Jordi Llorens-Terrazas
@jordi_llorens_t
Followers
62
Following
44
Media
0
Statuses
11
Lecturer in Econometrics @economicssurrey
Joined July 2022
Fantastic line-up of seminar speakers for the upcoming term @economicssurrey !
After a rich Fall 2023 seminar programme, our list of Spring 2024 econ external speakers @economicssurrey @UniOfSurrey is even richer!!! Check it out! https://t.co/nZJcsphw3j! TamonAsonuma RaffaellaGiacomini @jgrogger @TomoakiYamada3 @PaolaManzini ElisaFaraglia @chris_starmer
0
0
2
Extremely excited about this and looking forward to joining you soon 😀
We are celebrating a successful round of junior recruitment. We welcome @jordi_llorens_t, @AlistairMacaul7, @JCostasFer in September and https://t.co/vNXHZsca3Z in 2024 after a year as a postdoc at Oxford.
2
0
19
IMF growth-at-risk framework indicates that downside risks are now back in the red zone (worst 20th percentile of the last four decades) due to tighter financial conditions and extraordinary uncertainty. Read more in the Global Financial Stability Report: https://t.co/wBATTUt7B8
6
56
207
What is a standard error? Check out the @JEconometrics Special Session at the 2023 @ASSAMeeting in New Orleans, featuring @jmwooldridge and @elie_tamer, among others!
5
42
360
Do not contribute to misinformation… Stay away from MDPI journals. Source for so much noise. The infamous Lockdowns-do-not-work “meta”-study gave 92% weight to a single MDPI paper.
2
1
8
(3/3) We put the Projected DCC to work on an empirical application to portfolio selection using the constituents of the S&P 100 index, and find that the performance is superior to the standard DCC.
0
0
0
(2/3) We ask the question: given Q (positive definite but not unit diagonal), what is the closest correlation matrix to Q? This is known as a matrix-nearness problem. We replace rescaling in the DCC with this solution, and call this "Projected DCC".
1
0
0
(1/3) In the standard Dynamic Conditional Correlation model (DCC), the sequence of (pseudo) correlation matrices needs to be rescaled so that it has ones on its main diagonal. This is typically done via rescaling. While this is natural, it can be improved.
1
0
0
Happy to share that the article "Projected Dynamic Conditional Correlations" with Christian Brownlees has been accepted for publication in the International Journal of Forecasting (link to wp version: https://t.co/SYs0eww9ka). Check it out, looking forward to your comments!
1
0
2