\begin{landscape}
\begin{table}[htbp]
\centering
\small
\caption{Rolling Realized Beta Change Regressions (30-min Window)}
\label{tab:rolling_delta_beta_30min}
\begin{threeparttable}
\begin{tabular}{lcccccc}
\toprule
 & VIX Futures & 10Y T-Note & 5Y T-Note & Dollar Index & Crude Oil WTI & Gold \\
\midrule
Intercept & -0.2364 & 0.0098 & -0.0025 & 0.0058 & -0.0933 & 0.1275* \\
 & (0.1651) & (0.0192) & (0.0134) & (0.0074) & (0.0650) & (0.0758) \\
Stance & -0.0190 & 0.0118 & 0.0195 & 0.0137 & -0.0008 & 0.1244 \\
 & (0.2653) & (0.0198) & (0.0158) & (0.0093) & (0.0891) & (0.0869) \\
Novelty & 0.1096 & -0.0096 & -0.0064 & -0.0086 & -0.1290 & -0.0539 \\
 & (0.0917) & (0.0102) & (0.0073) & (0.0075) & (0.1020) & (0.0344) \\
Stance $\times$ Novelty & -0.0818 & 0.0064 & 0.0074 & 0.0011 & -0.1772 & -0.0852 \\
 & (0.1018) & (0.0118) & (0.0074) & (0.0080) & (0.2106) & (0.0755) \\
\midrule
$N$ & 71 & 88 & 88 & 88 & 88 & 87 \\
$R^2$ & 0.009 & 0.003 & 0.008 & 0.045 & 0.064 & 0.047 \\
Adj.\ $R^2$ & -0.036 & -0.033 & -0.027 & 0.010 & 0.031 & 0.013 \\
\bottomrule
\end{tabular}
\begin{tablenotes}[flushleft]
\small
\item \textit{Notes:} Dependent variable: $\Delta\hat{\beta}$ (change in rolling realized beta relative to ES, 30-min window). Coefficients are in beta units; a coefficient of $0.12$ means that a one-standard-deviation increase in the regressor is associated with a $0.12$ increase in co-movement with the S\&P~500. All regressors z-scored. Clustered SE in parentheses. Stars: BH-adjusted $p$-values. Semantic measures from MiniLM--BERT ensemble.
\item $^{***}p<0.01$, $^{**}p<0.05$, $^{*}p<0.10$.
\end{tablenotes}
\end{threeparttable}
\end{table}
\end{landscape}