How long-term contracts can mitigate inefficient renegotiation arising due to loss aversion

A loss averse buyer and seller face an uncertain environment. Should they write a long-term contract or wait until the state of the world has realized? I show that simple long-term contracts perform better than insinuated in Herweg and Schmidt (2015), even though loss aversion makes renegotiation sometimes inefficient. During renegotiation, the outcome induced by the long-term contract constitutes the reference points to which the parties compare gains and losses of the renegotiated transaction. Whereas Herweg and Schmidt assume that the long-term contract is always performed, it is actually preferably that it is not in a "bad" state. This alters the threat-point in renegotiation, making it easier to renegotiate and thus improves the performance of long-term contracts. Specific performance contracts perform better than in Herweg and Schmidt but are still problematic since one can not always prevent that the contract is enforced when it should not. Option contracts perform much better since only one party has the ex post trade decision and it is thus much easier to prevent that the contract is inefficiently enforced due to loss aversion. My findings suggest that loss aversion alone cannot explain why parties sometimes abstain from writing beneficial long-term contracts, but gives important insights on how long-term contracts should be written when parties are aware they are loss averse.