Options
Ardia, David
Nom
Ardia, David
Affiliation principale
Fonction
Ancien.ne collaborateur.trice
Identifiants
Résultat de la recherche
Voici les éléments 1 - 3 sur 3
- PublicationMétadonnées seulementLarge scale portfolio optimization with DEoptim(Boca Raton, Florida: BrownWalker Press, 2014)
; ;Boudt, Kris ;Mullen, KatharinePeterson, Brian - PublicationMétadonnées seulementThe short-run persistence of performance in funds of hedge fundsThere is extensive empirical evidence that funds of hedge funds (FoHFs) quickly change their investment bets as a function of the changing market conditions. In this chapter, we first analyze the stability of risk exposure and performance of FoHFs during the period January 2005–June 2011. We then study the short-run persistence of performance in the FoHFs industry. Past performance is measured using the 1-year trailing return as well as risk-adjusted measures such as the Sharpe ratio and the fund’s alpha based on the Carhart (1997) or Fung and Hsieh (2004) factor models. Over the examined timeframe, we consistently find that using risk-adjusted return measures improves the risk-adjusted performance of the momentum investment strategy. This finding holds for the financial crisis period as well as the pre- and post-crisis periods.
- PublicationAccès libreEfficient Bayesian estimation and combination of GARCH-type modelsThis chapter proposes an up-to-date review of estimation strategies available for the Bayesian inference of GARCH-type models. The emphasis is put on a novel efficient procedure named AdMitIS. The methodology automatically constructs a mixture of Student-t distributions as an approximation to the posterior density of the model parameters. This density is then used in importance sampling for model estimation, model selection and model combination. The procedure is fully automatic which avoids difficult and time consuming tuning of MCMC strategies. The AdMitIS methodology is illustrated with an empirical application to S&P index log-returns where non-nested GARCH-type models are estimated and combined to predict the distribution of next-day ahead log-returns.