Research

Refereed Publications

1. Estimation and inference for higher-order stochastic volatility models with leverage, with Jean-Marie Dufour & Md. Nazmul Ahsan, Journal of Time Series Analysis, vol. 46, no. 6: 1064–1084.
   Slides
   Short Abstract

   We propose simple, efficient moment-based estimators for higher-order stochastic volatility models with leverage [SVL(p)] that require only a few moment equations and support simulation-based Monte Carlo tests. Applied to daily returns on three major U.S. indices (S&P 500, Dow Jones, Nasdaq), SVL(p) models deliver more accurate volatility forecasts than competing specifications.
   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

Conference Proceedings

2. Simulation-Based Inference for the Synchronization of Business Cycles, with Jean-Marie Dufour, JSM Proceedings, Business and Economic Statistics Section. Toronto, ON: American Statistical Association, December 2023.
3. Simulation-Based Inference for Markov Switching Models, with Jean-Marie Dufour, JSM Proceedings, Business and Economic Statistics Section. Washington, D.C.: American Statistical Association, December 2022.

Working Papers

4. Monte Carlo Likelihood Ratio Tests for Markov Switching Models, with Jean-Marie Dufour
   Revised February 2026 | BoC SWP | Slides
   Short Abstract

   We develop likelihood ratio tests for Markov switching models that are valid in finite samples and robust to the identification problems common in this setting, covering multiple regimes and multivariate, non-stationary, and non-Gaussian models. We illustrate them on U.S. GNP growth and on Markov switching VARs testing business-cycle synchronization.
   Presentations

   Econometric Society World Congress, CIREQ Econometrics Conference: Recent Developments in Identification and Inference, Universidad del Rosario Seminar, 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, 17th CIREQ Ph.D. Students' Conference, 56th Annual (2022) Meetings of the Canadian Economics Association
5. Underlying Core Inflation with Multiple Regimes
   Revised November 2024 | arXiv | Slides
   Short Abstract

   We build a core inflation indicator from a high-dimensional factor model with multiple regimes, addressing the failure of standard measures to signal underlying inflation when it rose in 2021. Allowing for regime changes yields a simple real-time indicator that reduces revisions, improves headline-inflation forecasts, and is robust to transitory and sector-specific shocks.
   Presentations

   New York Camp Econometrics XIX, 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
6. Implied Factors: The Linear Skeleton of Machine Learning Forecasts, with Philippe Goulet Coulombe & Karin Klieber
   Draft and slides coming soon
   Short Abstract

   Any supervised learning algorithm whose predictions are linear in the target variable—regardless of nonlinearities in the features—admits a linear-factor representation, and a small set of these implied factors closely tracks its forecasts. Empirically, two or three implied factors reproduce out-of-sample ML model forecasts of U.S. GDP, inflation, and unemployment, matching or outperforming PCR, PLS, sPCA, SPCA, and SsPCA.
   Presentations

   BSE Summer Forum: Workshop on AI and Machine Learning in Economics, ECONDAT 2026 Spring Meeting (Bank of France), 60th Annual (2026) Meetings of the Canadian Economics Association, Bank of Canada Fellowship Learning Exchange
7. MSTest: An R-package for Testing Markov-Switching Models, with Jean-Marie Dufour
   Revised December 2025 | BoC SWP | arXiv (2024)
   Short Abstract

   MSTest is an R package implementing hypothesis tests for the number of regimes in Markov switching models—including Monte Carlo likelihood ratio, moment-based, parameter-stability, and standard likelihood ratio tests—together with tools to simulate and estimate univariate and multivariate Markov switching and hidden Markov processes.
8. wARMASVp: Estimation, Inference, Filtering, and Forecasting for Higher-Order Stochastic Volatility Models in R, with Jean-Marie Dufour & Md. Nazmul Ahsan
   Revised May 2026 | CRAN | Reference manual
   Short Abstract

   wARMASVp is an R package for estimation, testing, filtering, and forecasting in univariate higher-order stochastic volatility SV(p) models with Gaussian, Student-t, or GED innovations and optional leverage. Estimation uses closed-form Winsorized ARMA-SV moment estimators (no numerical optimization), with Local and Maximized Monte Carlo tests and three filters—the first such package in any language with heavy tails and leverage.
9. Volatility Forecasting with Higher-order Stochastic Volatility Models, with Jean-Marie Dufour & Md. Nazmul Ahsan
   Revised June 2026
   Short Abstract

   We evaluate multi-horizon volatility forecasts from higher-order stochastic volatility [SV(p)] models, estimated by a closed-form moment-based method with no numerical optimization. Across twenty international equity indices (2001–2020), we compare them against GARCH, heterogeneous autoregressive, and ten machine learning methods. Second- and third-order SV models consistently give the lowest losses—15–29% below the best traditional benchmark, with larger gains at longer horizons.
   Presentations

   IAAE 2025 Annual Conference
10. Estimation and inference for stochastic volatility models with leverage and heavy-tailed distributions, with Jean-Marie Dufour & Md. Nazmul Ahsan
    Revised June 2026 | BoC SWP
    Short Abstract

    We propose closed-form moment-based W-ARMA estimators for higher-order stochastic volatility SV(p) models with Student-t or GED innovations and leverage, with √T-consistency, asymptotic normality, and exact Monte Carlo tests for normality, heavy tails, and leverage. Applied to three U.S. equity indices (2000–2025), heavy-tailed SV with leverage is favored and gives the best variance forecasts and VaR/ES calibration.
11. Joint Determination of Counterparty and Liquidity Risk in Payment Systems, with Jorge Cruz Lopez & Charles M. Kahn
    Revised September 2024 | Slides | Awarded Best Paper on Risk Management at the NFA 2019 Conference
    Short Abstract

    We study how banks jointly manage funding liquidity and counterparty risk in an interbank payment system. Using intraday data from the Canadian Large Value Transfer System, we show banks coordinate payment timing to manage both risks, with incentives rising in risk exposures and funding costs—and disruptions to this coordination can amplify systemic risk.
    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

Policy & Technical Notes

12. Practical estimation of high-dimensional stochastic volatility models with application to macroeconomic uncertainty in Québec and Canada, with Jean-Marie Dufour & Md. Nazmul Ahsan, CIRANO, September 2025.
13. The Government of Canada Debt Securities Dataset, with Jeffrey Gao & Francisco Rivadeneyra, Bank of Canada, Technical Report No. 112, February 2018.

Works in Progress

• Practical and reliable estimation methods for high-dimensional multivariate stochastic volatility models with macroeconomic applications, with Jean-Marie Dufour & Md. Nazmul Ahsan
• Monte Carlo Test for Factor Models with Markov switching
• MNbreaks: An R Package for Estimating and Testing Multiple Structural Changes in Multivariate Linear Regression Models, with Pierre Perron & Zhongjun Qu