What is behind the magic of O-Score? An alternative interpretation of Dichev’s (1998) bankruptcy risk anomaly