Working Papers
Efficient Estimation and Filtering for Multivariate Jump-Diffusions (with Gustavo Schwenkler), Journal of Econometrics. Codes
This paper develops estimators of the transition density, filters, and parameters of multivariate jump-diffusions. The drift, volatility, jump intensity, and jump magnitude are allowed to be state-dependent and non-affine. Our density and filter estimators converge at the canonical square-root rate, implying computational efficiency. Our parameter estimators have the same asymptotic properties as true maximum likelihood estimators, implying statistical efficiency. The results of this paper enable the empirical analysis of previously intractable models of asset prices and economic time series.
Stochastic Volatility Models with Markov-Switching Jump Parameters, September 2015
In this paper, I examine continuous-time stochastic volatility models with jumps in returns and volatility in which the parameters governing the jumps are are allowed to switch according to a Markov chain. I estimate the parameters and the latent processes for the S&P 500 and the Nasdaq indices from 1990 to 2014. The Markov-switching parameters characterize well the periods of market stress, such as those in 1997-1998, 2001 and 2007-2010. Furthermore, I find that jumps account for a larger part of the total variation of returns when allowing jump parameters to be Markov-switching. In turbulent periods, stochastic volatility contributes less to the observed variation in returns. Several statistical tests favor the model with Markov-switching jump parameters. These results provide empirical evidence of the state-dependent and time-varying nature of asset price jumps, a feature of asset prices that has been recently documented in high-frequency data. The models are estimated by MCMC methods.
Affine Stochastic Volatility Models with Regime-Switching (with Fan Zhuo), March 2016
This paper provides tools to estimate Markov-Switching continuous-time affine stochastic volatility models with jumps in returns and volatility, where the jumps parameters are not Markov-switching. The estimation is performed via MCMC, which allows to obtain the latent processes induced by the modeling. Furthermore, we propose some misspecification tests and develop a Markov-switching test based on the Odds ratios. We estimate the parameters and the latent processes for the S&P 500 and the Nasdaq indices from 1990 to 2014. We show that the S& P500 is the only index exhibiting a Markov-switching behavior, even though plots of the Markov chain for the Nasdaq deceivingly show the contrary.
Optimal Pricing in Media Revenue Management (Master's thesis - in French), August 2010
TV Channels face many problems when they sell their inventory to media agencies. They have to choose which commercials should be aired, and simultaneously determine the prices of these commercials, while staying competitive. This master's thesis provides a bilevel programming model to tackle this revenue management problem. The novelty of this modeling lies in the introduction of patterns of commercial spots. As in the cutting stock problem, I use patterns to "cut" media inventories. This way, a pattern can be considered as a set of commercials, and TV channels sell selection of commercial spots. I develop an algorithm to solve the problem, based on well chosen pre-processive cuts. My algorithm allows to optimally schedule and price patterns of commercial spots for a typical prime time on TV.