Regimes, Non-Linearities, and Price Discontinuities in Indian Energy Stocks

We construct a representative index of largest Indian energy companies listed on the National Stock Exchange (NIFTY 50). We test for presence of regimes, non-linearities, and jumps in the price signal. We benchmark performance against alternative models, including single-regime models and models with no jumps. We then benchmark the quality of regime identification against other indices examined in the literature, such as Nikkei 225 and FTSE 100. Overall, find that our regime-switching model performs well in identifying the regimes in this comparative setting. Based on our model selection criteria, we prefer a regime-augmented model to a model that allows no regime identification. But overall, we prefer a model with jumps and regimes over those that do not allow for jump-diffusion and Markov regime-switching.