- Estimation and inference for higher-order stochastic volatility models with leverage, with Jean-Marie Dufour & Md. Nazmul Ahsan, revised & resubmitted, Journal of Time Series Analysis
We propose efficient and simple estimators for higher-order stochastic volatility models with leverage [SVL(p)], based on a small number of moment equations. The computationally simple estimators proposed allow one to easily perform simulation-based (possibly exact) tests. The methods developed are applied to daily returns for three major stock indices (S&P 500, Dow Jones, Nasdaq), confirming the superiority of SVL(p) models over competing conditional volatility models in terms of forecast accuracy.
Conference & Seminar presentations: 39th Annual Meeting of the Canadian Econometrics Study Group, North American Summer Meeting of the Econometric Society, 58th Annual (2024) Meetings of the Canadian Economics Association, 63e Congrès de la Société Canadienne de Science Économique, CIREQ Econometrics Conference in Honor of Eric Ghysels
- Underlying Core Inflation with Multiple Regimes
This paper utilizing procedures for detecting multiple regimes in high-dimensional factor models to introduce a new approach for estimating core inflation indicators. When inflation began to increase in 2021, many core inflation indicators failed at provding an appropriate and timely signal of underlying inflation. We solve this problem by considering time-varying parameters that account for changes in inflation regimes. The result is a simple real-time indicator that is useful in guiding monetary policy. It reduces historical revisions, provides improved forecasts of headline inflation, and is robust to transitory and sector-specific shocks.
Conference & Seminar presentations: Bank of Canada Conference on Real-Time Data Analysis, Methods and Applications in Macroeconomics and Finance, IAAE 2024 Annual Conference, Bank of Canada Brown Bag Seminar, 57th Annual (2023) Meetings of the Canadian Economics Association
- MSTest: An R-package for Testing Markov-Switching Models, with Jean-Marie Dufour
Revised November 2024 |
arXiv The R package MSTest implements hypothesis testing procedures to identify the number of regimes in Markov switching models. These models have wide-ranging applications in economics, finance, and numerous other fields. The MSTest package includes the Monte Carlo likelihood ratio test procedures of Rodriguez-Rondon and Dufour (2024), the moment-based tests of Dufour and Luger (2017), the parameter stability tests of Carrasco, Hu, and Ploberger (2014), and the likelihood ratio test of Hansen (1992). Additionally, the package enables users to simulate and estimate univariate and multivariate Markov switching and hidden Markov processes, using the expectation-maximization (EM) algorithm or maximum likelihood estimation (MLE).
- Monte Carlo Likelihood Ratio Tests for Markov Switching Models, with Jean-Marie Dufour
Job Market Paper | Revised October 2024 |
Slides This paper proposes a likelihood ratio test for Markov switching models that extends to a broader range of settings not previously addressed in the literature. This includes multiple regimes as well as multivariate, non-stationary, and non-Gaussian settings, which are common and useful in applied work. Importantly, the approach is valid for finite samples, which is relevant for many macroeconomic applications that utilize quarterly data, and it is robust to the identification problems often encountered in Markov switching models. Two macroeconomic applications are considered: one in a univariate setting using U.S. GNP growth data and another in a multivariate context involving Markov switching VAR models and testing for the synchronization of business cycles. These tests also have broader applications within macroeconomics and finance including indentification of SVAR models and causal relationships.
Conference & Seminar presentations: CIREQ-McGill Seminar, 76th European meeting of the Econometric Society, New York Camp Econometrics XVIII, Carleton University Brown Bag Seminar, NBER-NSF Time Series Conference, IAAE 2023 Annual Conference, Boston University Econometrics Seminar, 16th International Conference on Computational and Financial Econometrics, Latin American Meeting of the Econometric Society, Joint Statistical Meetings of the American Statistical Association, 56th Annual (2022) Meetings of the Canadian Economics Association
- Volatility Forecasting with Higher-order Stochastic Volatility Models, with Jean-Marie Dufour & Md. Nazmul Ahsan
Revised August 2023
We study the performance of higher-order stochastic volatility [SV(p)] models in forecasting volatility. Using different volatility proxies (squared returns and the realized volatility) of various international indices, we conduct two out-of-sample forecast experiments: (1) we forecast a moderately volatile period after the late-2000s financial crisis; (2) we forecast a highly volatile period, i.e., the core financial crisis. We compare the accuracy of volatility forecasts among SV(p) models, GARCH models, and Heterogenous Autoregressive model of Realized Volatility (HAR-RV) models. The results suggest that SV(p) models perform better than other models in most cases.
- Joint Determination of Counterparty and Liquidity Risk in Payment Systems, with Jorge Cruz Lopez & Charles M. Kahn
Revised September 2023 |
Slides | Awarded
Best Paper on Risk Management at the NFA 2019 Conference.
We investigate how banks jointly manage their funding liquidity and counterparty risk in the context of an interbank payments system. Using intra-day data from the Canadian Large Value Transfer System, we show that banks coordinate the issuance of payment orders to jointly manage their liquidity and counterparty risk. Coordination incentives increase with risk exposures and the cost of funding. We conclude that coordination disruptions may increase risk exposures that lead to funding constraints and systemic risk. Therefore, coordination should be considered in policies aimed at enhancing financial stability.
Conference & Seminar presentations: II Regional Conference on Payments and Financial Market Infrastructures, Banco de la República de Colombia & CEMLA, Payments Canada & Bank of Canada Research Symposium, 51st Annual (2017) Meetings of the Canadian Economics Association
- mbreaks: R Package for Estimating and Testing Multiple Structural Changes in Linear Regression Models, with Linh Nguyen, Pierre Perron, & Yohei Yamamoto
Revised January 2023
Th mbreaks R package to implements testing procedures to detect multiple structural changes in the coefficients and variance of linear regression models proposed by Bai and Perron (1998, 2003) and Perron, Yohei Yamamoto and Jing Zhou (2020). The package provides methods of constructing the confidence intervals of the break dates, testing for the presence of structural changes and selecting the number of structural changes.
