0497836 20250807152307.620181210235959.9 85045622136 000447768700002 10.1007/s10258-018-0145-5 11 s. P cav_un_epca*03117551617-982X173 (2018)173-183Springer Capital asset pricing model Detrended cross-correlation analysis Detrending moving-average cross-correlation analysis Fractal regressions Portugal cav_un_auth*0256902 Ekonometrie Department of Econometrics E E Department of Econometrics řé Krištoufek Ladislav UTIA-B CZ Ústav teorie informace a automatizace AV ČR, v. v. i. cav_un_auth*0351446 Ferreira P. PT http://library.utia.cas.cz/separaty/2018/E/kristoufek-0497836.pdf GJ17-12386Y GA ČR CZ cav_un_auth*0351447 We examine risk profiles of the Portuguese stock market index component stocks using a novel approach to the classical capital asset pricing model (CAPM). Specifically, we estimate the CAPM via fractal regressions that allow studying the marginal effects at selected scales. In this way, we can reveal whether the risk is perceived differently by market participants with different investment horizons. Apart from the analysis itself, we provide new statistical insights into the issue of separating and comparing the scale-specific effects with statistical validity. We find several stocks deviating from an expected risk perception homogeneity across investment horizons. This is true for both analysed periods, i.e. before and after the global financial crisis. There are also several stocks that changed their relationship to the market portfolio in between, which has strong implications for possible portfolio construction. The pro- posed methodology is not limited to financial topics but can be used in any discipline where the scale-specific marginal effects might be of interest. WOS AH 50000 50200 50202 2019 2 RVO:67985556 http://hdl.handle.net/11104/0290650 S 2 Article Economics ECONOMICS 0.444 0.222 16.1 0.00006 0.108 179 0.147 0.500 9 Q4 5.236 11.7 Q4 Q4 0.32 Q3 11.708 2018 85045622136 SCOPUS PUBMED 000447768700002 WOS cav_un_epca*0311755 Portuguese Economic Journal 1617-982X 1617-9838 Roč. 17 č. 3 2018 173 183 Springer