societies, his work has become the intellectual touchstone of all things information market. And really that’s what it comes down to. Crowd-sourcing: a replacement of expertise with market.

However, there are some things to think about here that make this ‘new’ system quite problematic. And I ain’t sayin’ so just because I’m an expert (after all, the policy people really don’t come talking to sociologists, despite my preferences). There are one specific and one theoretical.

The first specific is that some people are just crazy, and aside from creating a tail-end of a distribution curve, it’s not at all clear what these folks contribute to the crowd. Old but still hilarious is Andy Baio’s Amazon Knee-jerk Contrarian Game. Personally, I like the ratings game at Yelp, an often-loved but massively crowd-sourced guide. Take, for instance, the Museum of Modern Art in NYC (i.e., the, or one of the, best modern art museum in the US and the world):

Why 1 star? Its just a horrible place to visit never ever again, screw this contemporary art thing, the exhibits they had going on were……… yeah no way to describe the sheer disappointment in the place. The place is designed to shock and awe you, all it did was bore me.
and
Most of the exhibits at MoMA are just random objects or B.S. paintings–hardly classifiable as art.
I could just go down to my garage or get a toddler to paint on a canvas to receive the MoMA experience. No crowds or superinflated entrance fees there, either.
and
I was so jazzed to go there. Many people I know raved about it.
All I came away with from this place was one word: Overrated.
Quality Modern Art is subjective. In my mind, for the hype this place gets is unwarranted. So sad…

So how do these reviews contribute to overall ratings systems? More broadly, what if the feedback/view/idea/opinion from your customers is just wrong? In 2.0 way of thinking about things, this is like saying that a market price is incorrect – it is axiomatically impossible, barring something wrong with the system (an information problem being the first culprit). And there is no ‘expert’ to say otherwise.

More theoretically, it has never really be adequately explained why a ‘market-like’ information crowd-sourcing should work. I understand why markets might produce a price that incorporates most public and private information about a commodity. But the widespread substitution of expertise with data mining and crowd-sourcing is a market metaphor more than a market. Why should a metaphor work? This is at the heart of someone like Daniel Davies’ criticism. And I get that sometimes aggregation does work. But there’s no good reason why.

My own feeling is that, using March’s metaphor of ‘exploitation’ and ‘exploration’ (where the first is the plumbing of existing knowledge/arenas, and the second is the seeking out of new opportunities), aggregation mechanisms are better at exploitation than exploration. They do better with existing standards of knowledge, of tastes, of commodities, than they do with something that is new. You know, Blue Ocean and such. I think there are better solutions for a 2.0 world that combine expertise and aggregation (for instance, Five Thirty-Eight’s work on the 2008 elections that combined data mongering with theoretically-driven and field-visit-driven analysis). But this post is already too long.

Comments (2)

December 13, 2008

Abstract finance: Securitization

Filed under: Abstract Finance, Technology — Peter @ 11:40 am

One basic idea to help understand contemporary finance is securitization. To explain what securitization and how it works, first think about the following: what happens when you and your best friend decide to open a business together. How are you going to divvy up responsibilities, management decisions, profits, and losses?

One way is to just decide, you do this part, I’ll do that part, and we’ll split the proceeds. This is very local ownership. It is tied directly to you and your friend’s relationship, the specifics of the business venture, the particulars of the agreement you forge. For the current purposes, I want to call this concrete finance.

An alternative way to divvy up your company is to create shares of one sort or another.

When you create shares, you can assign ownership rights to them, so that each share might represent an equal percentage of end-of-year profits. If I put in half of the money, but my friend is doing most of the work, maybe he would get more shares than me. Maybe not. But to one degree or another, these shares represent ownership. Shares can be a legal agreement as well as an informal one.

What you’ve done formally is to create an instrument to stand in as financial value. It’s a kind of abstraction. The ownership no longer depends so specifically on the relationship between you and your friend. The rights to ownership, and maybe future profits and losses, are now formally invested in the shares, not just in your relationship. When you agree to partition the company into shares, you are turning the rights to future management and earnings into these more abstracted instruments.

