I don’t think that’s true. Marginal utility can explain the directional component but not the reference point component.
Let’s take the example from the comment above. A car salesman gets $100 in the morning and has to give it back in the evening if he doesn’t sell enough cars. Compare that to just handing out $100 if he’s successful in the evening. From an expected utility view both of these arrangements are equivalent.
Loss aversion argues that they’re not, because the salesman’s reference point for loss/gain shifts after obtaining ownership of the $100.
Let’s take the example from the comment above. A car salesman gets $100 in the morning and has to give it back in the evening if he doesn’t sell enough cars. Compare that to just handing out $100 if he’s successful in the evening. From an expected utility view both of these arrangements are equivalent.
Loss aversion argues that they’re not, because the salesman’s reference point for loss/gain shifts after obtaining ownership of the $100.