|
on Econometric Time Series |
| By: | Fabrizio Iacone (Universita degli Studi di Milano); Morten Ørregaard Nielsen (Queen's University and CREATES); A.M. Robert Taylor (University of Essex) |
| Abstract: | Lobato and Robinson (1998) develop semiparametric tests for the null hypothesis that a series is weakly autocorrelated, or I(0), about a constant level, against fractionally integrated alternatives. These tests have the advantage that the user is not required to specify a parametric model for any weak autocorrelation present in the series. We extend this approach in two distinct ways. First, we show that it can be generalised to allow for testing of the null hypothesis that a series is I(\delta) for any \delta lying in the usual stationary and invertible region of the parameter space. Second, it is well known in the literature that long memory and level breaks can be mistaken for one another, with unmodelled level breaks rendering fractional integration tests highly unreliable. We therefore extend the Lobato and Robinson (1998) approach to allow for the possibility of changes in level at unknown points in the series. We show that the resulting statistics have standard limiting null distributions, and that the tests based on these statistics attain the same asymptotic local power functions as infeasible tests based on the unobserved errors, and hence there is no loss in asymptotic local power from allowing for level breaks, even where none is present. We report results from a Monte Carlo study into the finite-sample behaviour of our proposed tests, as well as several empirical examples. |
| Keywords: | fractional integration, level breaks, Lagrange multiplier testing principle, spurious long memory, local Whittle likelihood, conditional heteroskedasticity |
| JEL: | C22 |
| Date: | 2020–06 |
| URL: | http://d.repec.org/n?u=RePEc:qed:wpaper:1431&r=all |
| By: | Miranda Gualdrón, Karen Alejandra; Ruiz Ortega, Esther; Poncela Blanco, Maria Pilar |
| Abstract: | Dynamic Factor Models, which assume the existence of a small number of unobservedlatent factors that capture the comovements in a system of variables, are the main "bigdata" tool used by empirical macroeconomists during the last 30 years. One importanttool to extract the factors is based on Kalman lter and smoothing procedures that cancope with missing data, mixed frequency data, time-varying parameters, non-linearities,non-stationarity and many other characteristics often observed in real systems of economicvariables. This paper surveys the literature on latent common factors extracted using Kalmanfilter and smoothing procedures in the context of Dynamic Factor Models. Signal extractionand parameter estimation issues are separately analyzed. Identi cation issues are also tackledin both stationary and non-stationary models. Finally, empirical applications are surveyedin both cases. |
| Keywords: | State-Space Model; Identi Cation; Em Algorithm; Dynamic Factor Model |
| Date: | 2020–06–25 |
| URL: | http://d.repec.org/n?u=RePEc:cte:wsrepe:30644&r=all |
| By: | Bao-Gen Li; Dian-Yi Ling; Zu-Guo Yu |
| Abstract: | When common factors strongly influence two cross-correlated time series recorded in complex natural and social systems, the results will be biased if we use multifractal detrended cross-correlation analysis (MF-DXA) without considering these common factors. Based on multifractal temporally weighted detrended cross-correlation analysis (MF-TWXDFA) proposed by our group and multifractal partial cross-correlation analysis (MF-DPXA) proposed by Qian et al., we propose a new method---multifractal temporally weighted detrended partial cross-correlation analysis (MF-TWDPCCA) to quantify intrinsic power-law cross-correlation of two non-stationary time series affected by common external factors in this paper. We use MF-TWDPCCA to characterize the intrinsic cross-correlations between the two simultaneously recorded time series by removing the effects of other potential time series. To test the performance of MF-TWDPCCA, we apply it, MF-TWXDFA and MF-DPXA on simulated series. Numerical tests on artificially simulated series demonstrate that MF-TWDPCCA can accurately detect the intrinsic cross-correlations for two simultaneously recorded series. To further show the utility of MF-TWDPCCA, we apply it on time series from stock markets and find that there exists significantly multifractal power-law cross-correlation between stock returns. A new partial cross-correlation coefficient is defined to quantify the level of intrinsic cross-correlation between two time series. |
| Date: | 2020–05 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2006.09154&r=all |
| By: | Nicklas Werge (LPSM); Olivier Wintenberger (LPSM) |