What makes them ‘abstract’? Well, they become abstract in two distinct but related ways. First, the shares themselves are worth something. So ownership of the shares are ‘worth’ whatever profits come out of your business at the end of the year, plus they are worth whatever someone will pay to get those profits. If your business is great, with bright prospects, someone might well be willing to pay a premium for your shares. Or expect a discount if your business’ prospects are lousy. This means that there are actually two values to shares, a representational value and a market value: the representational value is how much value the shares stand in for (i.e., your share of the profits), and the market value is how much someone would buy or sell those shares for.

The second way they become abstract is that the value the shares can be thought about in ways that are not specific to the business that you started. For you, the shares are a shorthand for the blood, sweat, tears, and rewards you get for taking chances on and working in your business. For someone else, the shares can be represented as cash returns over time. And these cash returns over time are compare-able to other cash returns over time from other kinds of things. For instance, owning a totally different kind of business, buying and selling rare coins, or lending money to arms dealers all have ‘cash returns over time’ that can be compared favorably or unfavorably to your shares. In this way, what was once only understandable as a local, concrete financial arrangement between you and your best friend gets brought into the wider universe of other financial stuff. This is what I want to call abstract finance.

If you’ve got this imagery (a concrete set of relations that is transformed into abstracted commodity), you’ve got the basis for all kinds of financial instruments. Just about any ‘stream’ of financial value can be split into shares. A company’s future earnings, a corporate loan, leases on heavy equipment, credit card debt, US tax revenue, mortgages, annual cell phone plan subscriptions, all of these things are concrete streams of financial value. They can be split up and sold as shares. These shares are secured by their underlying pools of value. When they are securitized, they become measured not in their own terms but in more formally abstract terms: rate of return, risk of default, time risks of pre-payments, costs relative to other kinds of investment.

Comments (0)

September 9, 2008

Human eror or cmoputer error?

Filed under: Markets, Technology — Peter @ 8:02 pm

Another story today about the relationship between technology and human discretion. Apparently, Google picked up an old story that was undated on the internet (really from 2002), and re-posted it as a story from today: that United Airlines was headed for immanent bankruptcy. Wanna see what happens when people suddenly think that you are a company going bankrupt?

NASDAQ’s response was to tell investors, tough cookies:

Once trading resumed 90 minutes later, UAL shares rebounded, but they still closed off 11% for the day at $10.92. Nasdaq, a unit of Nasdaq OMX Group Inc., said that it had reviewed transactions involving UAL shares during a 13-minute period before the halt and that all trades will stand. A trade for 100 shares at a penny apiece occurred on another exchange after the trading halt on Nasdaq and was later cancelled.

The reason for NASDAQ’s response is worth noting, even if it is not noted. Exchanges have a deep stake in defending the view that there are market errors, and then there are human errors. And a market error – an error that implicates the platform and mode of trading itself in the systemic mucking up of financial transactions – is deeply problematic for exchanges. Attributing the error to human errors – that buyers/sellers correctly placed orders which were correctly matched – allows exchanges to maintain their self-image as market infrastructure.

I’m sure Google and Income Securities Advisors Inc. will have more to say about the incident.

Comments (1)

September 5, 2008

Effects of Markets 2.0

Filed under: Markets, Technology — Peter @ 3:05 pm

[Warning: this post is a bit long. If you're not interested, you can try instead the music stylings of Gary Numan.]

In markets 2.0, I refer to the data that is generated as part of normal market transactions. The 2.0 references web 2.0, where the ’social data’ generated by web interactions and transactions has acquired a kind of life of its own. I make no super-special claims that markets now are wholly different from markets in the past. But some of their properties have become increasingly visible and important.

Let’s take the basic case. In a financial market, if Gordon Gecko and I make a trade, a couple things happen. First, money goes in one direction, and a commodity or security goes in the other direction.

