% ============================================================================
% Highlights — separate file required at Elsevier submission
% (3-5 bullets, max 85 characters each, results-oriented)
% Revised 2026-10-17 (version 2.1 of the working paper)
% ============================================================================
\documentclass[12pt]{article}
\usepackage[utf8]{inputenc}
\usepackage[T1]{fontenc}
\begin{document}
\section*{Highlights}
\begin{itemize}
  \item LLM and vision models turn listing text and photos into auditable hedonic regressors
  \item Two AI layers add 3.8 points of $R^2$ to a rich hedonic model; 3.0 come from the photo
  \item Façade-vision features lift every estimator, more than switching OLS to a forest
  \item Curb appeal, materials and luxury cues price sensibly under clustered and bootstrap SEs
  \item FSBO descriptions embed the asking price: text gains are credible only on the MLS
\end{itemize}
\end{document}