Loss aversion without the endowment effect, and other explanations for the WTA-WTP disparity