I’m playing a bit fast and loose here, but you get the idea. Let’s call this markets 1.0. Exchanges like the occur all the time in lots of different places. And if you wanted to generate outwards, I interact with all kinds of people in ‘markets 1.0′ mode – iTunes when I listen to a song, colleagues when I ask about a job, my bank when I enter into a transaction. I have in mind a particular case (capital markets) of what I think is a more general phenomenon.
(more…)

Comments (3)

May 14, 2008

A sociological analysis of the current market crisis

Filed under: Markets, Technology — Peter @ 2:54 pm

For my talk at the UCSD Culture conference, I spoke about market crises, commensuration, and market linkages. The slides are a .pdf of my keynote presentation, available here (the keynote presentation for those who can manage it, is a zip file available here). And this post goes along generally though not perfectly with the slides: (more…)

Comments (1)

April 30, 2008

They know the score

Filed under: Technology — Peter @ 3:13 pm

Two cases, separated by 150 years, about technology, missionaries, and institutional change:

The first:

In 1818 the directors of the London Missionary Society sent a mechanical clock to grace the church at its first station among the Tswana in South Africa. No ordinary clock – its hours were struck by strutting British soldiers carved of wood – it became the measure of a historical process in the making. Clearly meant to proclaim the value of time in Christian, civilized communities, the contraption had an altogether unexpected impact. For the Africans insisted that the “carved ones” were emissaries of a distant king who, with missionary connivance, would place them in a “house of bondage.” A disconsolate evangelist had eventually to “take down the fairy-looking strangers, and cut a piece off their painted bodies, to convince the affrighted natives that the objects of their alarm were only bits of coloured wood” (Moffat 1842: 339). The churchman knew, however, that the timepiece had made visible a fundamental truth. The Tswana had not been reassured by his gesture; indeed, they seem to have concluded that “the motives of the missionary were anything but disinterested.” And they were correct, of course. In the face of the clock they had caught their first glimpse of a future time, a time when their colonized world would march to quite different rhythms.

- Jean and John Comaroff, Of Revelation and Revolution, Volume 1, p. xi

And the second:

The first step toward implementing the system was a pilot study, involving one specialist and two or three stocks, to determine whether a specialist could physically monitor an automatically executing trading system while simultaneously performing other activities. Because of the narrow physical dimensions of specialists’ posts, Loss enlisted the Exchange’s carpenters to mount a wooden platform on a pedestal-pipe two feet above the pilot specialist’s post, where the ATS computer screen would stand and display its prices for all to see. A brochure explaining the new system was distributed to member brokers. The Friday before the pilot study was to begin, Loss left his office for what he described as his first relaxing weekend in months. The following Monday morning, all was in readiness for the start of the pilot program. Except…

As soon as Loss walked into his twenty-third-floor office he received a phone call from the Exchange’s floor manager. “You’ve gotta come down and see this,” said the manager.

“What am I going to see?” Loss remembers asking.

“Just come down.”

Minutes later, Loss saw for himself: In a pile of sawdust on the floor near the designated specialist’s post lay his Automated Trading System equipment. Someone had used an ax or a saw to cut a semicircle through the wooden platform and disconnect it from the post.

The jagged teeth marks on the wood, Loss recalled years later, “Looked like Jaws had just been through.”

- Marshall Blume, Jeremy Siegel and Dan Rottenberg, Revolution on Wall Street, pp. 195-196

Comments (0)

February 7, 2008

What is XBRL, and Who does XBRL help?

Filed under: Data, Institutional, Markets, Technology — Peter @ 10:02 am

Put it on your radar screens, the next big thing is going to be XBRL. It stands for extensible business reporting language, and it is meant to commensurate business reporting via standardization. So instead of entering text into an annual report, companies, governments, NGOs, anyone who would like to comply with governmental mandate will be using XBRL. You can think of XBRL as a set of metatags for financial and company data, so that instead of bracket-tags for header, title, links, etc. you would have bracket-tags for earnings, time periods, definitions of costs, etc.

From CoreFiling’s insight blog: “It won’t be very long before it is those documents – the bar-coded financial disclosures – that will be the primary materials consumed by financial market systems to help analysts and investors make decisions about the best way to invest. This is vastly more sophisticated than today’s processes that rely on slow and inaccurate re-keying of a subset of the financial information published by companies.”

This is commensuration more than just standardization, since the tags are designed to be specific to a particular business enough so that everyone is not required to give the same information, yet the tags are standardized enough that everyone is required to give information that can be made comparable. The pitch for companies (other than, because otherwise we’ll fine you and take away your business license) is that XBRL will make their financial reporting less costly, less prone to error, and ultimately more efficient.

