LAUSR

We estimate the relationship between high‐temperature exposure and land transaction using daily transaction data from 2013 to 2018 in China. Standard hedonic price method is used to perform multidimensional fixed effects regression on 458,564 transaction samples. This study shows that each additional day of the extremely high temperature increases the average bid price by 0.6%, which is equivalent to an additional annual increase of 15.018 billion yuan in land transaction costs. The effect of high temperature on land prices is considerably stronger for residential and commercial lands. We further find that high‐temperature‐induced changes of mood and cognition play a role. Extreme gradient boosting (XGBoost) algorithm is used to calculate price deviation, which overcome missing eigenvalues and reduce nonlinear measurement error. Our findings indicate that heat can make investors more aggressive, which occurs when bidders continue to raise prices to win. The number of bidding behavior will not drop due to the hidden costs of land transactions. Specially, bidding experience offsets the cognitive output bias. This study contributes to the behavioral finance and decision‐making literature, helps investors make better investment management decisions, and alleviates the negative effect of land transaction premium on the real estate market.