Another Nobel for Time Series Econometrics?

Thomson Reuters makes annual Nobel Prize forecasts in chemistry, physics, medicine and economics, based on citation counts from its Web of Science database (no surprise). Of course the exercise is largely a marketing tool for their database, but it's still fun and timely.

They boast that their algorithm "has accurately forecast 27 winners since its inception in 2002." It's not clear how to interpret that statement, and whether the algorithm's performance is nearly as impressive as the statement suggests. (Each year they forecast several prizes, typically shared two or three ways, in each of the fields. If I'm guessing correctly, they score a "correct forecast" whenever the name of any current winner appeared at any earlier time as part of any of their forecasted prizes.)

Anyway, as I said, it's still fun and timely. This year they identify three top candidate prizes in economics:

David Hendry, Hashem Pesaran and Peter Phillips: For their contributions to economic time-series, including modeling, testing and forecasting

Josh Angrist, David Card and Alan Krueger: For their advancement of empirical microeconomics

Sam Peltzman and Richard Posner: For extending economic theories of regulation.

Quite a fine collection!

Sheldon Hackney: A Truly Great Penn Man

Sheldon Hackney, Penn's president 1981-1993, recently passed away. See the fine coverage in the Almanac and Daily Pennsylvanian.

In my younger days as a Penn undergrad, Hackney took a lot of abuse. People felt that he didn't have much backbone. Exhibit 1 was always his failure to stand up to the water buffalo thing, letting political correctness run amok. I too felt that way.

But now, with the benefit of more information providing breadth and depth of hindsight, I see that I was wrong. Yes, he botched the water buffalo thing. But now I see that as just a small detour in a heroic career.

Thanks, Sheldon, for seeing things clearly in Alabama, for taking your case to Penn, the nation, and the world, for testifying to the truth, day in and day out. Thanks for relishing the joy of being a professor first and always. Thanks for engaging Penn's neighbors, and for laying the groundwork for Penn's eastward expansion. And thanks for your famous civility, your stunning grace under pressure, a model for us all. I recall the Fleetwood Mac lyric, from just slightly before your Penn arrival, "Can I sail through the changing ocean tides, can I handle the seasons of my life?" May I do half as well as you.

Deep Inside the LSE

Mary Morgan, Professor of History and Philosophy of Economics at LSE (Department of Economic History) is giving a talk at Penn today, in our Department of History and Sociology of Science. She's done interesting work in the history of econometrics, and more recently in the methodology of economics. We've never met, so I'm looking forward to her talk.

The occasion led me to pop into the LSE web site, which was fun because I haven't been at LSE in ages. Of course one is aware of the LSE leaders in one's own fields (for me, econometrics, economics, finance), but one forgets that there's much, much more there. Look below at all the topics (and click to see the people) stuffed into the LSE. Truly unique! Basically anything that would fit in The Economist fits somewhere at LSE. Even their logos look similar, with white letters on a bright red background. Must be a British thing.

• Accounting
• Anthropology
• Economics
• Economic History
• European Institute
• Finance
• Gender Institute
• Geography & Environment

• Government
• International Development
• International History
• International Relations
• Language Centre
• Law
• Management
• Mathematics

• Media & Communications
• Methodology
• Philosophy, Logic and Scientific Method
• Public Affairs (IPA)
• Social Policy
• Social Psychology
• Sociology
• Statistics

Theory gets too Much Respect, and Measurement Doesn't get Enough (60-Second Lecture Video and Transcription)

In the last post, I told you about Penn's "60-Second Lecture". Mine is now completed, and we had a good time. Watch the video, and you'll have a good laugh at the unflattering opening shot, complete with a barking dog in the background, and the blinding sun in my face throughout. A rough transcription follows.

Science is advanced by just two things, measurement and theory. Their interplay pushes science forward, as each disciplines the other.

Some people believe that good research requires tightly-integrated measurement and theory, present in equal amounts.

I submit to you, first, that such views are both naive and false. Measurement and theory are rarely advanced at the same time, by the same team, in the same work. And they don't need to be. Instead we exploit the division of labor, as we should. Measurement can advance significantly with little theory, and theory can advance significantly with little measurement. Still each disciplines the other in the long run, and science advances.

And I submit to you, second, and primarily, and perhaps provocatively, that theory gets too much respect in science, and that measurement doesn't get enough. A wry observer once remarked that theorists typically have the top-floor offices, while experimentalists and statisticians are tucked away in the basement. But Lord Kelvin got it right more than a century ago, when he argued that measurement is the essence of science. And moreover, theory is largely data distillation, attempting to eliminate noise and isolate signal, and hence dependent on measurement. The theory mill grinds data.

Of course it's futile, and it's not at all my intent, to assert that one or the other of measurement and theory is somehow uniquely important, or even relatively more important. Each is equally important, and each most definitely needs the other. But again, that's the point: given their truly equal importance, they should get equal attention and respect. Instead theory gets too much, and measurement doesn't get enough. Perhaps that will change in the emerging age of Big Data.

60 Seconds on Measurement vs. Theory in all the Sciences

At Penn we have something of a unique tradition, an ongoing series of "60-Second Lectures." Four Penn profs speak each semester, outdoors, at noon on four consecutive Wednesdays. And yes, each of the four lectures is capped at 60 seconds.

