Estimation of Precautionary Demand Caused by Financial Shocks

Abstract

Pioneering work of modelling financial anxieties was given by Kimura et al. in 1999 as psychological change of people due to financial shocks. They regressed financial position by nonstationary interest rate and estimated the variance of financial shocks in a heteroskedasticity frame work of TGARCH model. They regarded a variance of negative shocks as financial anxieties. Furthermore, with precautionary demand replaced by financial anxieties, they insisted that the cointegration property among (real GDP, money, share price, financial anxieties) still holds even in 1999 after the financial crisis of 1997 and 1998, while cointegration does not hold without a variable 'financial anxieties' in 1999. However, their estimation exhibits a strong influence of positive financial shocks in a bubble economy of the period [1987,1989] and it is difficult to say 'financial anxieties'. Extended data after 1999 further stressed the influence of positive shocks in a bubble economy. We improved their results by introducing a growth rate system in EGARCH model. Furthermore, estimating separately each variance of positive and negative financial shocks, we can show that precautionary demand is increased by negative shocks and decreased by positive ones. Money adjusted by precautionary demand thus obtained is shown to satisfy cointegration relationship in (adjusted money, real GDP, interest rate) even in 2003.