Tuesday, July 22, 2014

Chinese Diebold-Rudebusch Yield Curve Modeling and Forecasting

A Chinese edition of Diebold-Rudebusch, Yield Curve Modeling and Forecasting: The Dynamic Nelson-Siegel Approach, just arrived. (I'm traveling -- actually at IMF talking about Diebold-Rudebusch among other things -- but Glenn informed me that he received it in San Francisco.) I'm not even sure that I knew it was in the works. Anyway, totally cool. I love the "DNS" ("Dynamic Nelson-Siegel") in the Chinese subtitle. Not sure how/where to buy it. In any event, the English first chapter is available free from Princeton University Press, and the English complete book is available almost for free (USD 39.50 -- as they used to say in MAD Magazine: Cheap!).

Sunday, July 20, 2014

Some History of NBER Econometrics

My last post led me to reminisce. So, for history buffs, here's a bit on the origins and development of the NBER working group on Forecasting and Empirical Methods in Macro and Finance. (No promises of complete accuracy -- some of my memory may be fuzzy.)

In the early 1990's Steve Durlauf had an idea for an "Empirical Methods in Macro" NBER group, and he asked me to join him in leading it. Bob Hall kindly supported the idea, so we launched. Some years later Steve stepped down, Ken West joined, and we decided to add "Finance." I was also leading a "Forecasting" group with highly-related interests, so we merged the two, and Diebold-Durlauf "Empirical Methods in Macro" then became Diebold-West, "Forecasting and Empirical Methods in Macro and Finance." Quite a mouthful, but it worked!

We met at least once per year at the NBER Summer Institute, sometimes more. Papers drawn from the meetings sometimes appeared as journal symposia. I'm particularly fond of those in International Economic Review (1998, 811-1144), which contains Andersen-Bollerslev on realized volatility from underlying diffusions, Rudebusch on measuring monetary policy in VAR's (with Sim's discussion and Rudebusch's feisty rejoinder), Christoffersen on interval forecast calibration, Diebold-Gunther-Tay on density forecast calibration, among others, and Review of Economics and Statistics (1999, 553-673), which contains Baxter-King on bandpass filters for business-cycle measurement, Kim-Nelson on measuring changes in business-cycle stability using Bayesian dynamic-factor Markov-switching models, and Gallant-Hsu-Tauchen on range-based asset return volatility estimation, among others.

I eventually stepped down around 2005, and Mark Watson joined. (Mark and I had earlier edited another group symposium in Journal of Applied Econometrics (1996, 453-593).) So Diebold-West became Watson-West, and the group continued to thrive. In 2013, Mark and Ken passed their batons to Allan Timmermann and Jonathan Wright, who are off and running. This summer's program was one of the best ever, and the meeting was heavily over-subscribed.

Tuesday, July 15, 2014

Time to Re-Think NBER Programs?

Check out John Cochrane's recent NBER post if you didn't already. It ends with:
A last thought. Economic Fluctuations [an NBER program] merged with Growth [another NBER program] in the mid 1990s. At the time there was a great confluence of method as well as interest. Growth theorists were studying growth with Bellman equations, dynamic general equilibrium models of innovation and transmission of ideas, thinking about where productivity shocks came from. Macroeconomists were using Bellman equations and studying dynamic general equilibrium models with stochastic technology, along with various frictions and other propagation mechanisms. 
That confluence has now diverged. ...  ...when Daron Acemoglu, who seems to know everything about everything, has to preface his comments on macro papers with repeated disclaimers of lack of expertise, it's clear that the two fields [fluctuations and growth] really have gone their separate ways. Perhaps it's time to merge fluctuations with finance, where we seem to be talking about the same issues and using the same methods, and to merge growth with institutions and political or social economics.

[Material in square brackets and bold added by me.]
I agree. Fluctuations and finance belong together. (I'm talking about asset pricing broadly defined, not corporate finance.) Yes, the methods are basically the same, and moreover, the substance is inextricably linked. Aspects of fluctuations are effectively the fundamentals priced in financial markets, and conversely, financial markets can most definitely impact fluctuations. (Remember that little recession a few years back?)