Personally, I think this is a flat out misrepresentation of what’s going on here. XBRL helps one group of people orders of magnitude more than anyone else: investors. And the trade-off between increased government efficiency and business streamlining of compliance data on the one hand, and increased ability for data-gatherers for banks, hedge funds, and the investor class is totally totally off the charts. What this will end up doing is: 1) creating a standard way for companies to report financials; 2) creating some increased efficiency for government entities to keep tabs on the finances of these organizations; and 3) create a massive additional datastream for financial services and investment firms to work with. If you think it is a challenge for public firms to resist making short-term decisions based on financial analysts’ quarterly reports of earnings now, wait until this information is directly readable by quant trading models.

This would be an amazing dissertation topic. I would track: a) the creation of the standard; b) the adoption of the standard around the world; c) how XBRL is being incorporated into financial modeling; d) the before-and-after effects of XBRL on market prices for firms; and e) qualitatively, what gets excised from XBRL, or rather, what remains incommensurable about firms, governments, etc.

UBMatrix
XBRL’s main site
US SEC’s ‘Interactive Data Viewers’
Microsoft uses XBRL
US GAAP XBRL Taxonomy (GAAP is the accounting standard in the US)
CoreFiling

Comments (0)

February 6, 2008

Black Swans, Risk Management, and Undersea Cables

Filed under: Institutional, Technology — Peter @ 11:16 pm

I’ve taken issue before with Nassim Nicholas Taleb’s black swan thesis, that high-impact, low-probability events are responsible for market crises and accidents. The more general implication is, as Taleb and Pilpel note:

What matters in life is the equation probability × consequence. This point might appear to be simple, but its consequences are not.

Suppose that you are deriving probabilities of future occurrences from the data, assuming that the past is representative of the future. An event can be an earthquake, a market crash, a spurt in inflation, hurricane damage in an area, a flood, crops destroyed by a disease, people affected in an epidemic, destruction caused by terrorism, etc. Note the following: the severity of the event, will be in almost all cases inversely proportional to its frequency: the ten-year flood will be more frequent than the 100 year flood – and the 100 year flood will be more devastating.

Now comes word that some number (actually up to 5 now) of undersea cables have been cut, knocking a wide area of the Middle East off the internet, particularly the route between Europe and Egypt, and from there to the rest of the Middle East.

But where is the 100 year flood? What appears to have happened is a connected series of accidents and snafus, including possibly the weather, an anchor dragging along the sea floor, or who knows what. Mysterious. What I would contend, drawing from org theory, is that what is more dangerous than a 100 year flood is a sequence of preventable, unforeseen errors. That is, it is the disruption of the routine more than a freakish activity that is most likely to create accidents and crises. The routine fire in a particularly bad location, a minor earthquake in an unexpected place, a sequence of coupled organizational routines that lead one-to-another into disaster. It’s not that you shouldn’t be looking for the next giant storm that’s inevitably coming down the pike, but more problematic are the breaks in the caulk around the tub that floods the electrical box, that shorts the grid. Or a failure in the bathrooms at the airport.

Read your Saul Alinsky, and get in the game.

Comments (5)

January 31, 2008

Privacy, data, and the new Sociometrics

Filed under: Ambiguity, Technology — Peter @ 11:53 am

As always, Technology Review provides a great glimpse at the innovations coming down the pike. In this case, by innovation, I mean the continued ascendancy of sociological insight wrapped up in physics, taken up by engineering, and brought forward as the ‘next big thing’: you can actually identify people’s social networking in real-time and help them to, for example, work a crowd better; or find a more suitable financial broker; or see which institutional representatives interact with whom at an academic conference.

All else aside, I think we are approaching a new era of privacy issues, related to data mining and what counts as anonymity, what counts as data, who owns it, and who profits from it. The sociometrics model only works when we have things like smart-badges, ambient microphones, and unobtrusive surveillance. Alex Pentland, one of the MIT researchers (rightly) sees quite benign benefits from these technologies, including making non-face-to-face interactions more effective and efficient.

