Short and long-run dependence in Swedish stock returns

The behaviour of Swedish stock returns over short and long-run horizons is analysed. Using monthly data from 1919 to 1995 and, weekly and daily data for the 1980s and first part of the 1990s little evidence of long-run dependence was found. Using three different tests that are robust to short-term dependence and conditional hetroscedasticity it was found that the modified R/S (rescaled range) test and ARFIMA-GARCH tests provided no support for long-run memory in Swedish stock returns. Only the fractional differencing test, GPH, gave a significant result in two cases: for monthly real and nominal stock returns for the full and the first half of the sample at rather high frequency for the spectral analysis.