The NBER Summer Institute group in which I most actively participate, Forecasting and Empirical Methods in Macroeconomics and Finance (one of several so-called working groups under the umbrella of the NBER's program in Economic Fluctuations and Growth), has been blending fluctuations and finance for decades. But we're intentionally narrowly focused on applied econometric aspects. It would be wonderful and appropriate to see broader fluctuations-finance links formalized at the NBER, not just at the working group level, but also at the program level.

Sunday, June 29, 2014

ADS Perspective on the First-Quarter Contraction

Following on my last post about the first-quarter GDP contraction, now look at the FRB Philadelphia's Aruoba-Diebold-Scotti (ADS) Index. 2014Q1 is the rightmost downward blip. It's due mostly to the huge drop in expenditure-side GDP (GDP_E), which is one of the indicators in the ADS index. But it's just a blip, nothing to be too worried about. [Perhaps one of these days we'll get around to working with FRB Philadelphia to replace GDP_E with GDPplus in the ADS Index, or simply to include income-side GDP (GDP_I) directly as an additional indicator in the ADS Index.]

Plot of ADS Business Conditions Index in 2007

Source: FRB Philadelphia

One might wonder why the huge drop in measured GDP_E didn't cause a bigger drop in the ADS Index. The reason is that all real activity indicators are noisy (GDP_E is just one), and by averaging across them, as in ADS, we can eliminate much of the noise, and most of the other ADS component indicators fared much better. (See the component indicator plots.)

Note well the important lesson: both the ADS Index (designed for real-time analysis of broad real activity) and GDPplus (designed mostly for historical analysis of real GDP, an important part of real activity) reduce, if not eliminate, measurement error by "averaging it out."

All told, ADS paints a clear picture: conditional on the underlying indicator data available now, real growth appears to be typical (ADS is constructed so that 0 corresponds to average growth) -- not especially strong, but simultaneously, not especially weak.

Friday, June 27, 2014

The First Quarter GDP Contraction was Less Severe than you Think



As discussed in an earlier post, my co-authors and I believe that our "GDPplus," obtained by optimally blending the noisy expenditure- and income-side GDP estimates, provides a superior U.S. GDP measure. (Check it out online; the Federal Reserve Bank of Philadelphia now calculates and reports it.) A few days ago we revised and re-posted the working paper on which it's based (Aruoba, Diebold, Nalewaik, Schorfheide, and Song, "Improving GDP Measurement: A Measurement Error Perspective," Manuscript, University of Maryland, Federal Reserve Board and University of Pennsylvania, Revised June 2014).

It's important to note that GDPplus is not simply a convex combination of the expenditure- and income-side estimates; rather, it is produced via the Kalman filter, which averages optimally over both space and time. Hence, although GDPplus is usually between the expenditure- and income-side estimates, it need not be. Presently we're in just such a situation, as shown in the graph below. 2014Q1 real growth as measured by GDPplus (in red) is well above both of the corresponding expenditure- and income-side GDP growth estimates (in black), which are almost identical. 
Plot of GDPplus
Source:  FRB Philadelphia



Thursday, June 19, 2014

Fine Work by Mueller and Watson at ECB


The ECB's Eurotower in Frankfurt, Germany
ECB Eurotower

Ulrich Mueller and Mark Watson's "Measuring Uncertainty About Long-Run Predictions" is important and original. I like it more (and understand it more) every time I see it. The latest was last week in Frankfurt at the ECB's Eighth Annual Workshop on Forecasting. No sense transcribing my discussion; just view it directly.

Wednesday, June 18, 2014

Windows File Copy

Estimation
Of course we've all wondered for decades, but during the usual summertime cleanup I recently had to copy massive numbers of files, so it's on my mind. Seriously, what is going on with the Windows file copy "remaining time" estimate? Could an average twelve-year-old not code a better algorithm? (Comic from XKCD.)