Transcripted from Alex Pentland’s website (.wmv movie link):

We can really measure exactly when you nod your head, and exactly the inflection of your voice, and exactly where you look, and all those things that you sort of know are important, it’s the social language. You can have a microphone that aims at different people and has a little bit of processing in there. We don’t listen to anything that violates privacy, we’re just looking at features of language. And that’s not particularly worrying to people because the words are never recorded, the meaning is never recorded. It’s really just social signals. ‘You were being pretty pushy there’. Or ‘You weren’t really being very forward there’. And we can combine that with measurements of performance to ask, how is it that your social presence affects your performance? Your working with other groups? And if we can do that, the evidence from the literature is that we can improve the working of groups, the functioning of organizations, by a lot. Not just 1 or 2 percent, but 20 or 30 or 40 percent. And of course, those are the things that drive profitability, that drive performance.

And it is indeed true that they are not recording words, or meaning, just social signals. But in an era where phone companies are data mining all voice and data activity, and financial firms are looking at aggregated transactions in the search for suspicious transactions (.pdf link), I would suggest that ‘just social signals’ are no longer a free resource. At minimum, we need a new language to capture what we used to, but no longer mean by anonymous.

And I’m not against what Pentland et al are up to, not by a long shot. I just think there are some issues that come into play that we’ve only begun to think about practically, and that we’ve not at all come to terms with theoretically or in the underlying social scientific research.

Comments (0)

January 19, 2008

Airplanes and Accidents

Filed under: Technology — Peter @ 8:40 am

Well, this was bound to happen. I mention an article tells us there haven’t been enough data points for airline crash investigators, and a plane crashes. As usual, it was a mix of tech and happenstance – apparently, on-board computers sent a demand for more power to the engines, but they did not respond. It’s interesting how so much detail is reported the inhibition threshold may have been set too high, and the engine pressure ratio gauge had failed. Information without informing.

The photos are eye-opening.

Comments (0)

January 15, 2008

Data mining, airlines, precursors

Filed under: Technology — Peter @ 3:23 pm

Via the Washington Post comes an interesting article on data mining and the airline industry. Apparently, airplanes are not crashing enough for the airlines to be able to determine the sources and causes of accidents. That is, there is not enough variability in the outcomes (the last crash was August 2006) to do forensic analysis.

Instead, airlines are turning to ‘precursor’ anlayses, data mining a whole slew of events that have not led to accidents: unstabilized approaches, pitch rates at takeoff, pilot scheduling. The article suggests but does not detail the sheer number of variables and flights being analyzed, saying that Southwest has ‘mined data on more than 1 million flights’, but not really talking about what that means.

Organizationally, this is fascinating because so much more often we see organizations respond to events rather than trying to predict them. Or rather, as the vice chairperson of the NTSB put it, mining for precursors is like “‘reading tea leaves’ because it can require imagination to tie together incidents that don’t seem hazardous at first blush.” Arguably, it’s the imagination part that is so tricky in seeing what to make of precursors to mistakes and accidents. Even if you find them, often precursors only matter when they happen in conjunction (ie. in systems that are tightly coupled). So you can actually imagine a series of events that still would not result in a crash unless those events were temporally and organizationally tied together.

I would say that this is what we’re seeing now in the finance world, but it’s not. It’s worth another post, but there we’re seeing deliberate profit-seeking and many (though not nearly all or homogeneously) firms knowing that things could blow-up but not really caring.

h/t: Paul Kedrosky

Comments (2)

January 10, 2008

Organizational Complexity and the Checklist

Filed under: Technology — Peter @ 5:01 pm

The New Yorker has a great article on the effects of technology in emergency medical care, with findings that are worth drawing out more carefully. In particular, the article is about intensive care units, where extraordinary measures are taken to keep patients alive. The question the author asks is, what happens when increased organizational complexity leads to errors? And what do we do about it.

A familiar answer is specialization. And in this world, specialization works. Research findings suggest that putting an intensive care specialist on staff (in ICUs in Maryland at least) had the effect of reducing death rates in intensive care units by a third. But the more effective solution seems to be a rather mundane, analog technology: the check list.

The main proponent of checklists in ICU care is Peter Pronovost, and the article details a single arena of innovation, the IV line. The checklist here consists of: “(1) wash their hands with soap, (2) clean the patient’s skin with chlorhexidine antiseptic, (3) put sterile drapes over the entire patient, (4) wear a sterile mask, hat, gown, and gloves, and (5) put a sterile dressing over the catheter site once the line is in. Check, check, check, check, check.” This has two effects: 1) it helps with memory recall; and 2) it provides a minimum set of standards in a complex process.