| Abstract: | The Quasi-Maximum Likelihood (QML) procedure is widely used for statistical inference due to its robustness against overdisper-sion. However, while there are extensive references on non-recursive QML estimation, recursive QML estimation has attracted little attention until recently. In this paper, we investigate the convergence properties of the QML procedure in a general conditionally heteroscedastic time series model, extending the classical offline optimization routines to recursive approximation. We propose an adaptive recursive estimation routine for GARCH models using the technique of Variance Targeting Estimation (VTE) to alleviate the convergence difficulties encountered in the usual QML estimation. Finally, empirical results demonstrate a favorable trade-off between the ability to adapt to time-varying estimates and stability of the estimation routine. |
| Date: | 2020–06 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2006.02077&r=all |
| By: | Claudia Foroni; Massimiliano Marcellino; Dalibor Stevanovic |
| Abstract: | We consider simple methods to improve the growth nowcasts and forecasts obtained by mixed frequency MIDAS and UMIDAS models with a variety of indicators during the Covid-19 crisis and recovery period, such as combining forecasts across various specifications for the same model and/or across different models, extending the model specification by adding MA terms, enhancing the estimation method by taking a similarity approach, and adjusting the forecasts to put them back on track by a specific form of intercept correction. Among all these methods, adjusting the original nowcasts and forecasts by an amount similar to the nowcast and forecast errors made during the financial crisis and following recovery seems to produce the best results for the US, notwithstanding the different source and characteristics of the financial crisis. In particular, the adjusted growth nowcasts for 2020Q1 get closer to the actual value, and the adjusted forecasts based on alternative indicators become much more similar, all unfortunately indicating a much slower recovery than without adjustment and very persistent negative effects on trend growth. Similar findings emerge also for the other G7 countries. |
| Keywords: | COVID-19,Mixed-Frequency,Forecasting,GDP, |
| Date: | 2020–06–10 |
| URL: | http://d.repec.org/n?u=RePEc:cir:cirwor:2020s-32&r=all |
| By: | Linton, O.; Li, S. |
| Abstract: | We carry out some analysis of the daily data on the number of new cases and the number of new deaths by (191) countries as reported to the European CDC. We work with a quadratic time trend model applied to the log of new cases for each country. This seems to accurately describe the trajectory of the epidemic in China. We use our model to predict when the peak of the epidemic will arise in terms of new cases or new deaths in other large countries. |
| Keywords: | COVID-19, Forecasting, Virus |
| JEL: | C13 C22 C10 |
| Date: | 2020–04–02 |
| URL: | http://d.repec.org/n?u=RePEc:cam:camdae:2025&r=all |
| By: | Jie Fang; Jianwu Lin |
| Abstract: | One of the major characteristics of financial time series is that they contain a large amount of non-stationary noise, which is challenging for deep neural networks. People normally use various features to address this problem. However, the performance of these features depends on the choice of hyper-parameters. In this paper, we propose to use neural networks to represent these indicators and train a large network constructed of smaller networks as feature layers to fine-tune the prior knowledge represented by the indicators. During back propagation, prior knowledge is transferred from human logic to machine logic via gradient descent. Prior knowledge is the deep belief of neural network and teaches the network to not be affected by non-stationary noise. Moreover, co-distillation is applied to distill the structure into a much smaller size to reduce redundant features and the risk of overfitting. In addition, the decisions of the smaller networks in terms of gradient descent are more robust and cautious than those of large networks. In numerical experiments, we find that our algorithm is faster and more accurate than traditional methods on real financial datasets. We also conduct experiments to verify and comprehend the method. |
| Date: | 2020–06 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2006.09247&r=all |
| By: | Juan Carlos Escanciano |
| Abstract: | This paper provides new uniform rate results for kernel estimators of absolutely regular stationary processes that are uniform in the bandwidth and in infinite-dimensional classes of dependent variables and regressors. Our results are useful for establishing asymptotic theory for two-step semiparametric estimators in time series models. We apply our results to obtain nonparametric estimates and their rates for Expected Shortfall processes. |
| Date: | 2020–05 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2005.09951&r=all |