Title: Estimation of smoothly time varying coefficient partial adjustment model
Authors: Kosuke Oya - Osaka University (Japan) [presenting]
Abstract: The partial adjustment model can capture how financial asset prices adjust to information. Speeds of adjustment towards intrinsic values of asset prices are the measures of the degrees of over and under-reactions for new information. The conventional partial adjustment model has a constant coefficient which reflects the adjustment speed. Although we can see the degrees of adjustments speed through the estimated coefficients, it is hard to see how adjustment speeds change during the estimation period. To cope with the difficulty, we introduce the partial adjustment model with smoothly time varying coefficient. The time varying scheme is established through the same way as the smooth transition autoregressive model developed in the previous studies. For empirical application, the price discovery and tatonnement process during the preopening period in the Tokyo stock exchange is focused. Using the time varying adjustment coefficient, we can find whether market discovers the equilibrium price of asset until the market opens.