The introduction of checklists in IV line procedures was pretty miraculous:

In December, 2006, the Keystone Initiative published its findings in a landmark article in The New England Journal of Medicine. Within the first three months of the project, the infection rate in Michigan’s I.C.U.s decreased by sixty-six per cent. The typical I.C.U.—including the ones at Sinai-Grace Hospital—cut its quarterly infection rate to zero. Michigan’s infection rates fell so low that its average I.C.U. outperformed ninety per cent of I.C.U.s nationwide. In the Keystone Initiative’s first eighteen months, the hospitals saved an estimated hundred and seventy-five million dollars in costs and more than fifteen hundred lives. The successes have been sustained for almost four years—all because of a stupid little checklist.

This begs the questions, why and in what circumstances is something like a checklist a useful organizational tool. What is it? Alex Pang thinks it’s about predictability, and the solidification of practices and standards in the form of a predictable document. I’m tempted to see this as standardization and to start to tease out when and where standardization works and doesn’t. My old friend commensuration seems not really to apply here.

Incidentally, the author Atul Gawande also wrote a great piece a year or so back on the Apgar score and its effect on childbirthing practices. Similar scene, but there the issue is only sort of a checklist – it was a quantification issue.

Comments (0)

November 27, 2007

The Blow-Up

Filed under: Prices, Technology — Peter @ 6:01 pm

There’s a lot to say about Bryant Urstadt’s article about quant traders in this past fall’s Technology Review (free reg required), called “The Blow-up.” Combined with Amir Khandani and Andrew Lo’s “What Happened to the Quants in August 2007″ (.pdf link), it gives a fuller picture of how quantitative finance is altering markets.

Clumping quantitative derivatives traders together is getting to be a little too gross a distinction, but there are, it seems, a small number of strategies being pursued by these kinds of traders. Two of the most important are: 1) pairs trading and 2) long-short equity trading. Pairs trading is where an historical relationship between two securities are found, tested, and analyzed; and then when the pairs get out of this relationship for no good reason, trades are executed expecting that this relationship reasserts itself. This differs in spirit, but not entirely in practice, from arbitrage trading, which is trading on the differences between two kinds of products which are ’supposed’ to be identical (a company’s stock and its bonds, for example, or a company’s stock in two trading environments).

Long/short equity trading is when a market-neutral position is taken, whereby long positions are made up of ‘losers’ (under-performing stocks), while short positions are made up of ‘winners’ (over-performing stocks). This strategy is based on the idea that outliers will eventually revert to the mean – effectively betting on consistent market overreaction. Given the behavioral finance work on how people tend to, you know, overreact to news, this makes sense. And because your position is effectively market-neutral – long positions offset short positions – it is possible (for broker/dealers) to highly leverage these positions.

These are two main strategies, and for the most part, they work. So, what happened in August? Here’s my reading of the two articles: First CDOs – the bundled up derivatives from housing loans that are spoken about when we talk about the ’sub-prime lending crisis’, went into the sink. Second, an outlying series of days in the week of August 6 caused a dramatic drop in the long-short strategy.

The effects of these two events (possibly, though it’s not clear if they were linked) was the following: a bunch of funds began selling equities to liquidate positions in order to meet margin calls for their CDO investments, which pushed pairs trading into abnormal positions. This caused more hardship, and all sorts of things happened – funds sold blue-chips to raise margin cash; and bought back short positions causing price relationships to swing further out of historical norms. Khandani and Lo speculate that one or more of the long-short funds liquidated its position, and that it turned out that many other funds were also engaged in these kinds of trades, so that when one or two got out, everyone else got hammered.

What makes this interesting? A couple of things. First, as markets are brought into tension with one another via pairs trading, they effectively create a new product, with new features and sometimes not-well-understood properties. A Collateralized Debt Obligation (CDO) can package together such disparate cash flows as sub-prime loans and airplane leases; by virtue of the CDO itself, these otherwise distinct streams are linked together. Second, as traders and funds are brought into tension with one another via sharing similar trading strategies, it creates collective action events that simply do not resemble atomized traders. It’s not surprising that the expression of these problems is always liquidity – IMHO that’s a fancy finance way of saying that people act in concert, when they are ’supposed’ to act according to their own preferences.

