Nonlinear modeling of oil and stock price dynamics: segmentation or time-varying integration?

In this paper, we show the usefulness of the switching transition error correction model in reproducing the bilateral linkages between oil and stock markets over the last three decades. Our findings show that while linear models fail to apprehend significant relationships between oil and stock markets, the hypothesis of financial and oil markets integration cannot be rejected using nonlinear cointegration models. More interestingly, this cointegration relationship is represented by an on-going process partially activated per regime when oil price deviations move away from their equilibrium with stock prices and exceed some threshold.