Pepe Montiel Olea

Local projections are surprisingly robust to misspecification. By contrast, VAR confidence intervals with short-to-moderate lag length can severely undercover for misspecification that is small, difficult to detect, and cannot be ruled out a priori. econometricsociety.org/publications/e…

Applications for the Machine Learning in Economics Summer Institute (MLESI) are open! Grad students, apply and come hang out in Chicago for a week to learn about ML and Econ.

The Econometric Society: Announcing a new interdisciplinary conference on Economics and AI+ML econometricsociety.org/society/news/A…

@CavaliereGiu @UCSD That is my favorite ECMA paper of all times! I wish I had met him! (And thanks for all your great posts!)

This 41 y.o. classic #ECMA article by 𝐇𝐚𝐥 𝐖𝐡𝐢𝐭𝐞 (@UCSD) is definitely one of my favorite #econometrics papers. Short, focused, pathbreaking, well written, disruptive, theoretically super interesting, relevant in applications. Simply amazing! jstor.org/stable/1912526

@TobiasSalz @SloanFoundation @ShengwuLi @instrumenthull @mdeshecon @guoxu_econ @alexoimas Great news, Tobi! Congratulations to you and the rest of this amazing group!

🙏 @SloanFoundation for this honor! Congratulations to @ShengwuLi, @instrumenthull, @mdeshecon, @guoxu_econ, Mikkel, @alexoimas, and Yueran, and of course all the other amazing researchers outside of economics.

@RichRHC @dix_rafael ¡Muchas felicidades, mi Rich! ¡Te mando un gran abrazo!

I'm thrilled to see this paper forthcoming! Check out the very nice thread by @dix_rafael

@CavaliereGiu @JEconometrics Thanks for the tweet, Giuseppe! We had a lot of fun writing this! We started working on this project after reading this amazing paper from the late 80's: dl.acm.org/doi/10.1145/76…

Hi #EconTwitter! Describing a parameter region can be complicated. In this very nice @JEconometrics paper, Pepe MO @PpM_O and Nesbit show that random sampling methods can be seen as #supervised #MachineLearning. Cool for those doing VARs. sciencedirect.com/science/articl… #Econometrics

@fkocheng @LudvigsonSydney ¡Qué orgullo, Fede! ¡Felicidades! (Congrats to NYU as well for training such amazing scholars!)

@LudvigsonSydney Thank you very much, Sydney!

Congratulations to Federico Kochen (NYU PhD class of 2022) on winning the ECB Young Economist Prize! ecb.europa.eu/pub/conference…

@jmwooldridge @BruceEHansen [Extra] The posterior mean estimator corresponding to the Beta(\sqrt{n}/2, \sqrt{n}/2) can be shown to be minimax (under quadratic loss). From a decision-theoretic perspective, I find it hard to think about why to use any other estimator for this problem :)

@jmwooldridge @BruceEHansen 3) The posterior mean corresponding to any prior in the beta family can be written as a linear combination between the MLE and the prior mean. In my opinion, the prior that has a good finite-sample "frequentist" interpretation is the Beta(\sqrt{n}/2, \sqrt{n}/2).

Here's something I've wondered about off and on. For some reason, sitting in an airport facing a two-hour delay has made me think about it. Suppose I have X ~ Bernoulli(p) and I put the Uniform(0,1) prior on p.

@jiafengchen42 @economeager @c_r_walt The same argument for prediction can be used for estimation. See the short discussion in Section 2.4.

@jiafengchen42 @economeager @c_r_walt The classes under consideration fix an arbitrary distribution for the covariates, and then allow for different assumption on the residuals. Theorem 1 shows minimaxity of OLS (without necessarily assuming Gaussianity of the residuals or covariates) [2/3]

If you guys think all this negative weight stuff is bad, wait until #EconTwitter discovers the James/Stein (1961) result that OLS is inadmissible under squared loss when m>2

@y_shapovalova @PierreAlquier @CFECMStatistics @ngocmath1 @statsmartin Thanks for organizing, Pierre!!!

@PierreAlquier @CFECMStatistics @ngocmath1 @statsmartin Thanks for organizing Pierre!

@roydanroy @PierreAlquier It is a very elegant and interesting paper! Congrats!

7 years of work has come to fruition: Our paper on extended admissibility and its equivalence to "nonstandard" Bayes optimality is now published at the Annals. We paid for Open Access so that anyone can download it immediately, for free. Enjoy! projecteuclid.org/journals/annal…

@JlibDoesEcon "and music history is filled with examples of works now famous which were initially declared unplayable by lazy or intimidated virtuosos" 😄😁

@JlibDoesEcon One interesting comment about the rejection by Rory Guy (this appears in the back on the vinyl of the concert Perlman-Ormandy Philadelphia Orchestra): "In retrospect, the objections ... may be partly accounted for. For the musicians, there were taxing technical demands ..."

After getting a sneak peek at a new piano concerto, the great pianist Nikolai Rubinstein told the composer it was “worthless, absolutely unplayable; the passages so broken, so disconnected, so unskillfully written, that they could not even be improved… bad, trivial, common.”