1\begin{table}[tp]2\centering3\footnotesize4\begin{threeparttable}5\caption{Intraday spanning: in-sample $R^2$ of each held-out futures root on the first three principal components of the other five roots, 1-, 5- and 30-minute returns, 2019--2026}6\label{tab:spanning}7\setlength{\tabcolsep}{3pt}8\begin{tabular}{lrrrrrr}9\toprule10Held-out root & \multicolumn{2}{c}{1 minute} & \multicolumn{2}{c}{5 minutes} & \multicolumn{2}{c}{30 minutes} \\ \cmidrule(lr){2-3}\cmidrule(lr){4-5}\cmidrule(lr){6-7} & 08--16 ET & All hours & 08--16 ET & All hours & 08--16 ET & All hours \\11\midrule12ZT & 0.54 & 0.48 & 0.72 & 0.66 & 0.81 & 0.77 \\13ZF & 0.71 & 0.66 & 0.88 & 0.83 & 0.95 & 0.92 \\14ZN & 0.69 & 0.63 & 0.90 & 0.85 & 0.97 & 0.96 \\15TN & 0.80 & 0.76 & 0.93 & 0.91 & 0.97 & 0.96 \\16US & 0.72 & 0.67 & 0.90 & 0.86 & 0.95 & 0.94 \\17UB & 0.71 & 0.66 & 0.85 & 0.81 & 0.92 & 0.90 \\18\midrule Share of first three PCs & 0.969 & 0.963 & 0.989 & 0.985 & 0.995 & 0.994 \\19Observations & 895,677 & 1,772,160 & 191,899 & 502,927 & 32,063 & 91,939 \\20\bottomrule21\end{tabular}22\begin{tablenotes}[flushleft]\footnotesize23\item \textit{Notes:} Appendix table. For each root, the first three principal components of the other five roots' returns (sampled at 1, 5 or 30 minutes, 2019--2026, all six contracts trading) are used to predict the held-out root's return in sample; the table reports the $R^2$. ``08--16 ET'' (``US hours'' in the text) restricts to intervals starting between 08:00 and 15:59 Eastern time; ``All hours'' includes the overnight session. The last rows give the share of total variance explained by the first three principal components of all six roots (three decimals because the values are close to one) and the number of return intervals. No standard errors are reported.24\end{tablenotes}25\end{threeparttable}26\end{table}27