Saturday, June 14, 2014

Another 180 on Piketty's Measurement

My first Piketty Post unabashedly praised Piketty's measurement (if not his theory):
"Piketty's book truly shines on the data side. ... Its tables and figures...provide a rich and jaw-dropping image, like a new high-resolution photo of a previously-unseen galaxy. I'm grateful to Piketty for sending it our way, for heightening awareness, and for raising important questions."  
Measurement endorsements don't come much stronger.

Then I did a 180. Upon belatedly reading the Financial Times' Piketty piece, I felt I'd been had, truly had. Out poured my second Piketty Post, written in a near-rage, without time to digest Piketty's response.

Now, with the benefit of more time to read, re-read, and reflect, yes, I'm doing another 180. It seems clear that the bulk of the evidence suggests that the FT, not Piketty, is guilty of sloppiness. Piketty's response is convincing, and all-told, his book appears to remain a model of careful social-science measurement (thoughtful discussion, meticulous footnotes, detailed online technical appendix, freely-available datasets, etc. -- see his website).

Ironically, then, as the smoke clears, my first Piketty post remains an accurate statement of my views.

Monday, June 9, 2014

Piketty's Empirics Are as Bad as His Theory



In my earlier Piketty post, I wrote, "If much of its "reasoning" is little more than neo-Marxist drivel, much of its underlying measurement is nevertheless marvelous." The next day, recognizing the general possibility of a Reinhart-Rogoff error, but with no suspicion that that anything was actually remiss, I added "(assuming of course that it's trustworthy)."

Perhaps I really should read some newspapers. Thanks to Boragan Aruoba for noting this, and for educating me. Turns out that the Financial Times -- clearly a centrist publication with no ax to grind -- got hold of Piketty's data (underlying source data, constructed series, etc.) and published a scathing May 23 indictment.

The chart above -- just one example -- is from The Economist, reporting on the FT piece. Somehow Piketty managed to fit the dark blue curves to the light blue dots of source data. Huh? Sure looks like he conveniently ignored a boatload of recent data that happen to work against him. Put differently, his fits appear much more revealing of his sharp prior view than of data-based information. Evidently he forgot to talk about that in his book.

In my view, Reinhart-Rogoff was a one-off and innocent (if unfortunate) mistake, whereas the FT analysis clearly suggests that Piketty's "mistakes," in contrast, are systematic and egregious.

Saturday, May 31, 2014

More on Piketty -- Oh God No, Please, No...

Cover: Capital in the Twenty-First Century in HARDCOVER

Piketty, Piketty, Piketty! How did the Piketty phenomenon happen? Surely Piketty must be one of the all-time great economists. Maybe even as great as Marx.
Yes, parts of the emerging backlash against Piketty's Capital resonate with me. Guido Menzio nails its spirit in a recent post, announcing to the Facebook universe that he'll "send you $10 and a nice Hallmark card with kitties if you refrain from talking/writing about Piketty's book for the next six months." (The irony of my now writing this Piketty post has not escaped me.)
As I see it, the problem is that Piketty's book is popularly viewed as a landmark contribution to economic theory, which it most definitely is not. In another Facebook post, leading economic theorist David Levine gets it right:
People keep referring to economists who have favorable views of Piketty's book. Leaving aside Krugman, I would be interested in knowing the name of any economist who asserts that Piketty's reasoning ... is other than gibberish.

    So the backlash is focused on dubious "reasoning" touted as penetrating by a book-buying
    public that unfortunately can't tell scientific wheat from chaff. I'm there.

But what of Piketty's data and conclusion? I admire Piketty's data -- more on that below.  I also agree with his conclusion, which I interpret broadly to be that the poor in developed countries have apparently become relatively much more poor since 1980, and that we should care, and that we should try to understand why. 
In my view, Piketty's book truly shines on the data side. If much of its "reasoning" is little more than neo-Marxist drivel, much of its underlying measurement is nevertheless marvelous (assuming of course that it's trustworthy). Its tables and figures -- there's no need to look at anything else -- provide a rich and jaw-dropping image, like a new high-resolution photo of a previously-unseen galaxy. I'm grateful to Piketty for sending it our way, for heightening awareness, and for raising important questions. Now we just need those questions answered.