This paper extends the Bayesian semiparametric stochastic volatility (SV‐DPM) model. Instead of using a Dirichlet process mixture (DPM) to model return innovations, we use an infinite hidden Markov model (IHMM). This allows for time variation in the return density beyond that attributed to parametric latent volatility. The new model nests several special cases as well as the SV‐DPM. We also discuss posterior and predictive density simulation methods for the model. Applied to equity returns, foreign exchange rates, oil price growth and industrial production growth, the new model improves density forecasts, compared with the SV‐DPM, a stochastic volatility with Student's innovations and other fat‐tailed volatility models.

中文翻译：

---

具有随机波动性的无限隐马尔可夫模型

本文扩展了贝叶斯半参数随机波动率（SV-DPM）模型。我们没有使用狄利克雷混合过程 (DPM) 来模拟回报创新，而是使用无限隐马尔可夫模型 (IHMM)。这允许回报密度的时间变化超出参数潜在波动率造成的变化。新模型嵌套了几个特殊情况以及 SV-DPM。我们还讨论了模型的后验和预测密度模拟方法。与 SV-DPM、Student 创新的随机波动率和其他厚尾波动率模型相比，新模型应用于股票回报、外汇汇率、油价增长和工业生产增长，改进了密度预测。