Studies in time series analysis of consumption, asset prices and forecasting

Abstract

Sisällysluettelo: Kari Takala Studies in time series analysis of consumption, asset prices and forecasting 11 Kari Takala The consumption function in Finland revisited; An error-correction model for Finnish consumption 37 Kari Takala Excess sensitivity testing of consumption with Finnish data 79 Kari Djerf - Kari Takala Macroeconomy and consumer sentiment: Performance of the Finnish consumer barometer after ten years 119 Kari Takala - Pekka Pere Testing the cointegration of house and stock prices in Finland 151 Bharat Barot - Kari Takala House prices and inflation: A cointegration analysis for Finland and Sweden 173 Kari Takala - Matti Virén Bankruptcies, indebtedness and the credit crunch 211 Kari Takala - Matti Virén Testing nonlinear dynamics, long-memory and chaotic behavior with financial and nonfinancial data 241 Tiivistelmä: This collection of seven papers deals with three different areas of econometric applications: consumption, asset prices, and forecasting.The papers apply techniques related to the analysis of unit roots and cointegration methods.The first paper deals with consumption theories and formulates an error-correction forecasting model for consumption.A single cointegration relationship is found between consumption, income and net wealth, which is in line with the permanent income hypothesis.The second paper studies the excess sensitivity of consumption to current disposable income. Estimating the coefficient with timevarying techniques, we notice a decline in the coefficient during the period of financial deregulation toward the end of the 1980s and a rise during the recession.Third paper takes a closer look at how useful consumer barometer variables can be in forecasting variables such as consumption and inflation. The first paper on asset prices, is based on the theory of cointegration between house and stock prices, which asserts that real after-tax risk-adjusted returns on assets should coincide in the long run.This paper presents a model for house prices that uses stock prices as a leading indicator to improve the forecasting of housing prices.Another paper on asset prices considers cointegration between house prices and inflation, and finds eg that house prices adjust to consumer prices in the long run and that no excess real appreciation, apart from rental income, is derived from house ownership.The two last papers deal with bankruptcy forecasting and testing for nonlinearities and chaos.It is asserted that bankruptcies can be interpreted as error-correction between supply and demand.Many tests have been developed to study the presence of nonlinearities in economic series.The results of testing unambiguously support that there are strong nonlinearities in economic data, but the evidence for chaos is weak.