Some other interesting things from the Blow-up:
- it’s estimated that 38% of all equities are traded automatically, and that this percentage may rise to over 50% in the next three years;
- computers for some high-frequency traders execute hundreds of thousands of trades every day.

There is a giant blind spot here, rooted in financial theory, enabled by even cautious policy-makers and economists, and executed by very smart people. Sometimes at home when we watch commercials, my partner and I play a game called ‘good for them, or good for us’ – to see if a product or policy or feature is good for the company or good for consumers. I would guess that the number of people who would argue that derivatives trading today is ‘good for us’ is very small.

Comments (0)

November 6, 2007

prices and baselines, part 2

Filed under: Prices, Technology — Peter @ 11:02 am

In an earlier attempt to think through pricing, I was trying to understand the importance of public, baseline prices from which traders, investors, potential buyers and sellers could determine commodity prices. This leads me to discussions of ‘dark pools’ of liquidity..

(more…)

Comments (3)

October 29, 2007

Why exchanges shift from open outcry to electronic trading

Filed under: Prices, Technology — Peter @ 8:32 am

An article in the NYT on the shift away from open outcry today, which gets it right and gets it wrong. The right part is that the shift and merger of the CME and CBOT onto a single floor located at the CBOT spells the end for open outcry. It’s a natural transition point, and the break will break what’s left of pit trading. There will likely continue to be a floor for the largest volume contracts, but even those won’t likely survive for long. I’ve long been agnostic over whether or not the floors will die (historically, there were dire warnings about the death of open outcry about every 20 years since the 1950s). But this may well be it.

What I think the article gets wrong is the why. The imagery is one of the futility of human labor against the labor of a machine. As one trader notes, “Sometimes it feels like we’re John Henry going up against the steam hammer.” Kate Zaloom is quoted as noting that in electronic trading is “the idea of having a more pure market, one that doesn’t have the complications of flesh and blood.”

This idea, that technology displaces humans, is way too undifferentiated to be a useful explanation. Stuart Elliot’s research for the National Research Council, in part estimating the occupational displacements due to technology by 2030 (I include the key table at the end, just to demonstrate how scary-screwed many workers might be), show pretty wide heterogeneity across occupations. So one question unanswered by the John Henry argument is, why trading? Lawyers don’t seem to be going anyplace, and I’m not convinced that the technical requirements of work are great explanations.

The second problem is this idea of a ‘more pure market.’ It’s kind of BS. I did a literature review on what financial economists themselves suggested were the differences/comparisons/reviews of open outcry and electronic trading, and the results are decidedly mixed.

In the table, the research is placed on the side that is ‘better’ in the estimation of the author, using the measures they use – liquidity, transaction costs, obtaining adverse (that is, private) information. So a study on the open outcry side suggests that it’s better than electronic trading. My favorite finding is that when comparing the actual prices with the theoretical prices you should get if markets were perfectly efficient (or, more precisely, if markets followed the formulae exactly), both open outcry and electronic trading kind of suck.

So the question remains, why did electronic trading displace open outcry? I think the answer is pretty simple:
1) The clients changed. The modal trader in the 1970s was a high-income individual who was looking to increase returns to his (yes, his) investments via a riskier kind of financial investing. Commission costs, transactions costs, these are important, but the form of trading mattered little. At the end of the 1990s, it was estimated that something like 97% of the trading in financial futures came from institutions. Electronic trading is great for these clients. They bought their own seats, demanded a voice in decision-making. Their interests are different from both floor traders and from wealthy individual clients. It cannot be stressed enough that electronic trading is best for institutions who are able to capitalize on the kinds of things that electronic markets are good at – speed, cross exchange trades, digitized, already-model-manipulable data.

