Do decreasing hazard functions of price durations make any sense