60-Second Lectures sound insane. But somehow they work. They're on Locust Walk, the massively-trodden main drag. There's a podium and lights and a microphone, and some people actually stop to listen, and it's all quite an entertaining spectacle, even if 60-Second Cage Fights would be much better. The video dudes record it, and then it's posted for the masses in Penn's online archives as well as its Facebook, YouTube, and Vimeo pages. (Vimeo? Who knows...) Then the video goes completely viral, catapulting the 60-Second Lecturer to unprecedented heights of global intellectual celebrity. At least that's what the Dean's Office tells me. Or something like that. Anyway, you surely see through it. Only a complete sucker would agree to attempt a 60-Second Lecture. Yes, that's right, I agreed.

But seriously, the 60-Second Lectures are a wonderful Penn tradition, and I'm delighted to contribute my minute's worth. My title will be "Measurement vs. Theory in all the Sciences." Join us if you're in Philadelphia. It's Wednesday, September 18, 11:55 sharp. (If you're more than a minute late, you'll miss everything!) It's at Stiteler Plaza, 37th and Locust Walk, unless the cheering throng exceeds 35,000, in which case we move to Franklin Field.

Forecast-Error Insurance

Yale economists Mark Rosenzweig and Chris Udry have a new paper, "Forecasting Profitability." It's related to a project of theirs that examines the Indian Meteorological Department's annual monsoon rainfall forecasts and their effects on farmer profitability. (Farmers who rely on the forecast to guide their planting-stage investments, and many do, are exposed not only to rainfall risk, but also to "forecast-error risk.")

Mark raised an interesting related question in an email a few weeks ago, "Do you know of any insurance products for forecasts of any kind? Can one purchase insurance that indemnifies one against forecast errors, anywhere in the world for any forecasts?"

Hmm...quite fascinating! There are financial products insuring against rainfall events (e.g., see the discussion my 2005 JASA paper with Sean Campbell, "Weather Forecasting for Weather Derivatives"), but what about rainfall forecast-error events?

Of course the "forecast-error insurance" issue transcends rainfall. More generally, who might demand forecast-error insurance and why? Only users of forecasts? Perhaps also producers of forecasts (not unlike medical malpractice insurance)? Who might supply it, and how might it be priced (very tricky...)?

Here are some thoughts.

First, forecast object x and the associated forecast error e = x - x_f will generally be correlated, in which case insurance against x events is also partial insurance against e events. (Note well, however, that the implicit e insurance is only partial, and perhaps very partial, depending on the correlation strength).

Second, assuming that the relevant financial markets exist, one could construct an e hedge by holding an appropriate portfolio of x-sensitive stocks, which in an efficient market would reflect discounted x forecasts and hence move only due to "news" (e). Long or short positions in that portfolio would hedge against positive e or negative e. Better yet, one could implement the hedge using stock options rather than the underlying, going long a call or long a put (or both, a so-called "straddle," which would hedge against large e of either sign).

Maybe I've missed something? Are the questions well-posed, and if so, are there simpler or more obvious answers? In any event, many assumptions obviously lurk behind the thoughts / suggestions above, but nothing is impossible for the man who doesn't have to do it himself.

Tom Sargent, Quantitative Economics, and Python

Speaking of Tom Sargent, check out his latest at http://www.quant-econ.net/.  (Thanks to Frank DiTragila for forwarding a few days ago.)  Python features prominently...

Fed Chair Desiderata

I'm so bored with the endless Larry and Janet show.  Finally some real wisdom, from John Cochrane. Tom Sargent has my vote!

Is Economics too Important for Economists?

Like piranha fish in a feeding frenzy, different research tribes fight furiously to stake claims in new areas like financial engineering and risk management. Fringe players, in particular, often strive to move toward the center, or to redefine the center in ways that feather their nests.

Such competition is desirable, but healthy competition is based on merit, not mudslinging. Hence my disappointment when watching the video preview for "Financial Engineering and Risk Management Part I," a massively open online course (MOOC) by Martin Haugh and Garud Iyengar (H&I) at Columbia, to be given soon on Coursera. H&I come from Industrial Engineering and Operations Research, and they conclude their sales pitch with the brazen proclamation, "... it's often said that economics is too important to be left to economists. Well, we feel the same way about finance and financial engineering. It's too important to be left to economists..."

Wow, strong words. So what's in their syllabus? Here it is:

- Introduction to forwards, futures and swaps
- Introduction to options and the 1-period binomial model
- The multi-period binomial model and risk-neutral pricing
- Term structure models and pricing fixed income derivative securities
- Introduction to credit derivatives
- Introduction to mortgage mathematics and mortgage-backed securities

Huh? What? Isn't that largely financial economics, pioneered and continuously refined by an ongoing parade of financial economists? Of course. Indeed what else could it be?

A quick glance at the web indicates that H&I's research is high-quality, and I hope that the same will be true for their course. (I have registered.) I'm also glad that H&I are contributing to the wonderful Coursera MOOC phenomenon, and I applaud their declared desire to increase lay financial literacy. But I suggest that they and their tribe give credit where credit is due -- is that not necessary for true literacy? -- and think twice before glibly biting the financial economics hand that feeds them.

More on Heteros*edasiticity

Thanks to Rikky Vohra for bringing still more to my attention. Absolutely classic!