2) Electronic trading changed the products of the CME and CBOT. The exchanges’ products are contracts, liquidity, and prices. Electronic trading changed what counts as a price, so that while the two products look the same, they are not the same. The information that is captured in an electronic price is qualitatively different from the information that is captured in an open outcry price. They overlap, but they are distinct constellations of information. Here I agree most with Daniel Beunza and Yuval Millo, that electronic trading is losing information that might be useful in the change-over. But if that information is not useful to someone who could actually, you know, use it, it doesn’t really matter. In any case, the idea that electronic markets win out because they’re more ‘pure’ is just not correct.

I think there’s something sad about the passing of open outcry. I’ll miss the guy from the meats who sold beef jerky out of a duffel bag. And the quick wit and black humor. I also am not terribly sad. The guy who got women to give him blowjobs behind the desk, the references to Lind-Waldock as Lind-Welfare because they actually hired Black people, the eye-candy summer interns, all that is part of the open outcry world too.

[Source: Stuart W. Elliott, "Projecting the Impact of Computers on Work in 2030," p. 37, available online [PDF].] (thanks, Alex)

Comments (0)

October 26, 2007

Markets and visualization

Filed under: Technology — Peter @ 5:37 pm

Daniel’s discussion of visualization is interesting, and it’s something I’ve had on my mind since reading Tad Williams’ Otherland series (the mysterious Mr. Sellars has a virtual garden that provides the GUI in an astoundingly complex virtual environment).

I think Daniel conflates two separate issues. The context is that market visualizations continue to be conservative, with the ubiquitous stock chart and even the ticker continuing to be the standards. This despite innovations all over the place in other kinds of visualizations.

The two issues, I think, are 1) what data should be contained in a market visualization; and 2) how might these visualizations be implemented into existing market trading praxis. The first is a question of information, and the second is a question of innovation, related but distinct.

So first, imagine two different visualizations, one a traditional chart that includes a set of quality measures that don’t traditionally make into market information (say, the number of times a firm’s CEO had family relations that week) and the other a new way of visualizing the same information everyone else uses. A good example of this second type is Bashiba (here’s a completely fascinating .wmv video link). Bashiba takes volume, volatility, prices, and maps them to the size of waves, the music, the weather. A listen and glance can tell you all the information at once.

What I think would be a disaster is to try to incorporate both at once. That kind of wholesale innovation is nice to think about, but wholesale revolution occurs less often than one might think, even (or especially) in finance.

So, the second part seems quite a bit like an institutional innovation problem to me. Normally, per someone like Clemens, the argument would go something like saying that innovation works if it is introduced within the currently agreed-upon form. So might seem like something new, but it is recognizable as something familiar. Particularly in the trading world, where for all the hoopla over the new masters of quant, many traders prefer three columns of numbers for buy-side, sell-side, and current trade, the recognizability has to be pretty immediate, and pretty accurate. A colleague, Brad Paley, makes his living making it possible for traders to learn the screen by “grabbing it with their eyes,” as one person describes it. He makes a good case for innovating by reproducing the familiar in new medium – not by copying a notebook electronically per se, but by making his handhelds enough notebook-ish that they are easily understood.

I’m not sure why Daniel believes that a ‘low-end entrant’ to the industry would lead the way in innovation – other than suggesting that innovations come from the outside, I’m not sure what the basis might be for that. I would think that innovation will come when traders recognize that the ‘new data environment’ actually gives them better means to act decisively than their old means. This suggests a generational shift in traders more than a low-end industry outsider.

Comments (0)

April 17, 2007

Paper: Information, Prices, and Sensemaking

Filed under: Technology — Peter @ 12:29 pm

Information, Prices, and Sensemaking. I’m trying to make sense of the underlying ontology of price, though I don’t call it that in this particular paper.

Comments (0)

Futures Trading, Technology, and Institutional Change

Filed under: Technology — Peter @ 12:24 pm

Summary: A study of the institutional change associated with the shift from open outcry to electronic trading in financial futures. I conducted an in-depth ethnography of a face-to-face trading organization, comparing it to 4 electronic trading firms. My key finding is that what appears to be a technological change in how traders make markets is actually a deep, institutional change in the actors, objects, and activities involved in making markets. Electronic trading changes what counts as information, the expertise of traders, and the relationships between individuals and the market.

Electronic trading is to open outcry what the internet is to the telephone. Not just a more efficient way to do the same thing, but potentially a whole new thing altogether…

Comments (0)