There is considerable experimental evidence that there is a divergence between willingness-to-accept compensation to give up a good and willingness-to-pay to obtain a good. This divergence persists even when the good in question in small relative to income, a result in apparent conflict with standard economic theory. This paper develops a theoretical bidding model with costly information acquisition to explain this divergence. The model generates a gap between offers to sell and bids to buy consistent with the experimental results. We argue that the model does a better job of explaining experimental data than either of the two commonly invoked theoretical explanations: the endowment effect and the substitution effect.